Harvard Law Review Harvard Law Review Harvard Law Review

Administrative Law

The Presumption of Regularity in Judicial Review of the Executive Branch

The Trump Administration’s ban on entry into the United States by nationals from several majority-Muslim countries1×1. The latest version of the policy is Enhancing Vetting Capabilities and Processes for Detecting Attempted Entry into the United States by Terrorists or Other Public-Safety Threats, Proclamation No. 9645, 82 Fed. Reg. 45,161 (Sept. 24, 2017). sparked controversy about when courts should credit the executive branch’s factual assertions about its motives and decisionmaking processes. When a plaintiff alleges that the government skirted procedures or acted on illicit motives, courts will sometimes “presume” that “official duties” have been “properly discharged”2×2. United States v. Chem. Found., Inc., 272 U.S. 1, 15 (1926). until the challenger presents “clear evidence to the contrary.”3×3. Id. at 14–15. This “presumption of regularity” has common law origins.4×4. See, e.g., Chicago, B. & Q. Ry. v. Babcock, 204 U.S. 585, 593 (1907). In the litigation over the constitutionality of the entry ban, the government asserted that the presumption, “which is magnified here by respect for the head of a coordinate Branch, counsels crediting the Order’s stated national-security purpose absent the clearest showing to the contrary.”5×5. Brief for the Petitioners at 78, Trump v. Int’l Refugee Assistance Project, Nos. 16-1436 and 16-1540 (U.S. Aug. 10, 2017). In the latest round of briefing, the government tweaked its presentation of the presumption, writing “the ‘presumption of regularity’ that attaches to all federal officials’ actions, as well as the respect due the head of a coordinate branch, should have foreclosed the Fourth Circuit from invalidating a presidential proclamation based on an uncharitable interpretation of an offhand, six-word comment.” Brief for the Petitioners at 68, Trump v. Hawaii, No. 17-965 (U.S. Feb. 21, 2018) (quoting Chem. Found., 272 U.S. at 14). The challengers argued that President Donald Trump’s statements calling for a Muslim ban should be considered as clear evidence of an unconstitutional motive to exclude Muslims.6×6. Brief of Respondents at 11–12, Int’l Refugee, No. 16-1436 (U.S. Sept. 11, 2017).

The litigants’ confident assertions and the fierce debate about the presumption that ensued in amicus briefs7×7. E.g., Brief of Amici Curiae Executive Branch Officials in Support of Respondents at 4, Int’l Refugee, Nos. 16-1436 and 16-1540 (U.S. Sept. 18, 2017) [hereinafter Amicus]. and in the media8×8. See, e.g., Josh Blackman, IRAP v. Trump: Applying the “Presumption of Regularity” in “Uncharted Territories,” Lawfare (May 9, 2017, 10:30 AM), https://www.lawfareblog.com/irap-v-trump-applying-presumption-regularity-uncharted-territories [https://perma.cc/SDH7-JG2A]; Linda Greenhouse, Lessons from a Travel Ban Clash That Wasn’t, N.Y. Times (Sept. 28, 2017), https://nyti.ms/2yuttYv [https://perma.cc/5LZY-YYB6]; Leah Litman, On Presumptions of Regularity, and Incidents of Irregularity, Take Care (May 11, 2017), https://takecareblog.com/blog/on-presumptions-of-regularity-and-incidents-of-irregularity [https://perma.cc/GWF8-RH7Z]. belied the reality that the doctrine’s operation and foundations are little understood. The Supreme Court has only uttered the phrase “presumption of regularity” in fifty-nine cases since 1900, less than half of which involve the executive branch.9×9. These figures were obtained by searching Westlaw for all Supreme Court cases mentioning the “presumption of regularity,” which returned fifty-nine results. The cases applying the presumption to executive branch actions were then identified. This Note examines the Supreme Court’s application of the presumption of regularity to the executive branch, but the presumption also applies to private actors and to judicial and legislative actions. See, e.g., Parke v. Raley, 506 U.S. 20, 29 (1992) (state court judgments); City of Richmond v. J.A. Croson Co., 488 U.S. 469, 500 (1989) (state legislature); Rogers v. Hill, 289 U.S. 582, 591 (1933) (corporate law). The Court often invokes the phrase without elaboration10×10. See, e.g., U.S. Postal Serv. v. Gregory, 534 U.S. 1, 10 (2001) (asserting without further elaboration “that a presumption of regularity attaches to the actions of Government agencies”). or develops the doctrine without invoking the phrase.11×11. All nineteen cases involving the executive branch — the President, prosecutors, and agencies — that use “presumption of regularity” are discussed. (Cases about grand juries are omitted.) The Note also discusses cases that do not use the phrase but that are recognized by scholars and courts as applications of the presumption. See infra note 50 and accompanying text. And the presumption has never been the subject of focused academic treatment.12×12. The presumption gets passing treatment in several literatures. Its evolution is discussed in work about arbitrary and capricious review. See, e.g., Richard W. Murphy, The Limits of Legislative Control over the “Hard-Look,” 56 Admin. L. Rev. 1125, 1132–33 (2004); Sidney A. Shapiro & Richard E. Levy, Heightened Scrutiny of the Fourth Branch: Separation of Powers and the Requirement of Adequate Reasons for Agency Decisions, 1987 Duke L.J. 387, 410, 426–27. Critiques of it pop up in work on prosecutorial discretion, see, e.g., Rachel E. Barkow, Separation of Powers and the Criminal Law, 58 Stan. L. Rev. 989, 1026 (2006), and judicial review of administrative action, see Douglas H. Ginsburg & Steven Menashi, Our Illiberal Administrative Law, 10 N.Y.U. J.L. & Liberty 475, 487–88 (2016). As a result, the applications, foundations, and implications of the presumption are in need of measured analysis.

The presumption of regularity is a deference doctrine: it credits to the executive branch certain facts about what happened and why and, in doing so, narrows judicial scrutiny and widens executive discretion over decisionmaking processes and outcomes.13×13. Cf. Adrian Vermeule, Law’s Abnegation 120–21 (2016) (invoking the presumption to argue for deference to agency procedural decisions). The Supreme Court’s cases applying the presumption to administrative agencies, prosecutors, and the President reveal that the presumption applies only to a subset of factual disputes about administrative motivations and internal processes. By analyzing the evolution of the presumption’s reach, we can reconstruct an implicit theory of administrative regularity. Whether and to what extent the Court is willing to presume procedural or motivational regularity in a given context depends on the Court’s assessment of the relevant decisionmaking scheme across several dimensions: procedural fairness,14×14. See United States v. Morgan (Morgan IV), 313 U.S. 409, 421–22 (1941). These assessments have not been formally enshrined in the doctrine. That said, Justice Harlan’s concurrence in Oestereich v. Selective Service System Local Board No. 11, 393 U.S. 233 (1968), suggested that procedural fairness was a formal predicate for presuming regularity of military draft orders. See id. at 242 (Harlan, J., concurring in the result). accountability, and degree of discretion,15×15. See, e.g., Citizens to Pres. Overton Park, Inc. v. Volpe, 401 U.S. 402 (1971); Pan. Ref. Co. v. Ryan, 293 U.S. 388 (1935). as well as complexity (of the decisionmaking scheme and of the substantive choices at stake).16×16. See, e.g., U.S. Dep’t of Labor v. Triplett, 494 U.S. 715 (1990) (complex benefits scheme). In making these assessments, the Court takes into account the capacities of courts, legislators, and administrators to evaluate and monitor decisionmaking structures, processes, and outcomes.17×17. For a crisp theoretical exposition of the “institutional design problems” of overseeing bureaucracies, see Jerry L. Mashaw, Bureaucracy, Democracy, and Judicial Review, in The Oxford Handbook of American Bureaucracy 570 (Robert F. Durant ed., 2010). These assessments aren’t static; the Court’s understanding of regularity is reshaped at times when settled assumptions about administrative process, action, and legitimacy are in turmoil.18×18. Comity and economy are often cited as the foundations of the presumption. See, e.g., Stephen G. Breyer et al., Administrative Law and Regulatory Policy 781 (7th ed. 2011) (“[This] allow[s] officials to perform their duties without fear of harassment . . . and unseemly probing of their mental operation, thus encouraging efficient administration . . . .”); Nathan Isaacs, Judicial Review of Administrative Findings, 30 Yale L.J. 781, 787–88 (1921) (similar). These concepts fail to capture the depth and complexity of the theory driving the presumption and therefore cannot explain the presumption’s current shape or its evolution. Cf. Lisa Heinzerling, The FDA’s Plan B Fiasco: Lessons for Administrative Law, 102 Geo. L.J. 927, 979–82 (2014) (concluding that comity and efficiency cannot justify a facet of the presumption).

This Note does three things: First, it improves our understanding of the presumption of regularity through a basic typology of the Supreme Court cases. Second, in Parts II and III, it reconstructs the theory of administrative regularity that appears to underlie the doctrine. Finally, also in Parts II and III, it sheds light on how the presumption has evolved alongside the Court’s attitude toward administration.

I. Restatement of the Presumption of Regularity

The presumption of regularity’s current domain covers two categories of disputes. In the first type, the challenger and the government disagree about what happened during the decisionmaking process.19×19. See, e.g., R. H. Stearns Co. v. United States, 291 U.S. 54 (1934). In the second type, the challenger and the government disagree about why a government action was taken.20×20. See, e.g., United States v. Chem. Found., Inc., 272 U.S. 1 (1926). Often, the record submitted for review will establish the facts beyond dispute.21×21. See Ronald M. Levin, Scope-of-Review Doctrine Restated: An Administrative Law Section Report, 38 Admin. L. Rev. 239, 284 (1986). When it does not, and the challenger alleges procedural or motivational misconduct, a court will presume that the agency acted regularly unless the challenger makes a strong contrary showing, at which point a court might look more closely, perhaps even conducting direct discovery into what happened at the agency.22×22. See Nat’l Archives & Records Admin. v. Favish, 541 U.S. 157, 174–75 (2004) (sketching how the presumption has operated). The presumption of regularity is relevant to factual determinations; it does not settle disagreements about how to interpret or apply the law.

In disputes about the decisionmaking process, courts sometimes assume that the government has observed procedural requirements or principles.23×23. See Levin, supra note 21, at 283–84; ABA Section of Admin. Law & Regulatory Practice, A Blackletter Statement of Federal Administrative Law, 54 Admin. L. Rev. 1, 44 (2002). More specifically, the Court has used this aspect of the presumption to settle three types of claims: (1) claims that the government skirted requirements in statutes or regulations; (2) claims of misconduct during the agency’s deliberations (more specifically, claims about the extent and manner of the consideration of evidence, or claims that an adjudicator is biased by the administrative structure); and (3) claims that an agency operates a procedural scheme (for example, a benefits claims process) in an unconstitutional manner. Note that there are certain common procedural claims to which the presumption has not been applied. For example, the burden is on the government to demonstrate that it followed notice-and-comment rulemaking requirements.24×24. See, e.g., Shell Oil Co. v. EPA, 950 F.2d 741, 758–62 (D.C. Cir. 1991).

In motivational disputes, the presumption of regularity helps courts identify or verify why the government acted. The presumption of motivational regularity’s application to the President is unsettled.25×25. Compare Chem. Found., 272 U.S. 1 (declining to look into the President’s reasons), and Am. Fed’n of Gov’t Emps. v. Reagan, 870 F.2d 723, 728 (D.C. Cir. 1989) (not requiring findings or reasons in an executive order), with Pan. Ref. Co. v. Ryan, 293 U.S. 388 (1935) (requiring reasons). This facet of the presumption has been used to bar discovery into prosecutorial and deportation decisions.26×26. See United States v. Armstrong, 517 U.S. 456 (1996); Reno v. Am.-Arab Anti-Discrimination Comm., 525 U.S. 471 (1999). In arbitrary and capricious review, the presumption is weaker, only shielding agencies from discovery about their motives when the agency has offered a contemporaneous explanation for the action. These contemporaneous explanations, typically in the form of an administrative record, are themselves presumed regular — that is, complete and authentic.27×27. Levin, supra note 21, at 265.

II. Procedural and Structural Disputes

As this Part will detail, the earliest iteration of the presumption of procedural regularity settled disputes about whether the government observed a discrete, technical requirement. As government action took on new forms in the early twentieth century, so did administrative discretion and process.28×28. Federal administration took on its modern, regulatory form, in which bureaucrats empowered with discretionary authority reshape economic and social relations, starting after the Civil War. Samuel DeCanio, Democracy and the Origins of the American Regulatory State 13–24 (2015); see also, e.g., William J. Novak, The Legal Origins of the Modern American State, in Looking Back at Law’s Century 249 (Austin Sarat et al. eds., 2002) (exploring the role of law in this transformation). The Court adapted the presumption of regularity: first to presume the regularity of internal agency deliberations and, from there, to address other types of disputes about administrative structure.

Several realizations about administration and judicial review enabled and motivated these changes. In the early New Deal era, the Court saw active judicial policing of internal agency processes as essential to protecting values like due process, accountability, and rule of law.29×29. See generally Daniel R. Ernst, Tocqueville’s Nightmare: The Administrative State Emerges in America, 1900–1940 (2014) (developing a legal process–centric account of 1930s administrative law). Applying the presumption to internal deliberations was an about-face, one that paralleled a deeper evolution in views on the regularity of administrative process.30×30. See id.; Reuel E. Schiller, The Era of Deference: Courts, Expertise, and the Emergence of New Deal Administrative Law, 106 Mich. L. Rev. 399 (2007); Mark Tushnet, Administrative Law in the 1930s: The Supreme Court’s Accommodation of Progressive Legal Theory, 60 Duke L.J. 1565 (2011). The Court came to appreciate that new forms of regulation demanded bureaucratic forms of decisionmaking and that legislators and administrators were at times better positioned than courts to design and supervise processes consistent with those values.

A. Discrete Procedural Requirements

Into the early twentieth century, direct judicial review of administrative action was available only piecemeal — through writs like mandamus, in common law tort actions against officials, and in suits between private parties.31×31. Frederic P. Lee, The Origins of Judicial Control of Federal Executive Action, 36 Geo. L.J. 287, 287, 295–96 (1948); Thomas W. Merrill, Article III, Agency Adjudication, and the Origins of the Appellate Review Model of Administrative Law, 111 Colum. L. Rev. 939, 946–49 (2011). The issues for review were usually narrow — jurisdiction, legal authority, and attention to procedural requirements.32×32. Jerry L. Mashaw, Federal Administration and Administrative Law in the Gilded Age, 119 Yale L.J. 1362, 1379 (2010) (noting that review was “extremely limited”). And in all of these areas, the Court, in Professor Jerry Mashaw’s words, “ostentatiously declined to meddle.”33×33. Id. at 1465; see also id. at 1464 (“[T]he basic rule . . . that emerges . . . is that administrative discretion will not be disturbed by judicial intervention.”); Ann Woolhandler, Judicial Deference to Administrative Action — A Revisionist History, 43 Admin. L. Rev. 197, 211 (1991) (“To the extent the courts examined agency procedures . . . they generally did so only at the wholesale level . . . .”). Under these conditions, the presumption that the government observed discrete procedural requirements could be consequential. For example, in R. H. Stearns Co. v. United States,34×34. 291 U.S. 54 (1934). a key issue was whether the Internal Revenue Service had confirmed several taxpayer settlements in writing, as required by statute, and the Court presumed based on circumstantial evidence that the settlements had been properly confirmed.35×35. Id. at 62–64.

Underlying this mode of review, and the powerful presumption that accompanied it, was the dominant conception of the separation of powers, which placed courts and administrators in separate spheres.36×36. See Bruce Wyman, The Principles of the Administrative Law Governing the Relations of Public Officers 84 (1903) (stating that “the life principle in the rule of the separation of powers” is judicial noninterference with “the internal operations of the administration”); Mashaw, supra note 32, at 1399–412 (analyzing cases). In Monongahela Bridge Co. v. United States,37×37. 216 U.S. 177 (1910). the Court explained that a determination delegated by Congress to the Secretary of War “cannot be revised . . . , except that . . . the courts can see to it that Executive officers conform their action to the mode prescribed by Congress.”38×38. Id. at 195. Regularity — conformity to the “mode prescribed” — was presumed: “It does not appear that the Secretary disregarded the facts, or that he acted in an arbitrary manner, or that he pursued any method not contemplated by Congress.”39×39. Id. The Court added, “[i]t is for Congress, under the Constitution, to regulate” in this area; Congress had delegated the determination to the Secretary of War, so judges needed to trust the Secretary.40×40. Id.

B. Internal Deliberations

Our current model of judicial review emerged in the early twentieth century.41×41. See Merrill, supra note 31, at 942. The struggle over the proper judicial role primarily played out in due process challenges.42×42. See Jerry L. Mashaw, Due Process in the Administrative State 77–78 (1985). For a time, the Court toyed with aggressive scrutiny,43×43. Id. at 51–73; see also Ann Woolhandler, Delegation and Due Process: The Historical Connection, 2008 Sup. Ct. Rev. 223, 236–47. invalidating state administrative schemes and subjecting decisionmaking processes to exacting review.44×44. See, e.g., Chi., Milwaukee & St. Paul Ry. v. Minnesota, 134 U.S. 418, 457–58 (1890); see also Paul R. Verkuil, The Emerging Concept of Administrative Procedure, 78 Colum. L. Rev. 258, 262 (1978). Around the turn of the century, however, the Court began presuming the regularity of state administrative boards and stopped allowing direct discovery into their deliberations. In Chicago, Burlington & Quincy Railway v. Babcock,45×45. 204 U.S. 585 (1907). the Court admonished lower courts for subjecting administrators and the state’s governor “to an elaborate cross-examination with regard to the operation of their minds in valuing and taxing the roads” and instructed that the “best evidence” of a board’s “decisions and acts” was its formal record.46×46. Id. at 593. This rule followed from the Court’s assessment of the administrative board. Rejecting a procedural due process claim, the Court held that tax assessment through administrative boards was not per se unconstitutional.47×47. Id. at 598; Verkuil, supra note 44, at 263 (“[Babcock] dissolved the per se notion of administrative inadequacy.”). It praised the particular board as expert, “honest,” and “sensible.”48×48. Babcock, 204 U.S. at 598.

Today, this “mental process rule is . . . ‘one facet of the general presumption of regularity’ which attaches to decisions of [federal] administrative bodies.”49×49. Singer Sewing Mach. Co. v. NLRB, 329 F.2d 200, 208 (4th Cir. 1964) (quoting 2 Kenneth Culp Davis, Administrative Law Treatise § 11.06 (1958)). This development came almost fifty years after Babcock, in the Morgan quartet.50×50. Morgan v. United States (Morgan I), 298 U.S. 468 (1936); Morgan v. United States (Morgan II), 304 U.S. 1 (1938); United States v. Morgan (Morgan III), 307 U.S. 183 (1939); United States v. Morgan (Morgan IV), 313 U.S. 409 (1941). The Morgan quartet does not use the term “presumption of regularity,” but courts and scholars have recognized its mental process rule as an aspect of the presumption. See Singer, 329 F.2d at 208; 2 Davis, supra note 49, § 11.05; Levin, supra note 21, at 283; Paul R. Verkuil, The Wait Is Over: Chevron as the Stealth Vermont Yankee II, 75 Geo. Wash. L. Rev. 921, 929 (2007). In the Morgan cases, marketing agents challenged rates set by the Department of Agriculture, arguing that the agency’s decisionmaking procedure violated both due process and the governing statute, which required a “full hearing.”51×51. Morgan I, 298 U.S. at 474–75 n.1. The rate-setting process had begun in the early New Deal era (before the appointment of the Secretary who issued the order); it involved two years of hearings (which produced no trial report) presided over by various agency officials.52×52. Morgan II, 304 U.S. at 15–21. Next, a committee of agency lawyers and economists, based on 11,000 pages of evidence, compiled a draft order and findings of fact,53×53. Ernst, supra note 29, at 72. which the Secretary, after meeting with those involved, issued.54×54. Morgan II, 304 U.S. at 16–18.

Troubled by the diffusion of decisionmaking across the agency, the Morgan I55×55. 298 U.S. 468. Court indicated that, generally, he “who decides must hear.”56×56. Id. at 481. It remanded the case for the government “to answer the[] allegations” that its process had deviated too far from this ideal.57×57. Id. at 482. Similarly, Morgan II58×58. 304 U.S. 1. emphasized that administrative process must conform to basic tenets of “fair play” and again invalidated the order, objecting to the mixing of prosecutorial and adjudicatory work during the rate-making process.59×59. Id. at 15.

The Court was wary of the growing discretion being vested in novel administrative structures during the New Deal.60×60. See Ernst, supra note 29, at 51–77. Underlying Morgan I’s “carte blanche”61×61. Breyer et al., supra note 18, at 781. to lower courts to police agency deliberations was a sense that imposing formal, quasi-judicial process requirements could check discretion being exercised through institutional decisionmaking.62×62. See Ernst, supra note 29, at 71; see also 2 Davis, supra note 49, § 11.01 (on institutional decisionmaking). But by Morgan II, the Court was reconsidering whether this was the optimal way to check agencies. The Court concluded, on the basis of the Secretary’s testimony on remand, that his consideration of the evidence had been sufficient,63×63. Morgan II, 304 U.S. at 18. adding, “it was not the function of the court to probe the mental processes of the Secretary in reaching his conclusions if he gave the hearing which the law required.”64×64. Id.

By Morgan IV,65×65. 313 U.S. 409 (1941). a fundamental transformation in the Court’s attitude toward administrative process had taken place.66×66. Professor Daniel Ernst’s recent history of the emergence of the administrative state makes much of Morgan II, arguing that agencies judicialized their procedures after the decision, see Ernst, supra note 29, at 51–77, and that the Court’s effort to shape administrative process was essential to legitimizing the nascent bureaucracy, see id. at 141–43. This reading ignores Morgan IV’s eventual application of the presumption of regularity and overstates the extent to which the New Deal Court embraced not only an active judicial role but also a single vision of administrative process. As Professor Kenneth Culp Davis noted in 1958, since Morgan IV, “courts have backed away from killing or substantially weakening the institutional decision” that characterizes administration. 2 Davis, supra note 49, § 11.01; see also Daniel J. Gifford, The Morgan Cases: A Retrospective View, 30 Admin. L. Rev. 237, 237 (1978) (analyzing Morgan as coming to terms with “complex decision-making by large, centralized agencies”). In an opinion by Justice Frankfurter, a New Dealer and advocate of active agencies and hands-off courts,67×67. See Ernst, supra note 29, at 9–27. the Court finally upheld the order and announced definitively that administrators were not to be examined directly, citing Babcock.68×68. Morgan IV, 313 U.S. at 422. The presumed regularity of administrative processes and of administrators’ motives supported this rule. “It will bear repeating,” said Morgan IV, “that although the administrative process has had a different development and pursues somewhat different ways from those of courts, they are to be deemed collaborative instrumentalities of justice and the appropriate independence of each should be respected by the other.”69×69. Id. Administrators “are assumed to be men of conscience and intellectual discipline, capable of judging a particular controversy fairly on the basis of its own circumstances.”70×70. Id. at 421.

Morgan IV’s acceptance of administrative process as fair and legitimate reflected a broader trend.71×71. See sources cited supra note 30; Verkuil, supra note 44, at 264–76. The same year, the Attorney General’s Report on Administrative Procedure concluded that “the problem of fairness in administration cannot be solved by judicial review alone”: the Court’s role was confined to enforcing “minimum standards.”72×72. Attorney Gen.’s Comm. on Admin. Procedure, Final Report 78 (1941). Limitations inherent in episodic review by generalist judges meant that the primary checks must lie elsewhere, including in “internal controls in the agency.”73×73. Id. at 76; see also Gillian E. Metzger & Kevin M. Stack, Internal Administrative Law, 115 Mich. L. Rev. 1239, 1273–74 (2017) (noting the report’s embrace of internal law). Morgan IV has had lasting structural consequences, enabling administrative decisionmaking structures that delegate and diffuse authority across agencies.74×74. Breyer et al., supra note 18, at 781–85; 2 Davis, supra note 49, §§ 11.01–.05 (characterizing Morgan as enabling the development of what Davis calls the “institutional decision”); see also Gifford, supra note 66, at 241–44 (describing how Morgan shaped the APA).

C. Administrative Bias

The Court has continued to grapple with what limits the Due Process Clause might place on agency structure. It has used the presumption to address the interaction between agency structure and the right to an unbiased adjudicator.75×75. See Gibson v. Berryhill, 411 U.S. 564, 578–79 (1973) (applying the fair tribunal rule to administrative adjudications). In Withrow v. Larkin,76×76. 421 U.S. 35 (1975). the plaintiff challenged a state licensing board’s “combination of investigative and adjudicative functions” on the theory that this feature “necessarily create[d] an unconstitutional risk of bias in administrative adjudication.”77×77. Id. at 47; see also Vermeule, supra note 13, at 63 (noting that “combination of functions” could risk “decisional distortions” like “self-serving bias” and avoidance of “reputational cost[s]”). Rejecting the idea that structure can systematically distort decisionmaking, the Court said: “The mere exposure to evidence presented in nonadversary investigative procedures is insufficient in itself to impugn the fairness of the Board members . . . .”78×78. Withrow, 421 U.S. at 55. It grounded this principle in Morgan IV’s presumption of regularity: “Without a showing to the contrary, state administrators ‘are assumed to be men of conscience and intellectual discipline, capable of judging a particular controversy fairly on the basis of its own circumstances.’”79×79. Id. at 55 (quoting United States v. Morgan (Morgan IV), 313 U.S. 409, 421 (1941)). Schweiker v. McClure80×80. 456 U.S. 188 (1982). applied this same principle to a benefits claims scheme in which administrative law judges could be appointed by private Medicare carriers, rather than by the government.81×81. Id. at 195. Even here, the Court refused to consider the claim that this structure created an unconstitutional risk of bias: “We must start,” the Court explained, again citing Morgan IV, “from the presumption that the hearing officers . . . are unbiased. This presumption can be rebutted by a showing of conflict of interest or some other specific reason for disqualification.”82×82. Id. (citing Morgan IV, 313 U.S. at 421).

Withrow and Schweiker add to the theory of administrative regularity. Morgan IV linked the presumption with the legitimacy and basic fairness of administrative processes. In these cases, the Court goes further, indicating that fair, effective, and accountable administration demand bureaucratic structures and that Congress and administrators are better equipped than courts to design and monitor those systems.83×83. Of course, the Court’s assessment might be wrong; a deferential approach could be inadequately protective of individual rights and claims to benefits, cf. Charles A. Reich, Individual Rights and Social Welfare: The Emerging Legal Issues, 74 Yale L.J. 1245, 1253 (1965) (urging full hearings in welfare determinations), or inadequate to ensure effective administration, especially of those programs that serve stigmatized populations, which “have historically been less generous and secure, [and] more vulnerable to maladministration,” Karen M. Tani, States of Dependency: Welfare, Rights, and American Governance, 1935–1972, at 9 (2016). Withrow explained why structures designed by Congress and agencies were more likely to be regular. Because of “the growth, variety, and complexity of the administrative processes,” systems must be tailored to context, and “legislators and others concerned with the operations of administrative agencies have given much attention to” various structures’ risks and benefits.84×84. Withrow, 421 U.S. at 51. In contrast, courts lack the practical experience and the capacity for systems-level thinking that institutional design and monitoring require.85×85. See Jerry L. Mashaw, Bureaucratic Justice 8 (1983) (“Why should . . . issues raised in episodic litigation . . . provide the judiciary with sufficient information for it to understand the administrative, political, social, economic, or scientific reality of a congressional-administrative program . . . [or] to take effective action to mold that reality in desirable forms?”); see also Vermeule, supra note 13, at 115. And the episodic judicial perspective could make courts error prone, unable to foresee how meddling would ripple to upset a carefully struck balance among competing values.86×86. See Mashaw, supra note 85, at 4 (noting that courts could impose “dysfunctional consequences for the supposed winners” or provide “irrelevan[t]” solutions). For example, the scheme in Schweiker could have been intentionally crafted to promote efficiency, carrier compliance, or some other legitimate administrative value at the expense of procedural fairness.

D. Procedural Schemes

The Court has also applied the presumption in Mathews v. Eldridge87×87. 424 U.S. 319 (1976). –based challenges to benefits claims processes. These cases sometimes call on the Court to figure out how a procedural scheme devised by Congress operates on the ground. In U.S. Department of Labor v. Triplett,88×88. 494 U.S. 715 (1990). the issue was whether statutory attorney’s fees limits in a benefits scheme for coal miners amounted to a violation of the claimants’ due process rights. The Court would reach the constitutional question — whether the claimants were due lawyers — only if the fee limits made lawyers essentially unavailable.89×89. Id. at 717–19, 722. The lower court, in concluding that lawyers were unavailable, had relied on assertions by attorneys that lawyers were less willing to take claims because of the caps.90×90. Id. at 723. But the Court disagreed: “[T]his sort of anecdotal evidence will not overcome the presumption of regularity and constitutionality to which a program established by Congress is entitled.”91×91. Id. It cited Walters v. National Ass’n of Radiation Survivors,92×92. 473 U.S. 305 (1985). which explained that, where the scheme set up by Congress is known or presumed to be constitutional, anecdotal evidence “is simply not the sort of evidence that will permit a conclusion that the entire system is operated contrary to its governing regulations.”93×93. Id. at 324 n.11; see also id. at 319–26 (upholding the constitutionality of the scheme at issue under a deferential standard). That is, Congress is presumed to design schemes that are constitutional and, in implementation and operation, agencies are presumed to act regularly — that is, to remain faithful to Congress’s design.

Like the structural bias cases, the benefits scheme cases pique the Court’s fear of bureaucratic systems design. Triplett and Walters also reflect a related worry that judicial meddling will affect not just processes but also outcomes.94×94. See INS v. Miranda, 459 U.S. 14, 17–18 (1982) (per curiam) (presuming the regularity of the immigration visa process based on perceptions about the substance of the agency’s work); see also Vermeule, supra note 13, at 113–23 (discussing links between substantive expertise and procedural design); cf. Vt. Yankee Nuclear Power Corp. v. Nat. Res. Def. Council, Inc., 435 U.S. 519, 525 (1978) (“[A]gencies . . . will be in a better position than federal courts or Congress itself to design procedural rules adapted to the peculiarities of the industry and the tasks of the agency involved.” (quoting FCC v. Schreiber, 381 U.S. 279, 290 (1965))). Procedurally, the worry was that promoting lawyers could compromise efficiency and consistency, even as it promoted fairness and individualization. Substantively, budgeting for lawyers’ fees and for government lawyers could shift resources away from some other priority that might benefit claimants more.95×95. See Vermeule, supra note 13, at 113 (“The decision over how much procedure to afford . . . is just a decision about the allocation of limited agency resources . . . .”). The Court is implicitly trusting that agencies will allocate their resources in ways that benefit claimants; this trust might be justified, given the missions of benefits agencies. See Mashaw, supra note 85, at 171 (“[T]he agency as adjudicator is not neutral with respect to who wins and who loses. Its mission is to pay the eligible and reject the claims of the ineligible.”). Similar intuitions about the connections between substance and procedure drove judicial review of Vietnam-era draft orders. Eager to leave the day-to-day operation of the draft to more expeditious and expert draft boards, courts presumed that boards had followed the intricate web of guidelines and procedures governing order of call determinations.96×96. See Comment, Due Process in Selective Service Appeals, 39 U. Chi. L. Rev. 331, 332 & n.6 (1972); see also id. at 337–38 (describing the government interests). This presumption was developed by the courts of appeals; the Supreme Court discussed but never applied it. See, e.g., Weintraub v. United States, 400 U.S. 1014, 1014 n.3 (1971) (Douglas, J., dissenting from denial of certiorari); United States v. Sisson, 399 U.S. 267, 323 (1970) (Burger, C.J., dissenting). After World War I, the Court afforded a presumption of regularity to the determinations of tribunals that reclassified servicemembers. See Rogers v. United States, 270 U.S. 154, 161 (1926). Because contrary evidence was so difficult to gather, this presumption that the order of call requirements had been followed verged on a presumption that the orders were valid.97×97. See Comment, supra note 96, at 332 & n.6 (describing and critiquing this presumption).

In theory, presuming procedural regularity — presuming that agencies follow concrete and abstract procedural principles — leaves agencies freer to develop their own internal systems.98×98. The real consequences are less clear. See Heinzerling, supra note 18, at 981–82 (arguing that Morgan exposes agencies to perversion by political forces); cf. Metzger & Stack, supra note 73, at 1286 (arguing that judicial enforceability of agency procedural regulations upsets agency control). Underlying this arrangement is an assessment of administrative process that is contingent and contextual, leaving the contours of the presumption subject to change with the Court’s views on administrative regularity.99×99. Cf. Jeremy K. Kessler, The Struggle for Administrative Legitimacy, 129 Harv. L. Rev. 718 (2016) (reviewing Ernst, supra note 29); Metzger & Stack, supra note 73, at 1304–06 (linking distrust of agency internal law to “distrust of administrative governance,” id. at 1304).

III. Motivational Disputes

When the validity of government action turns on rationale, courts have several options for how to approach judicial review. One option is to uphold actions when a valid rationale is conceivable, regardless of the reasons the government relied on; this leaves the primary controls on undesirable administrative action to politics. At the other extreme, courts might find their own facts and reasons and uphold only those actions supported by trial; this seeks to maximize legal accountability through judicial process. The presumption of regularity is one tool courts have used to calibrate review between these extremes. Here, the presumption helps determine the record on which an action is reviewed and the extent to which courts credit the government’s explanations.

As this Part will show, in the early 1900s, the Court took the first approach, and the presumption of regularity was hard to disentangle from a presumption of validity. Over time, the Court has diluted the presumption based on a reassessment of administrative regularity. Policymaking became more discretionary as delegations became more substantial and complex, especially during the New Deal and the 1960s and 1970s.100×100. Broad delegations have existed since the early days of the federal government, see Jerry L. Mashaw, Creating the Administrative Constitution 46–52 (2012), but by the 1970s, open-ended delegations were considered characteristic, see, e.g., Richard B. Stewart, The Reformation of American Administrative Law, 88 Harv. L. Rev. 1667, 1677–79 (1975). The accompanying worry is that discretion breeds “unaccountable and arbitrary exercises of administrative power.” Gillian E. Metzger, Ordinary Administrative Law as Constitutional Common Law, 110 Colum. L. Rev. 479, 491 (2010). Direct policing of delegations is generally seen as impossible and undesirable. See, e.g., Mistretta v. United States, 488 U.S. 361, 372 (1989) (“Congress simply cannot do its job absent an ability to delegate power under broad general directives.”). Instead, the Court has sought to channel discretion using substitute guardrails. See Stewart, supra, at 1688–711 (describing administrative models that might address discretion). Meanwhile, innovations in administrative structure and procedure meant that these sweeping policy choices might be forged in informal and diffuse processes.101×101. See generally Verkuil, supra note 44. Worries about procedural adequacy, accountability, and the efficacy of outcomes spurred a revolution in judicial review of informal agency action102×102. See Breyer et al., supra note 18, at 386–87, 551–52 (describing the changes). facilitated in part by tweaks in the presumption. But recent applications of the presumption suggest that a reconception may be underway, at least in certain areas.

A. Regularity as Validity

Under the deferential mode of judicial review that reigned at the turn of the century, courts didn’t look too closely at the government’s explanations.103×103. See supra section II.A, pp. 2435–36 (describing judicial review in this period). Take, for example, United States v. Chemical Foundation, Inc.,104×104. 272 U.S. 1 (1926). a 1926 case that grew out of executive branch actions authorizing the confiscation and sale of German chemical patents and other property during World War I. The government later brought charges against some buyers, alleging that they had conspired to buy up a monopoly by “deception” of the officials, including the President, who were in charge of the sales.105×105. Id. at 4. In the lower courts, “the United States failed to establish any conspiracy, fraud or deception alleged,”106×106. Id. at 14. and in the Supreme Court, the presumption of regularity applied: “Under that presumption, it will be taken that [the official who approved the sales] acted upon knowledge of the material facts. The validity of the reasons stated in the orders, or the facts on which they rest, will not be reviewed by the courts.”107×107. Id. at 15. The United States also alleged that the President had made the orders without “knowledge of the material facts,” id. at 14, but the Court explained that “[t]he President will be presumed to have known the material facts and to have acted in the light of them,” id. at 16; see also id. at 14–16. Absent proven bias, the Court suggested, courts are not to question the given facts or reasons. In other cases, the Court compared discretionary administrative actions to legislation and required no explanation, affording a “presumption of the existence of facts justifying [the] specific exercise.”108×108. Pac. States Box & Basket Co. v. White, 296 U.S. 176, 186 (1935) (state administrative action). In cases like this, a presumption that the government acted without illicit motive (the presumption of regularity) is so strong that it becomes a presumption that the government’s action is legally valid (a presumption of validity).109×109. Cf. Isaacs, supra note 18, at 788 (characterizing the presumption as affording “de facto . . . immunity from judicial review”).

B. Revising the Calculus

Early in the twentieth century, administrative action tended to occur after a formal adjudicatory process — a hearing on the record — and in service of a cabined mandate. The erosion of these traditional safeguards on administrative action during the New Deal made the Supreme Court nervous.110×110. See A.L.A. Schechter Poultry Corp. v. United States, 295 U.S. 495, 532–34 (1935) (contrasting a traditional scheme with a new and controversial scheme). It began to reassess administrative regularity.

Panama Refining Co. v. Ryan111×111. 293 U.S. 388 (1935). struck down part of the National Industrial Recovery Act112×112. Ch. 90, 48 Stat. 195 (1933) (terminated by Exec. Order No. 7252 (Dec. 21, 1935)). (NIRA) as an unconstitutional delegation of legislative power to the executive branch.113×113. Pan. Ref. Co., 293 U.S. at 420–30. It also invalidated a related executive order for failure to state facts and reasons demonstrating that it was an “appropriate exercise of the delegated authority.”114×114. Id. at 431. In dissent, Justice Cardozo urged that the presumption of regularity made the order valid.115×115. Id. at 446–48 (Cardozo, J., dissenting). A reason-giving requirement applied to adjudicative administrative actions, Justice Cardozo insisted, but not to presidential actions.116×116. Id. at 447. The order was presumed regular because it was an official act by the President — the execution of a delegation from Congress.117×117. Id. at 444 (“It is enough that the grant of power had been made and that pursuant to that grant [the President] had signified the will to act. The will to act being declared, the law presumes that the declaration was preceded by due inquiry and that it was rooted in sufficient grounds.”). This view recalls the Court’s position in Monongahela Bridge Co. v. United States, 216 U.S. 177, 195 (1910). In a system of separated powers, it wasn’t the Court’s job to question the motivations behind such official acts by the political branches.

But the majority rejected this conception of regularity and the underlying theory of the separation of powers. Digging into the decisionmaking scheme, the majority saw irregularity. Executive discretion had to be exercised under certain conditions to be fair and accountable, with legislation laying down procedures and “rules of decision” for the executive branch, which was to “pursue the procedure and rules enjoined and show a substantial compliance therewith.”118×118. Pan. Ref. Co., 293 U.S. at 432 (quoting Wichita R.R. & Light Co. v. Pub. Utils. Comm’n, 260 U.S. 48, 59 (1922)).

The Court also declined to apply the presumption a few years later, in United States ex rel. Johnson v. Shaughnessy,119×119. 336 U.S. 806 (1949). over cries in dissent that a deportation order lacking a full explanation could be upheld based on the presumption alone.120×120. Id. at 814–15; id. at 818 (Reed, J., dissenting). Then, in 1943, SEC v. Chenery Corp.121×121. 318 U.S. 80 (1943). held that an administrative action could only be upheld based on its real, contemporaneous rationale, indicating that reason-giving was a fundamental feature of a regular administrative scheme.122×122. Id. at 87. For an argument that Chenery and Panama Refining responded to the same concerns — arbitrariness and discretion — see Kevin M. Stack, The Constitutional Foundations of Chenery, 116 Yale L.J. 952, 986–98 (2007). Reason-giving and active judicial review are thought to check arbitrary exercises of discretion by promoting deliberative and reasoned decisionmaking inside the executive branch.123×123. See Frederick Schauer, Giving Reasons, 47 Stan. L. Rev. 633, 657 (1995) (“[W]hen institutional designers have grounds for believing that decisions will systematically be the product of bias, self-interest, insufficient reflection, or simply excess haste, requiring decisionmakers to give reasons may counteract some of these tendencies.”); Stack, supra note 122, at 993–1000 (on the connection between nonarbitrariness, rule of law, political accountability, and reason-giving).

C. Retrenchment

Panama Refining’s assessment of regularity became a permanent feature of judicial review of agency action in 1971, in Citizens to Preserve Overton Park, Inc. v. Volpe.124×124. 401 U.S. 402 (1971); see id. at 407–09, 414. Broad delegations had proliferated in social and environmental legislation of the 1960s and 1970s, just as judges and scholars were beginning to question the New Deal–era idea that bureaucratic expertise legitimized bureaucratic discretion.125×125. Stewart, supra note 100, at 1677–79. Meanwhile, “[a]dministrative law . . . ha[d] entered an age of rulemaking,” with agencies turning to informal processes, especially notice-and-comment rulemaking, to make sometimes-sweeping policies.126×126. William F. Pedersen, Jr., Formal Records and Informal Rulemaking, 85 Yale L.J. 38, 38 (1975). A general “sense of uneasiness” about administration pervaded, along with a conviction that agencies had been captured or were otherwise systematically biased toward powerful, organized interests.127×127. Stewart, supra note 100, at 1684; see id. at 1684–85; see also Breyer et al., supra note 18, at 386. In response, courts began to read “paper hearing” requirements into the Administrative Procedure Act’s128×128. Ch. 324, 60 Stat. 237 (1946) (codified as amended in scattered sections of 5 U.S.C.). (APA) bare rules for informal rulemaking129×129. See 5 U.S.C. § 553 (2012); Breyer et al., supra note 18, at 551–52. The rule that agencies must demonstrate that they have observed these requirements is inconsistent with the presumption of procedural regularity but consistent with the assessment of informal agency processes outlined in this section. and began to review administrative policymaking searchingly, to ensure that agencies had taken a “hard look” before acting.130×130. See, e.g., Catherine M. Sharkey, State Farm “with Teeth: Heightened Judicial Review in the Absence of Executive Oversight, 98 N.Y.U. L. Rev. 1589, 1609 (2014) (describing “hard look review”). But see Jacob Gersen & Adrian Vermeule, Thin Rationality Review, 114 Mich. L. Rev. 1355 (2016) (arguing that the Court has loosened review).

In Overton Park, the Court reviewed an informal adjudication by the Department of Transportation (DOT) authorizing federal funds for a controversial highway project that would cross a park in a white and relatively affluent neighborhood of Memphis.131×131. Citizens to Pres. Overton Park, Inc. v. Volpe, 401 U.S. 402, 406–08 (1971); Peter L. Strauss, Revisiting Overton Park: Political and Judicial Controls over Administrative Actions Affecting the Community, 39 UCLA L. Rev. 1251, 1251, 1261 (1992). The Court first rebuffed the government’s argument that discretionary administrative actions like this order were unreviewable by courts under the APA.132×132. Overton Park, 401 U.S. at 410. See also 5 U.S.C § 701(a)(2) (excepting from judicial review “agency action . . . committed to agency discretion by law”). Next, in clarifying the appropriate standard of review, the Court explicitly rejected a reading of the presumption of regularity as a presumption of substantive validity, explaining that the “presumption is not to shield [the Secretary’s] action from a thorough, probing, in-depth review.”133×133. Overton Park, 401 U.S. at 415 (citations omitted).

The Court set out to review the order under this standard. But because the APA doesn’t require a hearing on the record in informal actions,134×134. See 5 U.S.C. § 553 (rulemaking); id. §§ 555, 558 (adjudication). DOT hadn’t documented its findings and reasons.135×135. See Overton Park, 401 U.S. at 408, 417. Yet under the rule from Chenery, the order couldn’t be upheld on a rationale offered during litigation, and Morgan’s presumption shielded DOT’s internal deliberations from direct discovery.136×136. Id. at 419–20 (citing United States v. Morgan (Morgan IV), 313 U. S. 409, 422 (1941)). To ensure records for review of informal action, Overton Park reshaped the presumption, holding that it would shield agencies from direct discovery only where they have offered contemporaneous explanations for their actions.137×137. Id. at 420. This rule applies in litigation under the APA, but the connection between the presumption of regularity and reason-giving outside this context is unclear. Some would require it outside this context. See, e.g., Joseph v. United States, 405 U.S. 1006, 1006 (1972) (Douglas, J., dissenting) (arguing that Vietnam draft orders should state reasons to facilitate judicial review). When agencies provide explanations, usually in the form of an administrative record, courts presume that record’s regularity — that is, its completeness and accuracy.138×138. See Levin, supra note 21, at 263–67; James N. Saul, Overly Restrictive Administrative Records and the Frustration of Judicial Review, 38 Envtl. L. 1301, 1311–12, 1312 n.81 (2008). Delineating authenticity, accuracy, and validity is tricky. In Latif v. Obama, 666 F.3d 746 (D.C. Cir. 2011), the D.C. Circuit applied the presumption to intelligence reports submitted as evidence in a habeas proceeding, explaining that the presumption supported the reports’ “authenticity” as well as the idea that “the government official accurately identified the source and accurately summarized his statement.” Id. at 750; see also id. at 748–59. But the presumption neither extended to the “truth of the underlying . . . source’s statement” nor “compel[led] a determination that the record establishe[d] what it [wa]s offered to prove.” Id. Because the presumption of regularity can entail assuming accuracy, it overlaps with a presumption of correctness sometimes applied to technical calculations. See, e.g., United States v. Fior D’Italia, Inc., 536 U.S. 238, 242 (2002) (presuming the correctness of a tax assessment); United States v. Nix, 189 U.S. 199, 205 (1903) (presuming the correctness of mileage calculations in a reimbursement application).

This reading of the presumption not only forced record creation but also facilitated further developments in judicial review. Reviewing the recission of a notice-and-comment rule in Motor Vehicle Manufacturers Ass’n v. State Farm Mutual Automobile Insurance Co.,139×139. 463 U.S. 29 (1983). the Court again rejected a strong, substantive presumption before adopting a searching form of arbitrariness review designed to promote reasoned and deliberative policymaking.140×140. Id. at 43 & n.9; see also, e.g., Jody Freeman & Adrian Vermeule, Massachusetts v. EPA: From Politics to Expertise, 2007 Sup. Ct. Rev. 51, 52; Metzger, supra note 100, at 491–92; Mark Seidenfeld, A Civic Republican Justification for the Bureaucratic State, 105 Harv. L. Rev. 1511, 1561 (1992) (on deliberation forcing).

This reshaping of the presumption is consistent with the Court’s theory of administrative regularity. The complex bureaucratic systems presumed regular in cases like Schweiker and Triplett involved structured processes and intricate guidelines that might help to guarantee fairness for individual claimants, accountability to law and to political actors, and consistent, effective outcomes. Scaling back the presumption in Overton Park and State Farm was an effort to ensure this kind of regularity in policymaking.141×141. Cf. Metzger, supra note 100, at 491–93 (describing the constitutional concerns that motivated hard look review). The change, one contemporary commentator predicted, “will . . . open up the administrative decisionmaking process to more searching inquiry; [will] require the administrator to put more of his cards on the table; [will] require even informal administrative processes to . . . satisfy at least certain basic standards of fairness.”142×142. Nathaniel L. Nathanson, Probing the Mind of the Administrator: Hearing Variations and Standards of Judicial Review Under the Administrative Procedure Act and Other Federal Statutes, 75 Colum. L. Rev. 721, 768 (1975).

In narrowing the presumption and giving courts a more active role, Overton Park and State Farm reject the faith in the political branches implicit in the strong turn-of-the-century presumption.143×143. Cf. Nicholas Bagley, The Puzzling Presumption of Reviewability, 127 Harv. L. Rev. 1285, 1300–03 (2014) (arguing that modern arbitrariness review represents a departure from an earlier “separate-spheres conception of the separation of powers that called into question the very constitutionality of overseeing the executive’s discretionary judgments,” id. at 1302). As Professor Peter Strauss puts it, Overton Park’s “premise was that political controls could not be effective in relation to administrative action.”144×144. Strauss, supra note 131, at 1266. But “intensified judicial review” gave “the unrepresented . . . some protection against the tendency of the state to respond to concentrated, self-interested power.”145×145. Id. at 1317; see generally Stewart, supra note 100 (interest group theory). On this theory, a weaker presumption is needed to correct for political forms of representation that are prone to the irregularity of systematic bias. In State Farm, the worry was politically motivated deregulation. As now–Chief Judge Merrick Garland argued in 1985, hard look review emerged out of the intuition that reversal of an earlier rule “could signal that the agency was no longer being faithful to congressional policies or that it was otherwise acting on improper motives.”146×146. Merrick B. Garland, Deregulation and Judicial Review, 98 Harv. L. Rev. 505, 518 (1985). In rejecting the suggestion that the President’s deregulatory agenda could excuse the agency’s failure to engage in reasoned decisionmaking, State Farm reflected Overton Park’s skepticism of political checks.147×147. The dissent disagreed. See Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 59 (1983) (Rehnquist, J., concurring in part and dissenting in part) (“A change in administration brought about by the people casting their votes is a perfectly reasonable basis for an executive agency’s reappraisal of the costs and benefits of its programs and regulations.”). Both stand for the idea that, without a deliberative process and accountability via judicial review, administrative motives can’t be presumed regular.

D. Recent Applications

Recent applications of the presumption of regularity suggest a shift away from this theory of regularity and this view of the judicial role, at least in certain areas. First, the Court used the presumption to bar direct discovery of the motives behind criminal charging decisions. In United States v. Armstrong,148×148. 517 U.S. 456 (1996). defendants claimed that prosecutors’ decisions to charge them were motivated in part by racial bias.149×149. Id. at 459. But the Court, referencing Chemical Foundation, held that the presumption of regularity barred discovery of prosecutors’ records and notes, absent clear and direct evidence of bias.150×150. Id. at 464, 468–70; see also Hartman v. Moore, 547 U.S. 250 (2006) (applying this rule to non–equal protection selective prosecution claims). Reno v. American-Arab Anti-Discrimination Committee151×151. 525 U.S. 471 (1999). indicated that the same rule applied in selective deportation claims against immigration agencies.152×152. Id. at 489–90.

Faith in the regularity of the criminal process could in theory underlie Armstrong.153×153. Cf. Parke v. Raley, 506 U.S. 20, 30 (1992) (presuming the regularity of a prior conviction by plea agreement on the grounds that the plea colloquy process proceeds in accordance with constitutional requirements). Yet as Professor Rachel Barkow has noted, the guardrails on prosecutorial decisionmaking — in an environment where nintey-seven percent of charges are settled through plea agreements154×154. See Dylan Walsh, Why U.S. Criminal Courts Are So Dependent on Plea Bargaining, The Atlantic (May 2, 2017), https://www.theatlantic.com/politics/archive/2017/05/plea-bargaining-courts-prosecutors/524112/ [https://perma.cc/GE6X-5EC5]. — are virtually nonexistent, far looser than those on administrative decisionmaking.155×155. Barkow, supra note 12, at 1024–31. Armstrong does rest on “the relative competence of prosecutors and courts,” a functional reason for presuming regularity that is in line with the focus on institutional capacity in Withrow and Triplett.156×156. United States v. Armstrong, 517 U.S. 456, 465 (1996). But Armstrong’s presumption is also founded on a conception of unreviewable executive discretion inconsistent with the separation of powers theory underlying Overton Park and State Farm. Armstrong rooted its presumption in “a concern not to unnecessarily impair the performance of a core executive constitutional function.”157×157. Id. And Reno invoked the same idea: selective enforcement claims, Justice Scalia wrote, “invade a special province of the Executive — its prosecutorial discretion — [so] we have emphasized that the standard for proving them is particularly demanding, requiring . . . ‘clear evidence’ displacing the presumption that a prosecutor has acted lawfully.”158×158. Reno v. Am.-Arab Anti-Discrimination Comm., 525 U.S. 471, 489 (1999) (quoting Armstrong, 517 U.S. at 465); see also Hartman v. Moore, 547 U.S. 250, 263 (2006) (“[The] presumption that a prosecutor has legitimate grounds for the action he takes is one we do not lightly discard, given our position that judicial intrusion into executive discretion of such high order should be minimal.”).

A “tradition of deference” to prosecutors might distinguish Armstrong from Overton Park and State Farm.159×159. Armstrong, 517 U.S. at 465. But Reno, about agency, not criminal action, could signal a shift in the Court’s theory of regularity. Compare Reno to Heckler v. Chaney,160×160. 470 U.S. 821 (1985). which held that an agency’s decision not to bring an enforcement action was an unreviewable exercise of discretion; the Heckler Court grounded this interpretation of the APA not only in the difficulty of reviewing amorphous prioritization decisions but also in the idea that Article II’s Take Care Clause161×161. U.S. Const. art. II, § 3. commits enforcement to the Executive.162×162. Heckler, 470 U.S. at 831–32. Heckler has been criticized for importing principles applied to prosecutors into administration. See id. at 846–50 (Marshall, J., concurring in judgment); Cass R. Sunstein, Reviewing Agency Inaction After Heckler v. Chaney, 52 U. Chi. L. Rev. 653, 665–75 (1985). The affinities between Heckler and Reno bring two potential changes into focus. First, Armstrong and Reno could signal the return of a stronger presumption of regularity, one verging on a presumption of unreviewability, or of substantive validity. Second, the ground underneath the presumption of regularity could be shifting from a functional assessment of the relevant decisionmaking scheme to a categorical conception of executive power.163×163. These competing visions of the presumption of regularity clashed in Latif, the case applying the presumption to intelligence reports. 666 F.3d 746 (D.C. Cir. 2011). The majority reasoned that the “separation of powers justifies a presumption in favor of official Executive branch records,” id. at 751, and that “[t]he presumption of regularity is founded on inter-branch . . . comity, not our own judicial expertise with the relevant government conduct,” id. at 752. But the dissent characterized the presumption as applying to “actions taken or documents produced within a process that is generally reliable because it is, for example, transparent, accessible, and often familiar.” Id. at 771 (Tatel, J., dissenting). This divide could be a species of the ongoing cycling between formalist and functionalist approaches in separation of powers doctrine. See John F. Manning, Separation of Powers as Ordinary Interpretation, 124 Harv. L. Rev. 1939, 1950–71 (2011) (describing both approaches).

The second recent application relates to President Trump’s rescission of Deferred Action for Childhood Arrivals (DACA), an Obama-era immigration policy. The district court ordered the Department of Homeland Security to supplement the administrative record submitted in APA litigation over the recission.164×164. See In re United States, 138 S. Ct. 443, 444–45 (2017) (per curiam). And the Ninth Circuit agreed that “the presumption of regularity that attaches to the government’s proffered record [was] rebutted.”165×165. In re United States, 875 F.3d 1200, 1207 (9th Cir. 2017). “Put bluntly,” the Ninth Circuit said, “the notion that the head of a United States agency would decide to terminate a program giving legal protections to roughly 800,000 people based solely on 256 pages of publicly available documents is not credible.”166×166. Id. at 1206.

The Supreme Court disagreed. It embraced a conception of regularity and deference as grounded in separate judicial and executive spheres, departing from the theories of regularity and judicial review in Overton Park and State Farm. To start, it suggested that DACA might be, like the action in Heckler, categorically unreviewable.167×167. In re United States, 138 S. Ct. at 445 (urging the lower court to consider whether the action is “unreviewable because it is ‘committed to agency discretion’” (quoting 5 U.S.C. § 701(a)(2) (2012))). Without commenting on the regularity of the government’s record, the Court worried that the lower court’s order was “overly broad” and noted the government’s argument that the order compelled privileged material.168×168. Id. One of the privileges asserted by the government, the deliberative process privilege,169×169. In re United States, 875 F.3d at 1210. is closely related to Morgan’s mental process rule.170×170. See San Luis Obispo Mothers for Peace v. U.S. Nuclear Regulatory Comm’n, 789 F.2d 26, 44–45 (D.C. Cir. 1986) (barring discovery of transcripts of a multimember agency and discussing Morgan IV and the privilege). The government had also argued that a species of executive privilege, rooted in “the Executive Branch’s interests in maintaining the autonomy of its office and safeguarding the confidentiality of its communications,” should protect any communications involving White House staff.171×171. Petition for a Writ of Mandamus at 26, In re United States, 138 S. Ct. 443 (No. 17-801) (quoting Cheney v. U.S. Dist. Court, 542 U.S. 367, 385 (2004)). In the travel ban litigation, the government makes forceful assertions about executive privilege in language redolent of recent presumption of regularity cases like Armstrong and Reno: “[P]robing the President’s supposed true reasons . . . invites impermissible intrusion on privileged internal Executive Branch deliberations and potential litigant-driven discovery that would disrupt the President’s execution of the laws.” Brief for the Petitioners at 67, Trump v. Hawaii, No. 17-965 (U.S. Feb. 21, 2018) (citations omitted). The relationship between the presumption and various forms of privilege remains underexplored. Absent from the Court’s order was an assessment of the fairness, reliability, or accountability of the decisionmaking context.

Several Justices would have allowed the documents to be compelled.172×172. In re United States, 138 S. Ct. 371, 371 (2017) (Breyer, J., dissenting from grant of stay). Justice Breyer, writing for the group, read the presumption narrowly, as applicable only in instances where plaintiffs sought discovery of “subjective mental reasoning,” not of material like agency documents.173×173. Id. at 373. Justice Breyer characterized the lower courts as “unanimously reject[ing] the Government’s position that the agency may unilaterally determine the contents of the administrative record that a court may review.” Id. at 372. That’s true; but courts have also required some showing from the plaintiff to get documents added to the administrative record. See Levin, supra note 21, at 263–68. Leaning on Overton Park and State Farm, Justice Breyer embraced review on an extensive record as necessary protecion against arbitrary or unlawful outcomes.174×174. In re United States, 138 S. Ct. at 372 (Breyer, J., dissenting from grant of stay); see also Heinzerling, supra note 18, at 981–83 (arguing that the mental process rule “should be softened,” id. at 981, to protect agencies from being perverted by “political intervention[s],” id. at 982). This inclination is consistent with our account of the Court’s theory of administrative regularity. DACA was an unusual initiative, both substantively — in that it formalized and institutionalized a policy of nonenforcement (such policies are traditionally implemented on an “ad hoc basis, guided by loose priorities laid out in . . . agency memos”175×175. Adam B. Cox & Cristina M. Rodríguez, The President and Immigration Law Redux, 125 Yale L.J. 104, 177 (2015); see also id. at 174–76 (summarizing relevant features of the initiative). ) — and procedurally — in that it made significant policy through an informal, although deliberated, interagency process directed by the Obama White House.176×176. Id. at 215–19. As the Ninth Circuit suggested, the reversal process directed by the Trump White House came quickly and seemed reliant on little evidence. These features might raise accountability, procedural, or rule-of-law flags that merit tweaks to the presumption.177×177. Cf. W. Neil Eggleston & Amanda Elbogen, The Trump Administration and the Breakdown of Intra-Executive Legal Process, 127 Yale L.J.F. 825 (2018) (arguing that courts are withholding deference from the Trump Administration because of deviations from regular process).

Justice Breyer’s analysis may also be consistent with the implicit conception of regularity applied by the lower courts in the travel ban litigation: both circuits have considered outside-the-order evidence of motive while emphasizing the irregularity of the Administration’s decisionmaking processes.178×178. Int’l Refugee Assistance Project v. Trump, 857 F.3d 554, 591–92 (4th Cir. 2017), cert. granted, 137 S. Ct. 2080 (2017) (per curiam); Hawaii v. Trump, 859 F.3d 741, 755–56, 770–82 (9th Cir. 2017) (per curiam), cert. granted sub nom. Trump v. Int’l Refugee Assistance Project, 137 S. Ct. 2080.

Conclusion

The Supreme Court’s assessments of regularity are contextual and contingent, and the contours of the presumption are subject to change as attitudes evolve. The Court systematically reassesses regularity, and with it the presumption, at times when settled models of administration are being upended, administrative discretion is expanding, administrative process is in flux, and old legitimating ideals are ill fitting. Underlying each new assessment is an implicit yet consistent theory of administrative regularity. This concept of regularity emphasizes due process, accountability, and rule-of-law values, and it looks at the extent to which a scheme ensures basic procedural fairness and protects against arbitrary, unlawful, or ineffective outcomes.

The presumption of regularity is a type of deference doctrine, but the presumption of regularity can also teach us something about judicial deference more generally. Procedural and motivational regularity are also predicates for other forms of deference. That is, the assumption that officials and entities follow procedures, act rationally, reasonably, and in good faith enables courts to credit the executive branch’s legal interpretations and policy judgments.179×179. See, e.g., Daphna Renan, Presidential Norms and Article II, 131 Harv. L. Rev. 2187, 2255–57 (2018); Dawn Johnsen, Judicial Deference to President Trump, Take Care (May 8, 2017), https://takecareblog.com/blog/judicial-deference-to-president-trump [https://perma.cc/2L2V-77YM]. Understanding the presumption of regularity can help us get a better grasp on the relationship between courts and administration.