Administrative Law Outline Vermeule, Autumn 2002 I. Introduction a. Problems thought to call for administrative regulation i. Market failures, economically defined 1. The need to control monopoly power a. Regulations aim at ―allocative efficiency‖ – set prices at what they would be if competitive, so market doesn‘t artificially have to produce more of B than A, where B is more expensive than A but for A‘s monopoly price. b. Difficulty in defining the market—is cable a natural monopoly in view of existence of broadcast tv, VCRs, Hollywood? c. Rationale includes noneconomic ends: fairer income distribution, avoiding price/service discrimination among customers, distrust of power of unregulated monopolist 2. The need to compensate for inadequate information a. Correct systematic information failure, but keeping in mind the fact that information is costly and that while consumers as a class have an interest in obtaining info they can‘t share the costs of obtaining it fairly (free rider problem). b. Heuristic errors people make in processing info (availability heuristic) c. Motivational errors people make in processing info (reducing cognitive dissonance through irrational optimism) d. Gov‘t action may be justified when: i. Suppliers mislead consumers whose available legal remedies (lawsuits) are expensive or impractical ii. Consumers cannot readily evaluate the info available, such as potential effectiveness of a drug iii. The market on the supply side fails to furnish the info needed or demanded. e. If info is inadequate, and it is expensive to provide info, maybe just ban the product—the argument being that products should be banned if, were info provided, no or very few consumers would want it 3. Collective action problems a. Public goods: nonrivalrous consumption plus nonexcludability (e.g., national defense) b. Prisoner‘s dilemma problems 4. The need to correct for externalities or for the existence of transaction costs that make bargaining difficult ii. Less secure economic grounds 1. The need to control ―windfall‖ profits a. Rents might skyrocket (windfall profits) if commodity prices suddenly increase (e.g., OPEC embargo would result in windfall for domestic oil reserve holders). b. When profits are high and don‘t reflect any particular talent on part of producers, maybe regulate to transfer ―undeserved‖ profits from producers (or owners) of the scarce resource to consumers (or taxpayers) 2. The need to eliminate ―excessive‖ competition a. If prices are too low, all but a few companies will go out of business, and then they‘ll be an oligopoly b. Also arises in industries w/ large fixed costs and cyclical demand, such that the firm may discover during economic downswing that they have insufficient revenue to continue production (but closing plant would be inefficient). This argument used in agriculture. c. Predatory pricing—dominant firm sets prices low to drive rivals out of business, so it can then charge monopoly prices. This argument used in antitrust. 3. The need to alleviate scarcity 4. Agency problems a. Market forces may be distorted when someone other than buyer makes purchasing decisions (or helps pay for buyer‘s purchase). Consumption may be greater than it would be if buyer had to make purchase and pay for it alone (e.g., medical care). iii. Redistribution 1. Transfer payments can be indirect and regulatory – e.g., minimum wage laws. Of course, the distributional consequences from some of these policies might be unfortunate or perverse. 2. On behalf of these policies, unequal bargaining power, based on unequal starting points, is rationale. Maybe overlaps w/ information failure. 3. But it may be that the effort to redistribute through regulation doesn‘t work well or at all. The public interest argument might be a disguise for private interest of well organized factions. iv. Nonmarket or collective values 1. Social/cultural norms may encourage political participants to seek regulation in absence of a market failure (e.g., wanting educational tv programming). v. Disadvantage and caste 1. Gov‘t attempts to shape preferences to overcome systematic forms of social disadvantage (e.g., civil rights law) vi. Planning 1. W/out gov‘t regulation, firms might not produce products efficiently. Another argument is that economic markets are insufficiently democratic, and that elected administrators might be able to take account of considerations the market doesn‘t vii. Paternalism 1. Gov‘t has obligation to protect individuals from their own confusion and irresponsibility. 2. Can be supported by research finding that at time of decision, people systematically misunderstand what they‘re experiencing. Or maybe people have preferences about their preferences, such that they seek external forces to help protect them from own misjudgment (e.g., Ulysses tying himself to the mast when faced w/ sirens). b. Classic regulatory tools i. Types of regulatory programs: 1. cost of service ratemaking 2. allocation in accordance with a public interest standard 3. standard-setting 4. historically based price-setting or allocation 5. screening or licensing 6. fees or taxes 7. provision of information 8. subsidies 9. noncoercive efforts to produce cooperation through moral suasion or political incentives ii. Regulator and regulated industries likely to have adversarial relationship (less true w/ provision of info, subsidies, and noncoercive efforst..) iii. Regulator is institutional bureaucracy likely to prefer easy to administer rules iv. New regulatory programs often copy old ones v. Regulatory decisions are subject to requirements of administrative law (e.g., APA). c. The historical development of administrative government and administrative law i. English antecedents 1. Grew out of English common law, in response to inability of c/l courts to properly remedy actions of gov‘t officials (b/c couldn‘t be pigeonholed into writs and b/c officials couldn‘t satisfy judgment out of pocket). Writ of mandamus originated to require officials to grant/restore entitlements to citizens. Writ of prohibition originated to preclude exercise of power not w/in official‘s jurisdiction. Writ of certiorari refashioned to apply to decisions of administrative bodies. Chancery courts‘ injunctive remedies applied to officials. ii. American experience to 1875 1. Somewhat laissez-faire, but gov‘t imposed taxes, distributed public lands, gave patents, regulated commerce, etc. administrative law largely about the writ system, w/ courts relying on jurisdiction concept to establish sharp distinction between actions w/in and w/out an administrator‘s authority. Court‘s ignored the possibility of a third category—an area of limited discretion subject to judicial review. iii. 1875 to 1930: the rise of administrative regulation and the traditional model of administrative law 1. State regulation of interstate railroad monopolies led to creation of Interstate Commerce Commission (ICC) in 1887 (abolished 1995). ICC founded to protect small shippers from exercise of monopoly power and rate discrimination by larger carriers. It shifted income from RRs to shippers, and vindicated more general public interest in efficient and centrally planned RR system. Alternatively, ICC was there to protect RRs from competition, and that inept regulation fatally weakened RRs by destroying innovation. 2. Courts were repeatedly invoked to bock regulatory decisions. C/l writ system supplanted by statutory provisions. Traditional model of administrative law evolved, w/ 4 elements: a. Legislature must authorize administrative sanctions on private persons through rules/standards that limit agency discretion b. Procedures used by agency must tend to ensure agency compliance w/ legislative directives c. Judicial review must be available to ensure that agencies use accurate and impartial decisionmaking procedures and comply w/ legislative directives d. Agency processes must facilitate the exercise of judicial review 3. Initially, courts treated administrative agencies w/ distrust iv. The New Deal and beyond: 1932-1945 1. Agencies beefed up and multiplied during New Deal. Critics attacked the agencies for combinind prosecutorial and adjudicatory functions and wanted procedural checks and judicial review. Supporters argued that effective performance required mixed powers and discretion (like that given to business leaders), and relied upon expert professionalism of administrators and political control by president to prevent abuses. v. 1945 to 1962: the Administrative Procedure Act (APA) and the maturation of the traditional model of administrative law 1. Compromise w/ critics resulted in Federal Administrative Procedures Act (1946) (codifed in 5 USC § 551 et seq). a. § 2 (5 USC § 551) contains definitions b. § 3 (5 USC § 552) establishes publication and public notice requirements for agency decisions. After amendment, all documents in agency‘s possession (w/ certain enumerated exceptions) must be made available to public and all meetings (w/ certain enumerated exceptions) must be open to public c. §§ 4-8 (5 USC §§ 553-58) deal w/ procedural formalities agencies must observe in decisionmaking. Distinguishes between promulgation of general regulations through rulemaking and case-by-case decisions through adjudication. Rulemaking consists of publication of proposed rules followed by public comment. For decisions ―on the record after opportunity for agency hearing,‖ APA provides for modified trial- type hearing followed by appeal to head of agency, w/ partial separation of functions between prosecuting staff and agency decisionmaking d. § 10 (5 USC §§ 701-706) deals w/ availability, timing, form, and scope of judicial review. Abuse of discretion review; cts must determine whether agency‘s factfinding supported by substantial evidence in the record as a whole 2. Administrative law develops sense of stability vi. 1962 to 1980: the rights revolution, critique of administrative process and administrative substance, and public interest administrative law 1. Regulation shifted from stabilizing the economy to protect public health/safety. Congress abandoned faith in administrative autonomy and prescribed clear rules/deadlines/etc for agencies 2. Critics: a. argued that agencies had been ―captured‖ by the firms they were supposed to regulate, and called for greater public openness in agency decisionmaking, statutory deadlines, increased participation by public interest advocates, more formal decisionmaking procedures, stricter judicial scrutiny, closer congressional scrutiny, elimination of conflicts of interests by regulators, and elimination of programs that primarily benefit the regulated firms. b. assailed administrative apparatus of welfare and public service state. Beneficiaries sought procedural protections – welfare as ―the new property‖ c. called for deregulation and consolidation, arguing that administrative law was largely inefficient, and noting that correcting market failures w/ clumsy regulatory tools created new kinds of market failures. 3. Courts extended right to participate in agency decisionmaking, extended coverage/content of procedural formalities, expanded availability/scope of judicial review. 4. Congress deregulated some industries and experimented w/ limited forms of economic incentives, such as marketable air pollution rights vii. 1980 to ?: presidential administration and the cost-benefit state 1. Reagan administration called for more deregulation. a. Executive Order 12,291 called for OMB control of regulations w/ careful attention to costs and benefits of proposed initiatives, a process ratified by Clinton in Executive Order 12,866. 2. Gov‘t has shown increasing interest in engaging in quantitative analysis, shown a greater interest in assessing tradeoffs, and an interest in smarter tools, especially by replacing command-and-control regulation w/ info disclosure and economic incentives 3. Courts have cut back on judicial review, limited extent to which courts can impose new procedural requirements on agencies, and required courts to pay attention to agency interpretations of statutes II. The Uneasy Constitutional Position of the Administrative Agency a. Agency‘s power to legislate i. U.S. Const. art I, § 1 provides that ―all legislative powers … shall be vested in a Congress of the United States….‖ Early cases prevented delegation of legislative power, but in 1897, Court said that although the power to prescribe rules ―in the future … is a legislative act, … Congress … might commit to some subordinate tribunal this duty.‖ ICC v. Cincinnati, New Orleans & Texas Pac. Ry. (upholding ICC Act) 1. Hence, the central issue in these kinds of cases is whether Congress has given an agency so much rulemaking discretion that Congress has abdicated its responsibility to exercise ―[a]ll legislative Powers‖ granted in Art. I. ii. Alternatively, the nondelegation doctrine can be thought to flow not from Art I, § 1, but also from a structural argument— separation of powers: it violates separation of powers for Congress to abdicate its legislative powers to the Executive (the inverse of Congress horning in on the role of the Executive) 1. Separation of Powers: the nondelegation principle is meant to preserve sep of powers, which promotes uniformity and impartiality in the application of sanctions. B/c of the risks posed by delegation, outside of certain narrowly defined foreign affairs and military situations, Congress must authorize any intrusions by executive officials on private liberty/prop. And delegations of legislative power to executive must be in such a form that it has rules that control administrative decisions. iii. Congress delegates in several common ways: 1. contingent legislation—statute becomes effective when administrative agency finds some fact or makes some determination 2. interstitial rulemaking—statute directs agency to use its expertise to fill in the gaps in a scheme of regulation, asking it to promulgate rules/standards iv. Why delegate? 1. Efficiency a. Take advantage of agency‘s comparative technical expertise b. Congress is clumsy rulemaker; agency has institutional advantage as a rulemaker (b/c it can concentrate on this measure rather than being one of dozens of pressing matters) 2. Bad reasons a. Reduce political accountability: Avoid tough political stances—instead of having to make hard choices, pass the buck to blue ribbon commission and tell that agency to make the decisions so you can talk the talk and blame the agency for any problems b. Advance special interest groups interests: Rather than having to vote in favor of air pollution, pass the buck to an agency that is susceptible to lobbying by your favorite interest groups and that are less likely to be subject to public scrutiny b/c of increasing marginal information costs for media in investigating congress, then all of agencies, and etc. b. The Nondelegation Doctrine i. Goals advanced by nondelegation doctrine: 1. Political accountability/deliberativeness – more particularly, the kind of accountability coming from requiring specific decisions from a transparent and deliberative body like Congress (which has an outlet for the voice of vocal minorities, and thus tends to craft compromises). Hence, the goal isn‘t accountability so much as promoting deliberative democracy. 2. Rule of law: This deliberative democracy is a safeguard of individual liberties, and promotes rule of law values (since it cabins discretion and provides for judicial review). ii. Evaluating these arguments 1. This is an agency problem (to reduce agency costs—e.g., shirking, self-dealing—the principal can adopt a number of strategies: (1) contract design (e.g., different compensation scale to create different incentives); (2) monitor agent‘s activities; (3) screen agents; (4) create institutional competition that will check shirking/diversion by agents (e.g., organize workers into different teams, so that each will rat out the other if they‘re not working hard enough). 2. Delegation problems create nested agency relationship. Congressional representatives are agents of the voter. Agencies are agents of Congress. In this framework, we can see the reasons Congress wants to delegate to agencies for good or ill. E.g., Congress might use delegation to shirk (i.e., avoid tough policy choices) or to divert resources (i.e., create agencies susceptible to lobbying by favored special interests). Thus, in this context, the nondelegation doctrine can be seen as a strategy of voters to monitor their agents through institutional competition— Congress. Preventing delegation makes monitoring of agents (Congress) easier b/c it is more public than the work of an alphabet soup of administrative agencies. iii. Nondelegation caselaw 1. Pre-1935 Supreme Court decisions prefigured doctrine but were deferential to Congress a. Contingency theory (Field v. Clark, 1892: prez given authority to impose retaliatory tariffs whenever other nation imposed tariffs the prez deemed unequal and unreasonable; held to be OK b/c prez just ascertaining matter of ―fact‖ (whether other nation‘s tariff is unequal and unreasonable) that was the contingency for triggering legislative act done by Congress b. Filling in the details principle (U.S. v. Grimaud, 1911: sec‘y of agriculture given rulemaking authority to protect public forests; held ok since statute didn‘t delegate legislative authority but merely permitting executive to fill in the details in this constrained little field—public forest protection law) c. J.W. Hampton, Jr. & Co. v. US (1928): statute that gives prez power to revise tariffs whenever he felt revision necessary to equalize the costs of production in the US and the competing country; held ok b/c the notion of adjusting tariffs to ―equalize costs‖ constituted an ―intelligible principle‖ (relying on a combo of the contingency theory and filling in the details principle) 2. 1935: Panama Refining and Schechter a. Panama Refining Co v. Ryan, 1935, was a challenge to NIRA‘s petroleum code, which authorized prez to ban interstate shipment of ―hot oil‖ but didn‘t give any guidance as to when prez should do so (b/c sentence making a violation of oil production quotas unlawful was missing). Court struck down that section of NIRA b/c it didn‘t provide standard (―an intelligible principle‖) governing when the prez could exercise the power (though Cardozo said that purpose section of NIRA and its context/background could provide the standard, Hughes replied by saying that the purpose section provided too many standards w/out indication of how the prez was to choose among them). b. A.L.A. Schechter Poultry Corp. v. U.S., 1935. Court struck down provision of NIRA that authorized prez to approve industry-created ―codes of fair competition‖ for the poultry industry and other industries. Can‘t delegate to private parties. Sheer breadth of delegation is unconstitutional (Cardozo likens it to roving commission to ―do good‖); even though prez could veto industry-created code, NIRA didn‘t give guidance to when prez should approve/veto. 3. Post-New Deal nondelegation caselaw a. Amalgamated Meat Cutters v. Connally, DC Cir, 1971. Court upheld Economic Stabilization Act which permitted prez to freeze wages/prices to combat cost-push inflation; statute calls for prez to develop standards to determine when/how to freeze. Court finds implicit limitations on prez‘s authority based in history of cost-push inflation/prior stabilization programs: freezes were to be fair/equitable and time-limited. Court suggests that subsidiary rulemaking and subsequent judicial review would sufficiently cabin agency‘s discretion (a position the USSC rejects in Whitman). b. Whitman v. American Trucking Associations, 2001. Court upholds Clean Air Act, which calls for agency to set primary/secondary air quality standards ―requisite‖ to protect public health. Court rejects notion that agency could adopt standards limiting its discretion that would ―save‖ the delegation; however, ―requisite‖ is intelligible enough—just b/c EPA has to make judgments of degree doesn‘t mean that there‘s no intelligible standard of what‘s ―requisite.‖ i. Scalia‘s (majority) view: the intelligible principle test sorts unconstitutional delegations of legislative powers from constitutional delegations of nonlegislative powers (e.g., interstitial gapfilling) iv. Constitutional test: Congress can delegate quasi-legislative power as long as it gives the agency or official an ―intelligible principle‖ to follow in exercising that power. This test has been construed to permit very broad rulemaking powers to be delegated, as delegation jurisprudence is driven by practical understanding that in our increasingly complex society, Congress simply cannot do its job absent an ability to delegate power under broad general directives. The ―intelligible principle‖ may be as broad as one that directs the agency to regulate ―in the public interest‖ or to set prices that are ―generally fair and equitable.‖ v. Nondelegation rules: 1. Congress can‘t delegate what it doesn‘t have a. Congress has enumerated delegated powers, and if it lacks the power to legislate in a certain area, then it can‘t delegate that power to an agency 2. Nondelegable powers a. E.g., Senate can‘t delegate the power to impeach to an administrative agency (this is an academic theory) 3. Nondelegation doctrine a. ―intelligible principle‖ (JV Hampton & Co v US)— this an ends-means rule: Congress has to specify the ends, and agency can then permissibly pick the means. Congress can‘t let the agency also pick the ends (e.g., no Commission to Promote the Public Welfare, but a Commission to Promote Workplace Safety is fine). Court had let Congress tell the FCC to regulate telecommunications ―in the public interest.‖ This rule is about preventing principal- agent problems. b. Delegation to private parties—unconstitutional (Carter v Carter Coal, Schechter). Also about principal-agent problems. c. Subsidiary rulemaking and judicial review—agency can ―save‖ a delegation by limiting own discretion (Connolly) i. Rejected in Whitman d. Sheer breadth of authority—some delegations are just too broad (Schechter) i. Whitman Court remarks that the degree of permissible agency discretion (i.e., the intelligibility of the ―intelligible principle‖) is proportional to the breadth of the delegation‘s scope. vi. Dead doctrine? 1. After American Trucking, the nondelegation doctrine is unenforced. But it‘s not dead; courts still use it as a principle of statutory construction (i.e., prefer a certain construction of a statute to avoid a nondelegation question). a. It was ditched b/c cts couldn‘t enforce it and b/c modern life demands the administrative state—but it remains a ―phantom‖ doctrine (according to Vermeule) 2. Doctrine is a sham a. History: no evidence that framers thought the federal constitution contained a nondelegation rule, if by that we mean that the constitution restricts delegations of legislative power to executive. The first congress did these delegations. b. Formal problem: i. Distinguish ―we delegate our de jure legal powers to X‖ (which is a delegation of legislative powers) from statute granted president the authority to generate rules. Traditionally, this is also thought to be a delegation if the prez is given too much discretion. But this analogy doesn‘t work; in the 2 nd situation, the Congress has proceeded by the constitutionally prescribed matter—it has enacted a statute, exercising its legislative power. And by prescribing rules or whatever, the prez is just executing the law. And so judicial review should look at what the prez did to see if it‘s in what the legislature has prescribed. (Compare to Youngstown, where Court rejected argument that executive has inherent powers to act in the absence of statutory authority.) c. But don‘t these delegations undermine accountability, deliberativeness, allow interest group transfers, and all that? But none of these support a nondelegation doctrine. These objections are vitiated by the fact that: i. the delegating statute itself has to go through the enacting process. If a delegation of responsibility diminishes congressional accountability, then you lobby against the statute. ii. none of these objections is unique to delegations. If we think that Congress is a shill for special interests, yeah we wouldn‘t like delegations but we wouldn‘t like any other statutes either. d. The slippery slope argument to Hitler: but couldn‘t Congress pass one big statute giving the prez the power to make rules on any subject w/in Congress enumerated powers? Yes, but we count on the political checks and institutional constraints that will prevent it. And even if they did do this delegation, maybe it wouldn‘t be bad—these sorts of delegations aren‘t usually usurpations/abdications, but rather emergencies. And the delegations tend to end when the emergency ends. Nothing in con law I would stop a Hitler at the Reichstag—the nondelegation doctrine does no work. c. Congress, the President, and the Administration i. Problem: administrative agencies are balkanized. No one has control. ii. USSC has invalidated four ways that Congress tried to retain control of agency action: (1) direct appointment (Buckley v Valeo); (2) having members of Congress serve on administrative bodies (MWA v Citizens for Abatement of Aircraft Noise); (3) controlling removal (Bowsher); and (4) legislative veto (Chadha). iii. The Legislative Veto 1. Has three elements: (1) statutory delegation of power to executive; (2) exercise of that power by executive; (3) reserved power in Congress to nullify that exercise of power. 2. INS v. Chadha, 1983. Court strikes down legislative veto, in which Congress gave power to either house to nullify AG‘s decision not to deport an alien. Court says that legislative veto is essentially legislative action in purpose and effect, and must therefore comply w/ Presentment and Bicameralism clauses of Art I. a. Textually unclear/wrong. The passage of the initial bill satisfied bicameralism/presentment; exercise of veto is just execution of INS Act. i. Presentment: Art I § 7 states that every bill that has passed HR and Sen must be presented to Prez. But this bill didn‘t pass both houses – so no presentment ii. Bicameralism: Art I § 7 states that every order/etc to which the concurrence of both HR and Sen may be necessary must be presented to Prez. But this bill didn‘t require concurrence of both houses b. Nevertheless, Burger states that the one-House veto was essentially legislative in purpose and effect, and therefore is subject to presentment/bicameralism (despite the fact that neither are textually applicable). i. Burger‘s rationale: w/out the veto provision, the result in the case couldn‘t have occurred w/out a statute requiring deportation. ii. But his rationale is unable to distinguish between permissible delegations of legislative rulemaking authority and an ―impermissible‖ exercise of what is ―legislative in character/effect‖ that is subject to presentment/bicameralism clauses. Hence, Burger‘s rationale requires throwing out all delegations. c. Although wrongly decided, Chadha is good law and stands for the proposition that Congress can‘t delegate power to itself/one house. d. Alternatively, Chadha is correct, but stands for the proposition that Congress can‘t retain power to execute the laws. iv. Appointment/Removal of Executive Officers 1. Art II, § 2, cl 2, says that the Prez shall appoint all officers of the US except the Congress may vest the appointment of inferior officers as they see proper in the Prez alone, the Dept, or Courts. a. Inferior officers: appointed as Congress sees fit by Prez, heads of Depts, or courts. b. Principal officers: appointed by Prez w/ advice consent of senate. Default rule. c. Congress never gets to appoint—they get to set up the office, deciding as to inferior officers as to who appoints them, and as to principal officers the senate gets to confirm i. Congress can, however, appoint officials to help it exercise legislative power (e.g., Library of Congress). d. Members of either house can‘t be officers of US: Incompatibility Clause (Art I § 6 cl 2). i. The Metropolitan Airport case: members of Congress cannot administer the law e. An ―officer‖ includes any appointee exercising significant authority pursuant to the laws of the US. An ―employee‖ is a ―lesser functionary‖ who is ―subordinate to officers of the US.‖ Buckley. f. Factors to consider in distinguishing principal/inferior: (from Morrison i. Nature and extent of official‘s duties, and whether or not they include policymaking functions ii. Amount of independence and source of supervision; e.g., whether official answers directly to prez, to a principal officer, or to someone lower in hierarchy iii. Position‘s tenure: continuing, temporary, or intermittent iv. How official may be removed. g. Interbranch appointments i. Textually, no limit on interbranch appointments of inferior officers. Appointments Clause says ―as they think proper‖ ii. SoP limit: If appointment had the potential to impair the constitutional functions assigned to one of the branches, that would be an unconstitutional interbranch appointment. 1. Ex: no SoP problems w/ judicial appointment of election supervisors (Ex parte Siebold) or prosecutors (Morrison). Maybe you can‘t have judges appoint FERC commissioners, since these are areas w/ which judges lack expertise. 2. Removal—constitution says that you can remove prez and principal officers by impeachment. Does that mean that that‘s the only way to remove an executive officer? No, says USSC. So what other means of removal are there, and can Congress specify them? a. Caselaw i. Myers v. U.S., 1926. Myers appt postmaster under statute that requires Senate consent for removal; Prez fires w/out consent. Held: the power to remove subordinates is inherently part of the executive power, which is vested in the prez; hence, the removal provision of statute is unconstitutional. Broad view of unitary executive: dicta that prez can always unilaterally fire executive officers. While Congress could limit prez‘s removal power vis a vis inferior officers by creating a Civil Service (Art II, § 2 allows Congress to ―vest the appointment of such inferior officers as they think proper, in the Prez alone, in the cts of law, or in the heads of departments‖), but postmaster is a major office—not chosen by prez alone, but by prez w/ advice/consent of senate. Hence, his removal must be left to prez‘s discretion. ii. Humphrey’s Executor v. U.S., 1935. : FTC created to enforce antitrust laws; commissioners removable only for cause. FDR wanted to remove a commissioner to replace w/ more vigorous enforcer of antitrust laws, and fired one—but not for the statutory criteria of ―inefficiency, neglect of duty, or malfeasance in office.‖ Held: b/c FTC commissioner is a quasi- judicial/legislative officer rather than a purely executive officer, Congress may circumscribe the prez‘s removal power w/ respect to that office. Myers limited. iii. Weiner v. U.S., 1958. Congress established War Claims Commission, giving commission the ―jurisdiction to receive and adjudicate according to law‖ claims for compensating internees/etc. Commission composed of 3 people appointed by Prez w/ advice/consent of Senate. Commission was to expire no later than 3 yrs after running of SOL for claims – no removal provision made. Prez removed a commissioner. Held: Prez can‘t remove quasi-judicial officer, b/c the statute doesn‘t provide for removal and impliedly denies removal power, and b/c there is no constitutional executive removal power w/ respect to quasi-judicial officers. iv. Bowsher v. Synar, 1986. Comptroller- General, head of GAO, given the responsibility to do cuts on appropriations bills to get us w/in budget. Under statute, Comptroller General could be removed at any time by joint resolution of Congress on a ―for cause‖ standard. Held: Comptroller General‘s role in exercising executive functions under Balanced Budget Act‘s deficit reduction process violates doctrine of separation of powers b/c the Comptroller General is removable only by congressional joint resolution or by impeachment, and Congress may not retain the power of removal over an officer exercising executive powers 1. Comptroller-General is creature of Congress: Though nominated by Prez and confirmed by Senate, Comp-Gen is removable only at Congress‘s initiative. Under the ―for cause‖ standard, Congress could effectively remove Comp-Gen for any number of actual/perceived transgressions. Plus, political realities mean that Comp-Gen isn‘t free from congressional influence. 2. Comptroller-General has executive powers: Act contemplates that Comp-Gen will exercise his independent judgment w/ respect to estimates and will make decisions normally made by officers charged w/ executing a statute. The Act gives Comp-Gen rather than Prez the ultimate authority in determining what budget cuts are to be made. v. Mistretta v. U.S., 1989. Congress created the Sentencing Commission, a body composed of 7 members (including judges and non-judges) statutorily located in the judicial branch, w/ the legal power to write sentencing guidelines binding on federal judges who sentence criminal defendants. Congress wrote fairly detailed standards in the statute, and provided that Prez could remove members only ―for cause.‖ Held: Commission doesn‘t violate separation of powers; Congress can call upon Judiciary‘s accumulated wisdom/experience in creating policy uniquely w/in judiciary‘s ken. vi. Morrison v Olson. Court upheld limitations on Prez‘s removal power over independent counsels. Court substituted a functional analysis for Myers/Humprey‘s Executor distinction: ask whether restrictions on removal impede the prez‘s ability to perform constitutional duty. Court concluded that since independent counsels are inferior (as opposed to principal) officers who have limited jurisdiction/tenure/policymaking authority, the impediments don‘t restrict prez‘s ability. b. Removal rules: i. May Congress remove? No (Bowsher) ii. May Congress require Senate consent to removal? No (Myers) 1. Rationale: the unitary executive. Article II contemplates a unitary and uniform executive. If Prez can‘t fire as he sees fit, then the unitary executive is frustrated in a functional way; moreover, it is distorted in that the Prez isn‘t made the sole head of the Executive Dep‘t (since the officer would also, in a sense, answer to the Senate). a. Textual basis: Art. II, § 1: ―the executive power shall be vested in a president‖ given a variety of enumerated powers requiring a powerful and unitary judgment such as commander in chief (§ 2), and who is charged to ―take care‖ that the laws be enforced (§ 3). b. Compare Art III § 1 (dividing judicial power between supreme court and inferior courts) 2. Contra: In the Federalist Papers, Hamilton thought the opposite; whenever you need Senate approval to appoint, you need Senate approval to remove. iii. May Congress make officers removable only for cause? Yes (Weiner, Humphrey‘s Executor iv. Bottom line: 1. Congress a. Congress cannot remove executive officers except by impeachment. (Bowsher) b. Congress can restrict Prez‘s power to remove certain officers. (Myers, Humphrey‘s Executor) i. Interbranch removals permissible if no SoP problems: ii. Mistretta v US: SoP not violated where statute permitted Prez to remove members of US Sentencing Commission (including judges) for cause. The commission itself didn‘t perform judicial duties. iii. Can vest removal power in executive official other than Prez if passes functional analysis (will limit on prez‘s removal power frustrate ability to Take Care…) iv. Morrison v Olsen: SoP not violated where removal of independent counsel vested in AG, not Prez. c. Congress can remove officials who exclusively serve legislative function. 2. President a. Prez can appoint principal officers w/ advice/consent of Senate, and inferior officers if Congress so provides. i. Principal/inferior officer distinction: if principal officer not appointed by Prez w/ advice/consent, then all actions are ultra vires. Look to nature/scope of duties and whether removable by higher executive officer. Maybe also look to see if supervised by higher executive officer. ii. Morrison v Olson: Held: independent counsel is inferior officer. Factors: (1) independent counsel could be removed (for specified reasons) by higher executive branch official; (2) independent counsels had only certain, limited duties of investigation/prosecut ion; (3) independent counsel‘s office limited in jurisdiction; (4) independent counsel‘s office limited in tenure— only one investigation. b. Prez can remove ―pure‖ executive officers (Myers), but Congress can restrict removal power for ―quasi- judicial/legislative‖ officials (Humphrey’s Executor, Weiner) i. Real question is function: do the restrictions on removal impede the prez‘s ability to perform constitutional duty? If so, then impediment to prez‘s removal power is unconstitutional; if not, then impediment is fine. (Morrison) ii. Important factors: principal vs. inferior officer; scope of jurisdiction; length of tenure; amount of policymaking/adminis trative authority. c. Removal power won‘t be implied if officer‘s term already limited in time. (Weiner) v. Presidential direction (i.e., even if Prez can‘t appoint/remove, can he nevertheless direct an agency?) 1. Textual source of authority: Vesting Clause (Art II, § 1 and Take Care Clause (Art II, § 3)—from which people theorize about the nature of the unitary executive. 2. Substantive vs. procedural. Do we mean that prez can tell them what to do in a substantive or a procedural sense? a. Reagan/Clinton Executive Order 12866—all significant regulatory decisions must be subject to cost-benefit analysis. All potential regulations must be given to OIRA, which reviews regulations (but not interpretive rules and statements of policy) for cost-benefit analysis. If OIRA doesn‘t like your cost-benefit analysis, it boots it back to you. This is all procedural, and fine. b. At the end of the day, the agency says ―we‘ve done cost-benefit analysis/etc; we‘re going to do X.‖ and Prez says ―no you don‘t.‖ What happens next? i. According to executive order: this is arbitrated by VP, who can countermand agency‘s decisions ―to the extent permitted by law.‖ Unclear what this means in the context of independent agencies (and will probably be worked out in political channels, since independent agencies, as a practical matter, need good will of prez b/c agencies need prez‘s support in disputes w/ Congress). 3. Executive officer vs. independent agency. Can the prez tell independent agency heads what to do w/in their realm of statutory authority? a. Can prez tell Executive agencies what to do? What if the executive agency adjudicate cases (e.g., Prez countermands AG‘s decision to grant waiver of deportation to Chadha)? i. Myers tells us that Prez can‘t make this order; ―there may be duties of a quasi judicial character … the discharge of which prez can‘t properly influence or control.‖ But he can use the decision after the fact as a reason for removing the dep‘t head. ii. Even the strongest of unitary executive believers go with this line; the more adjudicative the decision looks, the less the Prez can issue order—even though the Prez can remove at will. d. Agency‘s power to adjudicate—under what circumstances does Art III require a court to decide a matter? i. The Rules 1. Congress might try to commit initial adjudication of either legal or factual issues to agency rather than courts. Does Congress have power to do this? (Stakes is that later court probably won‘t review de novo.) 2. Crowell says for public rights cases, yes. For private rights cases, more complicated framework. Courts would review agencies‘ initial legal decisions de novo, and for facts, they‘d get reviewed deferentially unless it was a jurisdictional or constitutional fact, in which case de novo review would be required. 3. The Crowell framework holds up pretty well; however, current law about this question is somewhat different—see Schor. Schor says, lets make this question more pragmatic: let‘s ask whether Congress had good/rational policy reasons for committing the decision to initial adjudication by agency, and does this commitment pose a threat to core functions of Article III courts? What factors do you look at to decide these questions? Policy (expertise of agency, savings of litigation costs); Threat to Judicial Independence (Crowell factors: the more review in a later proceeding, the less threat; whether the area is specialized and marginal as thought in Schor, or is Congress carving out a major chunk of Article III jurisdiction and giving it to an agency) a. This isn‘t really too much of a worry, since institutional competition between Prez and Congress might prevent Art III from being evaporated. ii. Main concern—don‘t undermine Art III courts 1. similarly, art III also prevents the converse—undermining art III courts by giving them nonjudicial things to do: a. Fed Radio Comm‘n v. GE, USSC, 1930: Congress can‘t grant a court the power to award a radio license b/c awarding licenses is an administrative function that Article III doesn‘t authorize courts to perform. iii. Constitutional arg for allowing delegation of adjudicatory power to non-Art III courts: 1. Not clear that all judicial powers have to be exercised by courts. 2. Necessary and Proper Clause gives Congress authority to give agencies adjudicatory power 3. So long as agency decisions subject to judicial review, art III not infringed 4. Greater flexibility to do this w/ public rights than private rights (Art III arguably only applies to private rights) iv. Public right/private right distinction and the Role of Art III courts 1. Early cases said that Congress could delegate adjudicatory powers to non-Art III courts in three circumstances: (1) military courts; (2) territorial courts; and (3) tribunals for adjudicating ―public rights.‖ (list from Marathon Pipeline) a. Public right—person suing gov‘t; Private rights— involve liability between one person and another. b. Rationale for allowing non-Art III courts to hear public rights cases: i. In public rights cases, person suing gov‘t on claim over which gov‘t has waived sovereign immunity. Since gov‘t has ability to avoid claim completely with sovereign immunity, but waived it, then they still have ability to have agency rather than ct adjudicate claim. 2. Crowell v. Benson, 1932. Congress lets agency decide finding of fact that triggers agency‘s jurisdiction; agency decided workers‘ comp cases for certain maritime employees. Held: to the extent that the Act permits agency to decide the jurisdictional facts that trigger the statute‘s applicability (where such application affects individual‘s private right), then court must have opportunity to determine independently all questions of fact/law upon own record. Hence, statute construed to permit court to make determination upon own record. a. Dissent in part (Brandeis)—D/P may require an Art III court to play a role in some cases, as where constitutional rights are involved. (Majority decided judicial review of such claims is enough). b. Adjunct theory—non-Art III entities can do factfinding even w/ respect to private rights so long as the legal significance of those factual determinations was subject to determination by an Art III court. (This is consequence of how USSC construed statute in Crowell) 3. Northern Pipeline Construction Co v. Marathan Pipeline Co, USSC, 1982. Restricted even further Congress‘s ability to delegate adjudicatory powers to non-Art III decisionmakers and moved away from public/private rights distinction. Case was about whether Art I bankruptcy judges could decide all legal controversies arising in or related to bankruptcy proceedings. Permitting bankruptcy court to decide related claims (e.g., contract claims) would result in administrative gains. a. Holding (from plurality/concurrence): Congress can‘t vest in a non-Art III court the power to adjudicate, render final judgment, and issue binding orders in a traditional contract action arising under state law, w/out consent of the litigants, and subject only to ordinary appellate review. b. Brennan (plurality): Can‘t do this (sustains public/private distinction). i. Art III permits delegation to non-Art III judges in 3 instances: (1) Congress may delegate adjudicatory power to territorial courts; (2) Congress may delegate that power to ―courts martial‖; (3) Congress may delegate that power to ―legislative courts and administrative agencies‖ to adjudicate ―cases involving public rights‖—rights arising between gov‘t and others. ii. Since gov‘t didn‘t have to create the public right, it can decide how those rights are administered (so long as D/P is satisfied). iii. Contract claims, e.g., can‘t do this—not a public right. iv. Crowell distinguished: there are different kinds of private rights: those that Congress created and c/l rights. When Congress creates a substantive federal right (as in Crowell), it can prescribe how that right is administered. Crowell itself required closer ct supervision than would occur in bankruptcy context. First, agency in Crowell only made specialized narrow fact determinations (bankruptcy does all relevant findings of fact in many areas of law). Second, agency in Crowell had to seek ct enforcement of its orders (bankruptcy could enforce own orders). Third, agency orders in Crowell could be set aside if unsupported by evidence (bankruptcy only subject to clearly erroneous standard). 1. i.e., Crowell involves (1) a private right that Congress had created and (2) greater art III court participation/supervision c. Rehnquist/OConnor (concurrence): Can‘t do this i. State contract claim too traditionally judicial to have Congress permit nonjudicial body resolve it d. Dissent (White, Powell, Burger) i. Functional approach: appellate review and nonpolitical nature of claims meant that traditional sep of powers concerns satisfied. Appellate review = uniformity/conformity w/ law; nonpolitical = no undue executive/legislative efforts to influence outcomes or dangerous accumulation of power in other branches 4. Commodity Futures Trading Comm‘n v. Schor, USSC, 1986: Can Congress grant to an agency (CTFC) the power to adjudicate ordinary state contract claims between two individuals? (1) Congress created new proceeding, called a reparations proceeding, where a disgruntled commodities consumer could claim damages; agency adjudicates dispute. Customer permitted to bypass procedure and sue. (2) If customer used procedure, the agency can adjudicate any counterclaims. (3) Schor used procedure, then broker brought counterclaim against him. When he lost, he challenged the constitutionality of having agency adjudicate the claim. a. O‘Connor: repudiated Marathon plurality and reaffirmed Crowell i. Art III serves to protect independent judiciary and to safeguard litigants‘ right to have claims decided before independent judges. Art III doesn‘t confer absolute right to plenary consideration of every nature of claim by an Art III court. Schor waived any personal right he had to Art III trial on counterclaim when he tried the claim before agency. ii. Declined formalistic Marathon Pipeline approach and chose pragmatic balancing test. Weigh: 1. the extent to which the essential attributes of judicial power are reserved to Art III courts 2. the extent to which the non-Art III forum exercises the range of jurisdiction/powers normally only vested in Art III courts 3. origins and importance of the right to be adjudicated a. look for voluntariness 4. the concerns that drove Congress to depart from the requirements of Art III a. look for relationship between questionable claim and nonquestionable claims b. look for big administrative efficiency gains v. ―Jurisdictional Fact‖ doctrine (sometimes called constitutional fact doctrine)—the idea, as in Crowell, that even though agencies may adjudicate otherwise private controversies subject to limited art III review, certain ―jurisdictional facts‖ must be subject to de novo judicial review. 1. A constitutional fact is a fact that is a prerequisite for the exercise of Congress‘s powers (i.e., whether something moves in interstate commerce). A jurisdictional fact is a fact whose existence triggers agency‘s authority. 2. Extremely disfavored—only on facts of Crowell (employment relation and that injury occurred on navigable waters) 3. But where important personal interests are involved, the requirement of de novo judicial review of key factual issues retained; e.g., citizenship (Ng Fung Ho v. White, USSC, 1922). a. this is really a requirement of D/P—not some art III requirement. vi. The 7th Amendment as a Limitation on Agency‘s Power to Adjudicate 1. Adjudication of public rights may be assigned to agency w/out violating 7th am (Atlas Roofing Co) 2. Probably can‘t delegate adjudication of entire criminal cases, though a. Compare Peretz v U.S. (holding that non-Art III magistrates could conduct voir dire w/ parties‘ consent); with Gomez v U.S. (Art III barred magistrate judge from selecting jury in felony trial w/out D‘s consent) III. Scope of Review a. APA § 706 (p. 1198). cts decide all relevant questions of law de novo. Agency findings subject to more deference—only overruled if agency acted in arbitrary/capricious way, or if findings are unsupported by substantial evidence. b. Questions of Fact i. APA § 706 says that agency action, findings, conclusions, will be reversed by court if: 1. Arbitrary, capricious, an abuse of discretion, or otherwise not in accordance w/ law (§ 706(2)(a)) 2. Unsupported by substantial evidence in a case subject to §§ 556-57 or otherwise reviewed on the record of an agency hearing provided by statute (§ 706(2)(e)) ii. APA generally sorts everything agencies do into four boxes: Formal Rulemaking, Informal Rulemaking, Formal Adjudication, or Informal Adjudication 1. Rulemaking v. Adjudication: Rulemaking is agency‘s exercise of legislative power and adjudication is agency‘s exercise of judicial power. 2. Formal v. Informal: Formal action takes place ―on the record‖ (after a trial type hearing), whereas informal action does not. iii. Scope of review for each: 1. Formal Rulemaking: Allentown Mack—under § 706, SE test 2. Informal Rulemaking—under § 706, arbitrary and capricious 3. Formal Adjudication: under § 706, SE test 4. Informal Adjudication—under § 706, arbitrary and capricious iv. Substantial Evidence test: ―unsupported by substantial evidence on the whole record‖ 1. Like the directed verdict test, but agency must consider the whole record. For a while in the 30s, courts felt that agency didn‘t have to consider the whole record (so long as agency‘s decision supported by some evidence in the record). In the 40s and 50s, this idea was rejected by USSC (Universal Camera); APA amended to have this ―on the whole record‖ language added to § 706. 2. If no evidence either way, then agency‘s decision stands— either way: agency can always cite its expertise as a reason, or credibility of witnesses. SE test is very deferential. 3. The SE test isn‘t just a function of the evidence in the record, but also a function of the agency‘s policy judgments. Agencies can adopt presumptions about classes of cases to encourage the results that they want. 4. In applying SE test, agency is held to the announced evidentiary standards. a. Allentown Mack v. NLRB, 1998. NLRB made factual determinations that employer lacked good faith doubt in union‘s majority support among employees; announced evidentiary standard was ―good faith reasonable doubt,‖ but NLRB really required a strict head count. Held: NLRB‘s factual determinations weren‘t supported by substantial evidence in the record b/c no reasonable jury could have reached NLRB‘s conclusion, and b/c even though NLRB was effectively using evidentiary standards stricter than those announced by NLRB (strict head count), it is held to the standard it has announced (―good faith reasonable doubt‖). Court noted that NLRB could have, if it wanted, announced the strict head count standard (in which case, the case would‘ve gone the other way). i. This isn‘t a case about agency using an evidentiary presumption; rather, it is where the agency systematically undervalues certain evidence as if there were a presumption. ii. Vermuele thinks that this is wrong; there is no such requirement that presumptions be adopted like Scalia wants. Agencies are actually free to apply a new policy to a case it is adjudicating for the first time. If an agency can create a new evidentiary rule, then isn‘t it free to place a thumb on the evidentiary scale? v. Burdens of Persuasion and Burdens of Proof 1. Be careful to distinguish the burden of proof from the standard of review used by court in reviewing the decision of the trier of fact 2. APA § 556(d) provides that ―Except as otherwise provided by statute, the proponent of a rule or order has the burden of proof.‖ This includes the burden of persuasion and the burden of production. vi. Relevance of ALJ‘s factfindings 1. ALJs created b/c it was thought unfair to have all adjudications decided by the agency commissioner, who might be biased in a particular case b/c they combined policymaking, investigative, prosecutorial, and adjudicatory responsibilities. 2. Concern: a. Too little court deference to ALJ and agency head doesn‘t have to pay any attention to them; ALJs become mere evidence collectors b. Too much court deference to ALJ and virtues of administrative factfinding lost; commissioners can better consult experts, etc. It‘s proper for agencies to use individual cases to change policy. Commissioners may know about biases of ALJ and know when to set aside her findings. 3. Universal Camera: give ALJ‘s findings some undefined weight as a relevant part of the record, and set aside the agency‘s decision if it fails the substantial evidence test. c. Questions of Law i. The Chevron Era 1. APA § 706: ―The reviewing court shall decide all relevant questions of law, interpret … statutory provisions….‖ 2. Pre-Chevron caselaw gelled into two platitudes exerting countervailing gravitational pulls on the law. At one pole stands the maxim that courts should defer to ―reasonable‖ agency interpretive positions. Pulling in the other direction is the principle that courts remain the final arbitrators of statutory meaning. Courts generally took a statute-by- statute intentionalist/originalist approach, and credited agency interpretations if they were contemporaneous w/ enactment and had been consistent over time. 3. Legislative rules vs. Agency Interpretations: A legislative rule is the product of exercise of delegated legislative powers (i.e., the product of formal rulemaking). An interpretive rule is any rule an agency issues w/out exercising delegated rulemaking power (i.e., an informal internal guideline). a. Legislative rules—binding b. Interpretive rules—persuasive (Skidmore) c. How to tell the difference: i. Does rule itself provide enforcement action to ensure performance of duties 1. i.e., take away the rule: can agency still prosecute person for violating reg? if so, rule is interpretive ii. Is rule published in Code of Federal Regulations 1. justification: Fed‘l Register Act describers docs required to be published as agency docs having ―general applicability and legal effect.‖ 2. Dubious: while Act describes what must be published, it doesn‘t prohibit other docs from being published there; there is positive value in having agency‘s formal statements of its interpretations published where the public may more easily find them. iii. Agency tells you it‘s a legislative rule iv. Rule effectively amends a prior legislative rule 1. make sure that the agency didn‘t make a legislative rule w/ a catch-all provision that gives itself the power to avoid later notice-and-comment rulemaking; i.e., is the ―law‖ the agency claims to be interpreting too broad to support the specific ―interpretation‖ the agency gave to it? 4. Skidmore vs. Swift, 1944. Administrator of FLSA adopted flexible rule as to when waiting time counted as overtime; trial ct didn‘t pay any attention and adopted blanket rule that waiting time wasn‘t ever overtime. Held: court should defer to administrator. Although administrator‘s interpretations don‘t control courts, they do constitute a body of experience and informed judgment to which courts can resort for guidance, weighing the agency interpretation according to: (1) the thoroughness evident in its consideration; (2) the validity of its reasoning; (3) its consistency with earlier and later pronouncements; and (4) all those factors which give it power to persuade, if lacking power to control. a. Justifications: i. Agency‘s specialized knowledge: interpretation made in pursuance of official duty, based on specialized info that is better than what judge would have. ii. Rule of law reasons: good administration of FLSA and good judicial administration alike require that the standards of public enforcement and those for determining private rights shall be different only for good reason—the fact that administrator‘s reasons not reached through adversary proceeding isn‘t a good reason. iii. Persuasiveness as evidence of enacting Congress‘s intent: Administrator‘s interpretations aren‘t binding, but they are influential to a degree determined by their thoroughness of consideration, the validity of their reasoning, their consistency w/ earlier and later pronouncements, and so on. ii. Chevron and the Interpretive method 1. Chevron v Natural Resources Defense Council, 1984. EPA changed its construction of a statutory term, ―stationary source,‖ adopting bubble theory for factories (a legislative rule). Held: reviewing court should defer to agency construction of statute so long as statute doesn‘t speak expressly to issue and agency‘s interpretation isn‘t unreasonable. a. Chevron two step: i. Step one: using traditional tools of statutory interpretation, see if Congress already spoke on precise issue. If yes, go w/ that. If not… ii. Step two: if agency‘s construction is permissible (even if not one court prefers), then it‘s binding. Congress left an explicit/implicit gap, and this is delegation to agency to fill gap. b. Drops legislative rule/interpretive rule distinction and replaces statute-by-statute analysis w/ global rule (supposedly, but skidmore‘s still around…). c. Justifications for this default presumption that Congress intended agency discretion in construing statutes i. Delegation theory: Congress has implicitly told courts to defer to agency interpretations of law; i.e., that Congress, as a general rule, has given administrative agencies authority to resolve ambiguities in statutes. 1. Where does congress say this? APA § 706: ―The reviewing court shall decide all relevant questions of law….‖ There‘s not evidence for this sort of delegation argument. ii. Comparative advantage: agencies have comparative advantage over courts in statutory interpretation, b/c of their political accountability and technical specialization. 1. But just b/c this is true doesn‘t make it theoretically justified. If it is the constitutional duty of the courts to say what the law is, then we have to look beyond relative competence as a basis for ignoring that principle when agency action is at issue iii. Promote uniformity: a rule of agency deference can reduce the balkanization of federal administrative law by limiting circuit splits 1. Again, just b/c this is true doesn‘t make it theoretically justified. If it is the constitutional duty of the courts to say what the law is, then uniformity isn‘t basis for ignoring that principle when agency action is at issue iv. Separation of powers: Another view: Separation of powers requires Chevron— when there is ambiguity in statute, the resolution of that ambiguity requires policy judgments, and these judgments aren‘t for courts to make, but rather for political branches. In a world where Congress has abdicated and delegated huge policy goals/powers to agencies, then we shouldn‘t expect statutes to answer all of the questions that arise. Either courts or agencies have to fill in the gaps in statutory scheme; congress won‘t do it. Pre-Chevron idea is that courts fill in the gaps (through an originalist/intentionalist view of legislation). Chevron is a consequentialist approach to statutory interpretation: there isn‘t a correct answer to statutory interpretation, but some answers advance different policy goals—and let agency choose among them, because it is the branch that is politically accountable. 1. Scalia: This is nonsense and can‘t be a constitutional mandate: courts often turn to policy considerations. Moreover, if there were constitutional objections to policymaking, then would USSC be unable to interpret Clean Air Act if Congress had included proviso saying that in all cases of ambiguity, the USSC shall give no deference to agency‘s views? a. But Chevron doesn‘t mean that courts are constitutionally required to defer to agencies; rather, the Chevron rule is merely a pragmatic one about the division of decisionmaking labor. Courts are free to give policymaking discretion to agencies for pragmatic reasons. b. Scalia likes Chevron because of its rule-like character—it replaces case-by-case statutory scheme w/ global rule to defer to agencies. Again, this is another pragmatic reason to like Chevron iii. After Chevron? 1. Chevron step 1—the traditional tools of statutory construction expands a. INS v. Cardoza Fonseca, 1987. Longstanding INS rule was that two portions of organic statute called for same legal standard. Held: on Chevron step 1, the two standards aren‘t the same and require different standards. Looks to legislative history, and musters all the tools of statutory construction. i. The problem that‘s emerging is that the ―traditional rules of statutory construction‖ are all-encompassing. If you resort to them (especially in elaborate detail), the statute‘s very likely not going to be ambiguous. Chevron has to be different than ordinary statutory interpretation cases where there is no agency to defer to; however, here, in a chevron case, the court musters all its tools of statutory interpretation, the court will almost always come to an answer. That‘s b/c the tools come up w/ a preferred meaning w/in a range of permissible meanings; but this shrinks agency‘s discretion to a point— that preferred meaning. But Chevron has to mean that some of the tools of statutory interpretation aren‘t used; the idea of Chevron is that agency‘s have some sort of range to work in. b. Step 1 includes policy considerations: i. Dole v. United Steelworkers, 1990. Dept of Labor wanted to make mfgs assemble data for their workers; OMB disagreed, but its approval was only necessary if Paperwork Reduction Act applied, which defined an ―information collection request‖ as including a ―recordkeeping requirement‖ or a ―collection of information requirement,‖ which was defined as including the obtaining of facts by an agency. USSC said Paperwork Act didn‘t apply, since the words ―by an agency‖ meant that info had to be collected for agency, not 3d parties (e.g., workers). No Chevron deference since this was pure question of statutory construction, and the traditional tools of statutory construction—including policy considerations—revealed Congress‘s intent. Dissenters hooted that majority had to write for more than 10 pages to conclude that statute was ―unambiguous.‖ c. Step 1 includes other statutes: i. Maislin Industries v. Primary Steel, 1990. Statute A forbids carrier to charge different rate than that filed w/ ICC. ICC interpreted Statute B, which required that carrier‘s ―practices‖ to be ―reasonable,‖ authorized it to forbid carriers to charge a shipper a lower rate then later try to collect a higher charge that it falsely claimed to have filed w/ ICC. USSC said Statute A forbid ICC‘s interpretation of Statute B. d. Step 1 includes everything: legislative history, broad purposes, canons of construction. i. Babbitt v. Sweet Home, 1995. Agency decided that ―take‖ in Endangered Species Act criminalized habitat destruction. Held: agency position upheld under Chevron Step 2; all tools mustered to support notion that it was permissible construction of statute. Dissenters muster everything to show that it isn‘t e. Step 1 includes the broad regulatory scheme: i. FDA v. Brown & Williamson, 2000. In 1996, after having disavowed any such authority since its inception, FDA asserted jurisdiction to regulate tobacco, concluding that nicotine is a ―drug‖ w/ in the meaning of organic statute. Held: FDA position rejected at Step 1; Congress‘s intent to preclude FDA from asserting jurisdiction over tobacco is clear (b/c if tobacco was w/in Act, FDA would have to ban it instead of just regulating it); it would be inconsistent w/ FDCA‘s overall regulatory scheme and in the tobacco-specific legislation it enacted after organic statute. Court emphasizes that this is an ―extraordinary case.‖ Dissent says that textually, tobacco is a drug; points out that majority‘s rationale (organic statute doesn‘t fit tobacco b/c it requires FDA to prohibit tobacco b/c they are dangerous, and yet the agency concedes that a ban isn‘t the proper remedy) proves the opposite point: that FDA does have jurisdiction but that it must ban cigarettes 1. Maybe this is all a nondelegation idea: no Chevron deference if things undercut the premises supporting Chevron: a. Breadth of claimed authority and merely cryptic authorization for agency‘s view = no implicit congressional authorization for FDA‘s view; ergo, no Chevron deference. b. But really, what‘s cryptic about it? Really, all we‘re left with is the breadth idea— this is too big for the FDA to regulate; we have to wait for Congress to make it more explicit. This is repeating a nondelegation argument in statutory interpretation guise (which is terrible, since the USSC doesn‘t ever decide the relevant issue). 2. No deference to ad hoc agency litigation positions or where agency is prosecutor a. Bowen v. Georgetown Univ. Hosp, 1988. Ct refused to give deference to agency‘s interpretation of statute offered in litigation, since the ―agency itself had articulated no position on the question … Deference to what appears to be nothing more than an agency‘s convenient litigating position would be entirely inappropriate.‖ 3. Substantive limit on how much agency can change a statute under Chevron deference. Chevron vs. the Nondelegation Doctrine a. MCI v. AT&T, 1994. Communications Act of 1934 allows FCC to modify any requirement. FCC modified it by saying that only AT&T was subject to statute, arguing that its interpretation was reasonable b/c one dictionary‘s definition of ―modify‖ made that term ambiguous. Held: FCC interpretation rejected; one oddball dictionary that postdates the statute‘s enactment does not an ambiguity make. Court construes statute under Step 1 to preclude FCC‘s interpretation b/c, (implicitly), otherwise there‘d be a nondelegation problem. Dissent argued that there wasn‘t a problem since FCC had arrived at this position thru incremental stages, so whether this reg is a modification depends if you choose the statute or the past reg as the baseline of measurement. i. This is actually a nondelegation case: if we interpret ―modify‖ to permit agency to change the statutory scheme so much that it can unregulated 60% of the telecommunications industry, then we‘ve got a nondelegation doctrine problem b/c the agency‘s discretion is unlimited, and we‘re not going to assume that Congress did this sub silentio. ii. This is the deepest threat to Chevron. Why? It‘s a fantasy to think that Congress will fill in gaps and ambiguities, and so agencies have to do it (Chevron idea). Nondelegation idea is that we don‘t let agencies do this— demand that Congress does it (remember that the nondelegation doctrine was a way of policing Congress to force it to pull its weight and pass tighter statutes). 4. Make sure Congress had intent to delegate a. U.S. v. Mead Corp, 2001. Before the entry of goods, Customs issues ―ruling letters‖ that are the agency‘s official position w/ respect to that transaction, and to be applied only to identical goods. Ruling letters can be modified w/out notice, and statute provides that no one should rely on the ruling letter. These letters aren‘t published, but must be made available for public inspection. Typically, ruling letters contain little or no reasoning. After 4 yrs Customs Service switched how day planners are counted, justifying its switch on ambiguity in dictionary definitions. Held: no Chevron deference to Customs‘s interpretation; Chevron deference applies only when it appears that Congress delegated authority to the agency generally to make rules carrying the force of law and that the agency interpretation was promulgated by an exercise of that authority; this delegation may be show in a variety of ways, including power to adjudicate or notice-and-comment rulemaking; the ruling letter fails to qualify, but it may deserve deference under Skidmore. i. Dissent: this retreat to statute-by-statute process and revival of Skidmore is bad for 4 reasons: (1) Uncertainty: it‘s unclear how much formality is enough (which is a necessary evil since we‘re pretending that we‘re still applying Chevron); (2) Artificially induced increase in informal rulemaking. Now agencies have to employ rulemaking to get deference (but shouldn‘t conferral of authority be enough?); (3) Ossification of statutory law: Chevron at least allowed interpretations to change. But now we‘ve exempted many statutes from Chevron deference—so they‘ll be static…; and (4) Compounds confusion by resurrecting Skidmore, which is a statement of the obvious: a judge should take into account the well-considered views of experts. Skidmore is too wishy washy for modern law. ii. Adds new step to Chevron: Chevron pre- Step 1: did Congress intend to delegate interpretive authority to the agency? 1. To decide this, we look at procedural formalities chosen by Congress for the agency to use; and we look to the legal effect that the rulemaking/adjudication/whatnot has, to see if they give rise to an inference of an implicit delegation a. Majority suggests that if agency‘s decision is procedurally formal, then we‘ll find an implicit delegation. Informal adjudication doesn‘t so imply (it encompasses almost everything the agency does whenever someone calls up the IRS on the phone). b. Court notes that even if you don‘t use formal adjudication/rulemaking, you may still be in. If your informal adjudication has the right kind of legal effect (i.e., has precedential weight and so forth), it might have enough to raise an inference that Congress intended the agency to have the power to make law. 2. Scalia thinks this kills Chevron—but maybe it saves it. We‘ve seen how the court has often forgotten Chevron—so maybe if we restrict Chevron to a narrower category of cases, then courts would apply chevron w/ greater force. i.e., we‘re reducing Chevron‘s scope to increase its weight. 5. Ways to limit Chevron—the more you fold these (all traditional tools of statutory interpretation) into step I of Chevron, the more attenuated the doctrine becomes. a. Expansive, rich readings of text and canons of statutory construction (Babbitt, Cardoza Fonseca) b. Legislative History (Babbitt, Cardoza Fonseca) c. Invoke statutory purposes (Babbitt, Dole) d. Broader regulatory scheme and other statutes (Brown & Williamson Tobacco, Maislin) e. Nondelegation doctrine as a canon of construction (e.g., MCI, Brown & Williamson) i. This is for extraordinary cases—Brown & Williamson f. Pre-Step 1: Did Congress have intent to delegate? (Mead) i. Chevron applies to legal interpretations adopted in adjudications, but Mead suggests that not all interpretations adopted in adjudication require deference. g. No deference to agency litigating positions h. No deference where agency is prosecutor i. Agencies can‘t interpret APA j. Agency doesn‘t receive deference if it‘s interpreting a statute enforced by many agencies d. Arbitrary and Capricious (―Hard Look?) Review i. APA: courts shall hold unlawful and set aside agency action that is ―arbitrary, capricious, [or] an abuse of discretion.‖ § 706(2)(a) ii. This is judicial review of policy decisions; we‘re looking at agency‘s decision w/in the bounds of discretion granted to agency by congress 1. The arbitrary/capricious standard is essentially procedural—it requires proof that agency exercised its independent judgment and considered al the arguments, etc. 2. Different than Chevron: If agency‘s choice is found impermissible under Chevron, agency must have new choice; under hard-look, agency can come back with exact same choice, but just needs to explain it with different reasons that justify it to the court 3. Review is ―a substantial inquiry‖: ―a thorough, probing, in- depth review, and [a] searching and careful [inquiry into the facts].‖ Overton Park. 4. Looking to see if decision was (1) based on consideration of relevant factors and (2) clear error judgment. However, presumption in favor of agency; court can‘t substitute its judgment for that of the agency. Overton Park. iii. Under APA, no record requirement for informal adjudication; nevertheless, to conduct arbitrary & capricious review, court needs a record 1. Citizens to Preserve Overton Park v. Volpe, 1971. Under organic statute, agency prohibited from authorizing the use of fed‘l funds to finance the construction of highways thru parks if a ―feasible and prudent‖ alternative route exists. If none, construction may be approved only if there has been ―all possible planning to minimize harm‖ to park. Sec‘y approved funds to build highway thru Overton park (if it went around park, you‘d have to displace poor minorities‘ residences). Route was approved by lower officials, and then Sec‘y said ok once the state acquired right-of-way on both sides of the park. Sec‘y never announced findings why there was no feasible prudent alternative and why design changes couldn‘t be made to reduce harm to park. Held: While formal findings not required, Sec‘y‘s acts are subject to judicial review and there must be more records than just litigation affidavits a. Conceives of review as multi-step process: First, construe the relevant statute to determine the scope/terms of agency‘s authority to determine whether agency acted w/in the authority conferred (may be subject to Chevron deference); Second, if agency has discretion to choose among alternatives, court must determine whether the agency exercised its discretion based upon consideration of relevant factors. Third, even if agency had discretion and based its decision on relevant factors, was the weighing ―arbitrary/capricious‖? iv. The Relevant Factors 1. Just b/c ct remands back to agency to reconsider based on consideration of the relevant factors, that doesn‘t mean that the agency will reach a different result a. National Coalition against Misuse of Pesticides v. Thomas, DC Cir, 1987 and 1990. EPA set zero tolerance for a pesticide, then raised it b/c zero tolerance would destroy economies of less developed nations exporting mangoes. DC Cir vacated agency action b/c agency wasn‘t given authority to consider the economic welfare of foreign nations in setting pesticide levels. EPA then did the same thing, but this time said that a zero tolerance level could adversely affect public health b/c it could damage international relations upon which food safety laws depend. EPA‘s action then upheld. 2. Only have to consider the organic statute, not all possibly relevant areas of law a. Pension Benefit Guaranty Corp v. LTV Corp, 1990: agency action reversed b/c decision was ―arbitrary‖ for failing to consider the impact of diverse areas of law. USSC reversed this, b/c the statute only called for the agency to act ―where appropriate and consistent w/ its duties under this title.‖ As a matter of administrative law, ―if agency action may be disturbed whenever a reviewing court is able to point to an arguably relevant statutory policy that was not explicitly considered, then [given the large number of federal statutes and policies] a very large number of agency decisions might be open to judicial invalidation.‖ b. This might justify greater presidential control of agency decisionmaking to ensure consistency among related policies for which different agencies are responsible 3. Reviewing court evaluates agency‘s decision on the record before the agency at time of decision, unless: a. Agency doesn‘t provide an opportunity for a person to provide info for the record, and person has material info that might affect decision. b. Person raises serious claim as to integrity of decisionmaking process (e.g., bias, prejudice, wrongful influence). v. A ―Clear Error of Judgment‖ 1. Not only must consider all factors, must connect choice to factors a. Motor Vehicles Mfg Assoc. v. State Farm, 1983. Question was whether NHTSA acted arbitrarily/capriciously in revoking requirement that cars be equipped with passive restraints. Holding: yes; agency failed to present an adequate basis and explanation for rescinding the passive restraint requirement. Not only must agency examine the relevant data, it must articulate a satisfactory explanation for its action including a rational connection between the facts found and the choice made. Normally, an agency rule would be arbitrary if agency: i. relied on factors Congress hadn‘t intended for it to consider, ii. entirely failed to consider an important aspect of problem, iii. offered an explanation counter to the evidence, or iv. is so implausible that it couldn‘t be ascribed to a difference in view or the product of agency expertise. IV. Procedural Requirements a. Constitutional Requirements i. Does Constitution require agencies to make rules? Only in Free Speech cases subject to overbreadth challenge. 1. Boyce Motor Lines v US, 1952. Statute calls for ICC to regulate transportation of explosives, providing that ―Such regulations shall be in accord w/ the best-known practicable means for securing safety in transit.‖ ICC‘s reg just replicated the statute, calling for truckers to use prearranged routes when practicable, and criminalizing knowing violations of reg. Held: knowledge requirement and the long history of regulation of this area vitiates the lack of notice problem. a. Davis argument: const. requires agencies to do so to fix overly broad delegations such as this. b. USSC: no it doesn‘t; Whitman v American Trucking – agency‘s rulemaking doesn‘t fix any sort of problem w/ delegation. Even if agency adopts a narrowing rule-like regulation, the agency had the discretion to adopt a different regulation— and that presents a constitutional nondelegation problem. But USSC upholds statute—even a broad standard is enough to survive c. This case runs together the vagueness and nondelegation arguments. After Whitman, there is no nondelegation problem. But we still want agencies to proceed through promulgating specific rather than general cases. This Rule of Law crowd appeals to vagueness doctrine of due process clause for this constraint. But nothing seems to support this requirement with respect to agencies in particular (i.e., if regulation is vague, then statute is vague as well, so why strike one and not the other?). (Idea is similar to nondelegation rebuttal: if there is any vague reg that agency could adopt, then statute can‘t be saved by a narrowing rule-like regulation.) 2. Forsyth County v The Movement, 1992. To prevent violence problem w/ civil rights marches, to get a license, you had to pay a fee set by the Board up to a $1K cap. Held: statute violates 1st am. in that it permits a gov‘t administrator to vary the fee for parading to reflect the estimated cost of maintaining public order. a. Another attempt to find constitutional basis for requirement that agencies must enact specific rules. This is a free speech nondelegation case; where administrative discretion impinges on free speech, then we‘ll have special nondelegation doctrine. ii. Due Process Clause: Londoner/Bi-Metallic: D/P puts few constraints on rule making (legislative process is enough); more on adjudication 1. Cases: a. Londoner v Denver, USSC, 1908. State statute provided that Board might, after notice and opportunity for hearing, order the paving on petition of a majority of property owners, so long as order approved and implemented through adoption by City Council, of ordinance authorizing the paving. Following completion of paving, Board determines total costs and apportions it among property owners. Property owners filed written objections, saying that procedures weren‘t followed. Held: D/P applies, and something more than just the opportunity to provide written objections required to satisfy D/P. b. Bi-Metallic Investment Co v State Board of Equalization, USSC, 1915. Board increases valuation of all real estate in Denver by 40% w/out giving taxpayers opportunity to be heard. Held: D/P doesn‘t apply to general lawmaking. Londoner v. Denver is different b/c there, a relatively small number of persons were concerned, and who were exceptionally affected, in each case upon individual grounds. c. Southern Railway v VA, USSC, 1933. Whenever agency thinks RR crossing is unsafe, can require RR to replace grading w/out notice. Held: statute entitling administrator w/out notice/hearing to command a RR to replace grade w/ overhead crossing when administrator thinks justified by public safety violates D/P. 2. Hence, D/P applies only when adjudicative facts rather than legislative facts are at issue. D/P requires hearing in adjudications (not as broad as APA sense), not rulemaking. a. Londoner had fewer people, affected on unique grounds, and thus had relevant info to disclose to decisionmaker (adjudicative facts). In Bi-Metallic, everyone was affected, and thus no one especially had good info for decisionmaker (legislative facts). i. Why not just appoint a representative rather than have no one in cases like Bi-Metallic? ii. In adjudications, can‘t rely on political process to protect people, but you can in general rulemaking b. In Southern Railway, there was rulemaking in the abstract, and the question was adjudicative: applying rule to facts. RR had info for decisionmaker (whether grading was safe). 3. Constitutional right to hearing: a. If you are singled out, and can produce adjudicatory facts that will affect outcome, you have a constitutional right to a hearing b. Constitution does not grant to members of public generally a right to be heard by public bodies making decisions of policy; Inherent in republican form of gov‘t that direct public participation is limited. c. If agency promulgates a rule, hearing rights will only attach to whether the facts fit them, not whether the rule is a valid interpretation of the statute. This means that as a constitutional matter, there‘s a distinction between adjudication and rulemaking, in that rulemaking isn‘t really subject to constitutional constraints. i. The question of whether a rule just applies to one person or applies generally is decided not substantively but formally. So long as rule formally applies generally, that‘s enough to make it a Bi Metallic scenario. Florida East Coast Highway. ii. A large part of the mission of APA is to subject rulemaking to procedural constraints that the constitution doesn‘t subject it to. So the constitutional rules in the background are pretty weak and don‘t provide much constraint. b. Rulemaking vs. Adjudication i. Adjudication = agency process for issuing an ―order.‖ Order defined as final agency disposition of everything other than rulemaking but including licensing ii. Rulemaking: ―rule‖ defined: § 551(4): whole or part of agency statement of general or particular applicability and future effect designed to implement/interpret/prescribe law/policy. 1. Constitutional difference between rulemaking and adjudication is different than APA difference (adjudication is much broader in APA) a. E.g., Overton Park (the highway thru the park case). In the constitutional framework of Londoner/bi- metallic, this is rulemaking (this isn‘t like agency requiring a party to give money to state; there‘s no one person being singled out, there are lots of interested party, though). In the APA framework, this is adjudication b/c agency is issuing an order in the APA sense (and the agency isn‘t engaging in rulemaking). iii. Sources of procedural requirements: 1. Organic statute creating agency or vesting it w/ powers a. Watch for interplay between APA and organic statute: e.g., (1) APA doesn‘t of its own force ever require formal procedures—it just says ―if there is a certain type of trigger in organic statute,‖ then formal proceedings are required (magic words: ―Decision on the record after opportunity for agency hearing‖); (2) organic statute can be independent source of procedures b. Question arises if agency receives any deference as to procedures they say their statute requires 2. If agency adopted procedural regulations, it must follow them (Arizona Grocery) 3. APA‘s procedural requirements of general applicability 4. Federal common law a. Sometimes, DC Cir require agencies to use different procedures, then don‘t spell out source of law for these holdings… (which is formally illegitimate) b. Vermont v Yankee: c/l of agency procedure is illegitimate (yet nevertheless continues to exist b/c DC Cir has weird jurisdiction that gives it unique advisory role over development of administrative law) 5. Judicially defined constitutional requirements of due process (Londoner/BiMetallic) iv. 4 basic categories of administrative procedures: 1. Formal On-the-record adjudication a. If statute governing agency‘s action in an adjudication requires that decision be made on the record after opportunity for agency hearing, requirements of §554 apply, which in turn invoke §556-57 i. These three sections establish a set of trial- type procedures for formal adjudication ii. Normally conducted by an administrative law judge who prepares an initial decision, which then may be appealed to agency head, based on the record and submission at the hearing iii. §706(2)(E) provides for judicial review of agency fact-finding under substantial evidence standard b. Courts tend to interpret the statute as providing for a hearing on the record in cases where the agency is imposing a sanction or liability on a party i. Presumption in favor of requiring formal on- the-record adjudication may be undercut by the rulemaking cases, e.g., Florida East Coast, which only require on-the-record rulemaking when the organic statute explicitly provides for a hearing on the record. ii. Seacoast AntiPollution League v Costle, USSC, 1972. Clean Water Act provided for decision ―after opportunity for a public hearing‖ but didn‘t require on the record. USSC said that EPA, in making licensing decisions, have to have the hearing in §§ 556-57. 1. APA explicitly includes licensing as adjudication. So look at organic statute for magic words… and find part of them. So this case is an outlier. USSC says ―we‘re not gutting the magic words requirement, we just say that point of magic words requirement is to exclude governmental functions that traditionally aren‘t subject to these procedures.‖ 2. Arg: magic words test is required in rulemaking b/c rulemaking is administrative equivalent of legislation, where hearings/etc don‘t normally follow trial-like procedures. Adjudication, by contrast, is administrative equivalent of judicial determination, and hence there should be presumption for formal adjudication. 3. But isn‘t the bright line magic words test better? In, e.g., licensing, all the relevant questions are policy questions. And so it‘s not susceptible to oral testimony, etc. iii. Lower courts divided on this question: Compare Seacoast (where ―after opportunity for a public hearing‖ required formal adjudication) with Chemical Waste Management (where ―public hearing‖ did not require formal adjudication). 2. Formal On-the-record rulemaking a. In cases where relevant statute provides that rules be made on the record after opportunity for an agency hearing, §553(c) requires that an agency engaged in rulemaking use the procedures of §556 and §557 b. Traditional understanding was that formal rulemaking was generally required in setting rates and similar requirements that determined revenues and profits of regulated firms i. This was restricted by USSC in Florida East Coast, when they said statute must explicitly provide for a hearing on the record 3. Informal Notice and Comment Rulemaking a. In a case of rulemaking where the applicable statute does not provide for a hearing on the record, the basic procedures for rulemaking is the notice and comment procedure provided in §553, which calls for: i. General notice of proposed rulemaking in the Federal Register 1. Doesn‘t apply to interpretative rules or general statements of policy ii. Opportunity for interested persons to comment iii. Concise general statement of basis and purpose once rule is promulgated iv. Shall not be effective in less than 30 days after promulgation v. §553 does provide for exceptions though where agency may avoid these requirements b. Purpose is to enlighten decision-maker by exposure to viewpoints of interested persons c. Substantial evidence standard of review does NOT apply d. Agencies have increasingly turned from adjudication to rulemaking in order to decide basic issues or regulatory policy 4. Informal adjudication a. In instances where relevant statute does not require adjudicatory decisions be made on the record after opportunity for agency hearing, there are no APA standards to follow b. Overton Park stated that review should be based on administrative record under an ―arbitrary and capricious‖ standard c. Remember, if adjudication is really informal, agency position doesn‘t get deference (Meade) v. Traditionally, most major regulatory agencies relied on adjudication; recent shift to using rulemaking. 1. Agency‘s point of view a. If you want to avoid coherence (b/c, e.g., you want to avoid saying ―Union always wins,‖), you can just proceed case-by-case. So go to adjudication b. But lower enforcement costs with rulemaking. The major difference between rulemaking and adjudication is the effect the process has on the content of the questions parties are allowed to present to agency. If agency can proceed by rulemaking, people who are later prosecuted can‘t challenge the content of the rule—only the validity of the rule. Compliance w/ the rule is really the only defense. i. This complies w/ D/P because D/P allows the gov‘t to parcel out your hearing among a variety of different forums and different stages/times. (―we held a rulemaking hearing, but you didn‘t attend.‖) 2. Social point of view a. Efficiency and coherence of policy b. Avoids the scapegoat problem—announcing broad rule in particular cases. c. Rule-making more ―open‖ than adjudication: more people can participate, but this is less true since we permit multi-party adjudication d. Minimizes delay: cheaper enforcement costs 3. Presumption that organic statute permits agency to engage in substantive rulemaking a. National Petroleum Refiners v FTC, DC Cir, 1973. Organic statute tells FTC to proceed by adjudication, but authorizes it to make rules/regs. FTC promulgated substantive rules of business conduct, aka Trade Regulation Rules, designed to give greater specificity and clarity to the broad standard of illegality in its organic statute. Held: FTC has authority to promulgate broad substantive rules under organic statute. Lots of dicta on why it‘s nice for Rule of Law reasons for agencies to proceed by enacting legislative rules rather than by proceeding case-by-case. c. On-the-Record Rulemaking i. Agencies must be consistent and agency action may only be sustained on the reasons that the agency itself articulated 1. SEC v Chenery Corp I. Statute calls for public utility reorganizations to by certified by SEC as fair and equitable. SEC vetoed Chenery‘s reorganization initially, explaining that the reorganization plan violates equitable principles (a totally bogus arg); on appeal, they tried to justify action based on policy expertise (a valid arg). Held: agency can‘t change the rationale for its decision on appeal; if it made decision based on principles of equity as would be applied by court, reviewing court does not have to defer to public policy args raised after the fact in litigation. a. An agency action may only be sustained on the reasons that the agency itself articulated. Why? SoP—court can‘t exercise agency‘s powers: if agency doesn‘t exercise policy judgment, court can‘t do it for it. b. Result: agencies leave huge paper trail. 2. SEC v Chenery Corp II. On remand, SEC did same thing, but this time justified action on policy expertise, announcing that it wasn‘t promulgating a rule to govern these situations, but would go case-by-case. Held: agency action upheld. Agencies have discretion to choose between rulemaking and adjudication. There‘s a reason to go case- by-case: to avoid premature judgment (general problem of advisory opinions). ii. On the record rulemaking only triggered by magic words in organic statute: ―hearing on the record‖ 1. US v Florida East Coast Highway, USSC, 1973. ICC empowered to set reasonable rates for RR carriers after a ―hearing,‖ which ICC interpreted to mean informal paper rulemaking hearing. Held: the ―hearing‖ received by RRS, permitting them to file written responses, was adequate b/c it didn‘t trigger formal rulemaking procedural requirements from APA. D/P doesn‘t require a hearing b/c formally the rule applied generally (and thus this case is controlled by Bi-Metallic). Organic statute doesn‘t require more than just opportunity to file written objections; though originalist understanding of ―hearing‖ might require more, the default rule is presumption that notice-and-comment is fine (don‘t want to tie agencies down, and Congress can always explicitly opt out). a. Note that this case also inquires into whether additional procedural requirements are imposed by (1) APA, (2) D/P, and (3) organic statute. b. Why not defer to agency as to what procedures organic statute requires? i. Mead framework used procedural formality as evidence of Congressional intent to delegate authority. But this is circular—b/c we need to know also what the meaning of these procedures are. Maybe, for example, we adopt global presumption that Congress hasn‘t delegated authority to define procedures to agency. This is like deferring to agency on its own jurisdiction. Florida East Coast‘s holding is consistent w/ notion that no agency has interpretive authority over meaning of APA, b/c deciding the organic question here (what does ―hearing‖ mean) really also decides the APA question. (of whether Congress intended by ―hearing‖ to trigger the formal rulemaking provisions of APA) d. Rulemaking Plus and Rulemaking Minus i. Courts cannot impose additional procedural safeguards not required by APA or organic statute. 1. Vermont Yankee Nuclear Power Corp v Natural Resources Defense Council, USSC, 1978. In deciding how to handle spent nuclear power plant fuel, the Atomic Energy Commission decided not to deal w/ issue thru adjudication, but thru notice-and-comment rulemaking. Relying on expert testimony that environmental hazards were negligible, AEC issued zero release rule giving numerical values for various hazards posed by plant construction. It then licensed construction of Vermont Yankee w/out giving environmental groups any say. DC Cir reversed, saying that rulemaking proceeding was defective for failing to expose and permit adequate adversary probing of waste disposal issues. Held: so long as agency complies w/ statutory minimum in organic statute/APA, courts can‘t impose additional procedural requirements. DC Cir reversed. 2. This case restricted hybrid rulemaking, in which, to police problems of informal rulemaking to which agencies increasingly rely, courts imposed additional procedural requirements above and beyond APA from c/l background of APA. 3. Nevertheless, the courts still impose procedural requirements implicit in hard look review (Overton Park). But there, source of authority is APA, not c/l. ii. Rules Excepted from § 553 (either formal or informal rulemaking) 1. Rules involving military or foreign affairs (§ 553(a)) a. Rather strictly interpreted limitation. Only when the purpose of the exception—not to impede military operations or interfere w/ conduct of foreign relations by subjecting decisions to publicity of informal rulemaking—is served will exception be recognized. 2. Rules for matters involving agency management or personnel, public property, loans, grants, benefits, and contracts. (§ 553(a)) a. This was compromise in APA to shield rulemaking that didn‘t involve regulating private behavior from § 553. b. This becomes subject of criticism, so agencies (e.g., HUD, HHS, Transportation, Interior, Agriculture, Labor) began waiving this exception. i. This waiver is binding until regulation is repealed thru rulemaking. Batterton v Marshall. c. Congress began piecemeal revoking this exemption from organic statutes. iii. Rules Excepted from § 553‘s Notice-and-Comment Requirements (§ 553(b)-(c)) 1. American Hospital Association v Bowen, DC Cir, 1987. Congress passed statute creating Peer Review Orgs to oversee Medicare expenditures. PROs would contract to conduct the review. If hospital failed, it wouldn‘t get reimbursed. Many regs about PROs and Dep‘t of HHS placed standard terms in PRO Ks. HHS didn‘t follow notice and comment. Held: These regs were procedural, and standard terms were merely set goals rather than imposing standards and hence were statements of policy or procedural rules. a. Purpose of exceptions is to accommodate situations where the policies promoted by public participation are outweighed by countervailing considerations of effectiveness, efficiency, expedition, and reduction in expenses. 2. The exceptions: a. interpretive rules i. Exception for interpretive rules is to allow agencies to explain ambiguous terms w/out having to undergo cumbersome proceedings ii. Substantive rules grant rights/impose obligations/effect private interests/change existing law or policy. Interpretive rules merely clarify. iii. Test: (1) statement of policy may not have present effect; (2) does purported policy statement genuinely leave agency free to exercise discretion? 1. Community Nutrition Institute v Young, DC Cir, 1987. FDA can condemn food as adulterated. FDA established thru reg adopted w/out notice/comment action levels informing food producers of max amount of unavoidable contaminants that it would permit. Held: action levels aren‘t statements of policy b/c they have present effect and are binding. 2. Other things to consider: a. Whether in absence of the rule there would not be an adequate legislative basis for enforcement action or other agency actions to confer benefits or ensure the performance of duties i. Does rule itself provide enforcement action to ensure performance of duties ii. i.e., take away the rule: can agency still prosecute person for violating reg? if so, rule is interpretive b. Whether the agency has published in the Code of Federal Regulations i. Federal Register Act describes the docs required to be published in the CFR as those agency docs having ―general applicability and legal effect.‖ ii. Dubious: while Act describes what must be published, it doesn‘t prohibit other docs from being published there; there is positive value in having agency‘s formal statements of its interpretations published where the public may more easily find them. c. Whether the agency has explicitly invoked its general legislative authority d. Whether the rule effectively amends a prior legislative rule i. make sure that the agency didn‘t make a legislative rule w/ a catch-all provision that gives itself the power to avoid later notice-and-comment rulemaking; i.e., is the ―law‖ the agency claims to be interpreting too broad to support the specific ―interpretation‖ the agency gave to it? ii. National Family Planning & Reproductive Health Ass‘n v Sullivan, DC Cir, 1992: If you amend a notice/comment rule, you must provide notice/comment for the amendment. iii. Shalala v Guersey Memorial Hospital, USSC, 1995: no notice/comment required if new interpretation isn‘t inconsistent w/ old view. b. general statements of policy i. Exception for general policy statements is to allow agencies to announce tentative intentions w/out binding themselves. The more a policy statement looks like it will be used to decide future cases (i.e., creates a binding legal norm), the more likely it‘ll be called a legislative rule. Hence, look for evidence that policy statement is tentative. See Community Nutrition Institute v Young, supra. ii. However, agencies can‘t promulgate policy statements to coerce regulated entities into actions the agencies couldn‘t mandate thru legislative rules. 1. Chamber of Commerce v US Dept of Labor, DC Cir, 1999. OSHA issued policy statement that it wouldn‘t inspect workplaces as often/thoroughly if employer adopted a workplace safety plan that exceeded fed‘l requirements in certain ways. OSHA couldn‘t have mandated this, and b/c of the costs/burdens of OSHA inspections, the offer to reduce those inspections exercised substantial coercive power on employers to adopt these workplace safety plans. Held: this statement of policy is invalid and beyond agency‘s power. c. rules of agency organization, procedure, and practice i. Exception for rules of agency org/procedure/practice is to ensure that agencies retain latitude for internal operations. ii. Air Transportation Ass‘n v Dept of Transportation, DC Cir, 1990. Congress amended FAA‘s organic statute to authorize FAA to establish a program of administrative penalties for violating act‘s requirements. FAA issued its Penalty Rules w/out notice/comment. Held: though nominally procedural, the Penalty Rules encode a substantial value judgment in that they alter rights of Ds and make contentious choices about how much d/p to give them. 1. Whether a rule is procedural/substantive depends on functional analysis. Where nominally procedural rules encode a substantial value judgment or substantially alter the rights/interests of regulated parties, the rules must be preceded by notice/comment. d. agency findings for good cause that notice and public procedure are impracticable, unnecessary, or contrary to public interest i. Substantive and procedural requirements. 1. Substantively, agency must find good cause not to provide notice/comment a. Mainly, this is for cases where announcing that agency is considering a particular rule would interfere w/ accomplishing the rule‘s goals; e.g., a price freeze (if announced, price would skyrocket as people try to beat the freeze). 2. Procedurally, agency must put this finding and reasons for it into rule when adopted. In practice, when agency publishes rule in Federal Register, it includes the finding of good cause and reasons why in the preamble. If agency doesn‘t comply, then rule won‘t qualify for this exception. e. Due Process and Hearing Rights: Protected Interests i. Magna Carta View of D/P 1. D/P means only that Gov‘t can‘t deprive anyone of life, liberty, property unless under the law of the land 2. Binds the executive to follow the decisions of the legislature a. Couldn‘t have a deprivation w/o statutory authorization (c/l: ―old‖ property; statutory: ―new‖ property) b. But if executive had other source of authority then he could take w/out any procedure ii. Property interest: category test—claim (1) comes from positive law (2) when gov‘t cabins discretion. 1. Goldberg v Kelly, USSC, 1970. AFDC procedures allowed a termination of benefits after 3 officials concurred but with no pre-termination (only post-) hearing. Held: lack of pretermination hearing violates D/P. People on welfare have too much at stake for us to risk an erroneous decision on a no-hearing procedure a. The triggering of due process protections seems to be the fact that the recipients are entitled by statute to receive the benefits if they meet statutory criteria of eligibility. The ―brutal need‖ of the plaintiffs is only relevant in the balancing test of deciding how much process is due. Moreover, the fact that the state conceded the applicability of D/P sort of means that USSC didn‘t need to decide the question. 2. Board of Regents v Roth, USSC, 1972. Roth is prof w/ 1 yr K that isn‘t renewed w/out any hearing. Held: no D/P violation; reemployment isn‘t a liberty interest b/c reason for not being rehired didn‘t cause stigma/etc; isn‘t a property b/c no statutory right to reemployment. 3. Perry v Sindermann, USSC, 1972. Sindermann is prof under series of 1 yr Ks that was informal tenure system, then was not renewed w/out any hearing. Held: D/P maybe implicated; a person‘s interest in a benefit is a property interest if there are such rules or mutually explicit understandings that support his claim of entitlement to the benefit. Case remanded to see if state law would interpret Sindermann as having legal claim for continued employment, which in turn would be legitimate claim of entitlement to protected property interest. 4. Arnett v Kennedy, USSC, 1974. Gov‘t employee fired for cause w/out pretermination hearing; civil service statute permits this, leaving whether there is to be a pretermination hearing up to supervisor‘s discretion. a. Rehnquist plurality: procedure and substance should come from state law (and thus D/P not implicated): take the bitter with the sweet—no liberty interest implicated where statutory expectancy specifies the proceedings to be used. This view rejected by 6 other justices. b. Powell/Blackmun concurrence: procedure comes from fed‘l constitutional law. And while D/P is implicated, after balancing, no need for more protection. iii. The less discretion is cabined, the less likely you‘ll get constitutionally mandated process. But we might think that the more discretion officials have, the more process we‘d want to prevent arbitrariness of decisionmaking. (but if it‘s all discretion, what would you even have a hearing about?) iv. If there is a property entitlement created by state law, the process that is due is determined by federal constitutional law, not by state law (consequence of Rehnquist‘s failure to get majority for Arnett. Arnett said that substance and process ought to be creatures of state law; court rejected that position.) What‘s wrong w/ the Arnett plurality: if substance is to be a creature of state law, why should process be determined by federal constitutional law? 1. Sources of authority of judicial imposition of procedural requirements a. Legislative Candor. -- The legislature may be evading accountability for its actions when it purports to grant statutory entitlements but simultaneously specifies enforcement procedures that substantially but surreptitiously diminish the protection afforded. Democratic processes could arguably be strengthened if courts were to insist on procedures that forced Congress to deliver on its substantive promises i. This argument, however, ignores the political necessity and virtue of compromise. Limited procedural protections may have been necessary to obtain sufficient legislative support for the substantive entitlement in the first place. More fundamentally, the most basic separation-of- powers considerations prohibit courts from seizing on the due process clause as a basis for imposing their own notions of political responsibility on the process by which elected officials make laws. If Congress wishes to enact legislation that is to some degree hortatory, it is unclear from what source the courts obtain authority to provide otherwise b. Honoring Expectations. -- A variant of the ―candor‖ argument emphasizes that procedural qualifications of substantive entitlements disappoint reasonable expectations derived from the public's understanding of legislation. Under this theory, citizens are aware of substantive statutory entitlements, but are ignorant of the procedural riders that compromise those entitlements. In its traditional role as guardian of reasonable expectations, the judiciary may demand procedures that will enable citizens to plan their activities with an enhanced sense of security. i. Do citizens really know that statutes contain formal entitlements, and is it likely that those who do know are ignorant of the effect of procedural riders? Even if statutory procedures might result in surprise or disappointed expectations, what is the constitutional basis for striking such procedures down if the legislature believes that they serve important policies, such as reduction of cost and delay? Moreover, arguments based on disappointed expectations must take account of the legislature's recognized power to disappoint expectations through amendment or repeal of statutes that confer new-property entitlements c. Nondelegation Considerations. -- Under the nondelegation doctrine, the legislature is required to confine the authority of administrative officials through relatively precise standards or directives. This requirement serves three principal purposes. First, it promotes political accountability by ensuring that policies are made by elected representatives rather than administrators. Second, it promotes predictability for individuals benefited or burdened by regulatory programs. Third, it protects against arbitrary or discriminatory action by minimizing the discretion of administrators. By serving the latter two functions, the nondelegation doctrine simultaneously promotes due process and equal protection goals. In short, although courts may lack the authority and competence to require that Congress legislate with specificity, when Congress does so of its own volition courts may impose procedural safeguards to ensure that Congress‘s specific standards are carried out evenhandedly. In this respect, considerations of equal protection and due process also argue in favor of judicial creation of procedural safeguards for statutory entitlements when the formal criteria are met, at least when the legislature has not clearly ruled out such procedures. 2. Vermuele: Renquist‘s Arnett opinion is right. Roughly until WWI, D/P clause had a well accepted understanding. The idea was that it was a separation of powers provision; the emphasis was on law rather than process. You could only be deprived of liberty/property through legislatively provided procedures. It restricted ability of executive/judiciary from taking away property/liberty; you need statutory authorization to deprive of life/liberty/property. Moving away from this understanding hasn‘t been entirely successful. The current approach might be intellectually incoherent. Plus, it has very limited operational value; we haven‘t been able to cash it out in successful ways. The virtue of the old understanding of due process is that it provided a clarity of legal obligation that itself has pragmatic benefit and that is notably lacking in modern caselaw. f. Due Process and Hearing Rights: What Process is Due? i. Rulemaking (in constitutional sense)—legislative process is enough (Bi Metallic) ii. Adjudication 1. Mathews v Eldridge. Look at due process on the margin— must a particular procedure be added? Add the procedure if and only if the cost to the government is less than the increased risk of erroneous deprivation if the procedure is denied (conversely, the increased accuracy w/ the procedure) multiplied by the value to the claimant. 2. Possible bits of procedure that a tribunal might use (from Judge Friendly‘s article): a. An unbiased tribunal, Notice of the proposed action and the grounds asserted for it, Opportunity to present reasons why the action shouldn‘t be taken, Right to call witnesses, Right to know the evidence against you, Right to have decision based only on the evidence presented (―closed record‖), Right to counsel, Making of a record at all (―open record‖), Statement of reasons for the decision, Right of the public to attend the hearing, Judicial review. g. Estoppel i. Equitable estoppel: When a person affirmatively misleads another who reasonably relies on the misleading info to his detriment, a court won‘t consider evidence contrary to the info that first person provided. 1. Office of Personnel Management v Richmond, USSC, 1990. 1st policy: you lose disability if you make >80% of old salary in each of two yrs. 2 nd policy: only one year window. P told (twice, over five years) and given official form w/ the old rule. He relied on this and worked too much—lost disability for 6 mths. Held: no estoppel in this context b/c would violate Appropriations Clause. a. i.e., equitable estoppel doesn‘t work against agencies. b. To avoid saying that there‘s estoppel, court sometimes relies on estoppel-like arguments, but using other doctrines: i. U.S. v Winstar, 1996: USSC held that agency had entered into enforceable contracts (they had promised that S&Ls could do something, then made that thing unlawful in a rule). ii. Socop-Gonzales v INS, 2000: P relied on INS officer‘s mistaken advice. Not estoppel; instead, INS had to apply its exceptional circumstances doctrine to provide relief ii. Collateral estoppel: Once a court has decided an issue of fact or law necessary to its judgment, that decision is conclusive in a subsequent suit based on a different c/o/a involving a party to the prior litigation. 1. US v Mendoza, USSC, 1984. P challenged denial of naturalization, citing collateral estoppel on legal issue the INS had lost in a similar case. Held: a non-party to a prior judgment could not assert collateral estoppel against the government. a. Application of nonmutual offensive collateral estoppel would thwart development of important questions of law by freezing the first final decision rendered on a particular legal issue, would frustrate judicial interests, including the court's own policy of awaiting conflicts among the circuit courts before granting certiorari, and would impair the government's discretion in deciding which cases to appeal. h. Rule Following and Retroactivity i. Adjudication based policy and changed by adjudication 1. This is quite common. Generally, agencies must give a non-arbitrary reason for changing adjudicatory regime— but then again, so do courts (that‘s the force of stare decisis). If the second order also applies retroactively, there will be a problem under Majestic Weaving (which is an exemplar of the whole line of squib cases). Majestic says agencies may use retroactive adjudication if it passes balancing test a. Majestic Weaving, 2d Cir, 1966. D negotiated w/ union before union had majority support—an ―unfair labor practice,‖ sez NLRB, overruling prior longstanding decision approving of it. Held: NLRB action impermissibly retroactive b/c didn‘t give notice of rule change. Dictum: we don‘t like calling something ―fair‖ then retroactively calling it ―unfair.‖ We‘d rather see you extend jurisdiction (Atkinson) or increase penalty for something already prohibited. b. Things to consider in balancing: i. Primary or secondary reliance interest being infringed? 1. NLRB v Atkinson, 9th Cir, 1952. NLRB declared D guilty of unfair labor practice for discharging an employee for not paying union dues (i.e., NLRB concluded that closed shop agreement was invalid). 2 changes in policy after the firing but before order brought: (1) that when employment swells w/out new union election/evidence of majority support, no closed shop agreement and no union recognition; (2) NLRB extended jurisdiction to this industry. Held: NLRB‘s order is invalid as retroactive. The hardship of applying rule retroactively outweighs benefit to be gained. D couldn‘t have known it was guilty of any conduct that would bring NLRB down on its head 2. Court distinguishes between primary and secondary reliance. E.g., here, didn‘t know that primary conduct was going to be made impermissible, so no retroactivity allowed. 3. A change in remedies (fine goes from X to 10X) or enforcement policy (enforcement increases from 40% to 50%) is only secondary reliance. Courts reason that you always knew your conduct was illegal, so no change in first order reliance interest. (see, e.g., Landsgraf). a. of course, this is all hogwash under modern law and economics, but distinction persists in admin law. ii. Huge administrative difficulties if policy change not made retroactive? 1. Leedom v IBEW, DC Cir, 1960: NLRB adopted, then shortened, period during which representation proceedings couldn‘t be instituted by a union not a party to the initial bargaining agreement. NLRB‘s action sustained by court, despite being retroactive. Court held that balancing of hardship/benefit was in favor of NLRB. Rule made retroactive not just to be comprehensive, but also to prevent administrative nightmare where some companies are subject to new rule and some companies subject to old rule. You don‘t want to have different ―Classes‖ of companies in this manner. ii. Adjudication based policy changed by rulemaking 1. Bowen v Georgetown University Hospital, USSC, 1988. HHS sets limits on amount of money that medicare w/ reimburse hospitals. In 1981, it promulgated rule saying it wouldn‘t look at wages when calculating indices for calculating reimbursement limits (hospitals/doctors didn‘t like this rule and so challenged it). 1981 rule invalidated on procedural grounds; Congress tinkers a bit w/ substantive portions of statute. In 1984, HHS changed rule: it would now look at wages, and would do so solely retroactively (b/c prospective portions are governed by the new statutory provisions). Held: organic statute doesn‘t permit this kind of retroactive rulemaking. a. Presumption against allowing retroactivity that can only be overcome by text of statute b. Scalia concurrence: APA definition of ―rule‖ forbids retroactive rulemaking (§ 551(4)). 2. Smiley v Citibank—limits Bowen: Bowen only applies where new retroactive rule replaces a rule that previously governed the same conduct. Bowen doesn‘t apply if retroactive rule replaces a confusing situation where there was no clear rule. iii. Rulemaking based policy changed by rulemaking—also follows Bowen 1. If new rule applies solely prospectively, no legal problems. Though there might be chevron/hard look challenges, but nothing else. But now assume it has a retroactive component (maybe it applies prospectively and retrospectively; or it only applies retroactively), then we have a problem under Bowen, which limits agencies‘ ability to promulgate retroactive rules. iv. Rulemaking based policy changed by adjudication 1. Arizona Grocery tells us that this sequence is illegal. It doesn‘t matter whether order purports to be prospective, etc. No matter what, a rule can‘t be changed by an adjudication under Arizona Grocery (which is about the binding effect of agency rules). 2. Arizona Grocery, USSC, 1932. Held: can‘t do this. a. Is this requirement consistent w/ Vermont Yankee, which held that courts cannot impose on agencies procedural requirements not in positive law. Possible explanations: i. APA‘s definition of rule requires agencies to comply w/ rules while they‘re on the books 1. you could say that it was a common law, and the drafters of the APA drafted the law against the backdrop of that common law, so it‘s implicitly in there. But it‘s very hard to find this principle in the language of the APA. ii. Sometimes agencies that violate their own rules violate D/P iii. It is arbitrary/capricious for an agency not to follow own rules b. Vermeule wants to replace Arizona Grocery w/ 2 questions: (1) does organic statute permit orders like this, and (2) is there anything arbitrary/capricious – but these are requirements of any agency action. c. This principle only applies when challenging agency action; can‘t invoke it as an exclusionary rule, for example. i. US v Caceres, USSC, 1979. Holding: D in a criminal prosecution may not exclude evidence obtained in violation of IRS regs requiring DOJ approval before wiretapping meetings between taxpayers and IRS agents was undertaken i. Agency Structure i. Framework : 3 Structural Grounds For Objection to Agency decision making 1. Insufficient Separation of functions (A hangover from Humphrey’s Exec.) a. ALJ Level b. Agency Head Level 2. Excessive Separation of function Morgan quartet cases: Agency has pulled apart decisionmaking entities that should be together 3. Bias — Combination of prosec. and adjud. creates bias a. pecuniary interests b. prejudgment ii. Insufficient Separation of the ALJ 1. §554(d): ―The employee who presides at the reception of evidence pursuant to § 556 (formal adjudication) shall make the recommended decision or initial decision required by § 557 (formal adjudication), unless he becomes unavailable to the agency. Except to the extent required for the disposition of ex parte matters as authorized by law, such an employee may not: a. Do any Ex parte communication unless everyone notified and given opportunity to participate b. Be responsible to or subject to the supervision or direction of an employee or agent engaged in the performance of investigative or prosecuting functions for an agency. c. An employee engaged in performance of investigative/prosecuting function may not, in that or a factually related case, participate or advise in the decision, recommended decision, or agency review, except as a witness or counsel in public proceedings. iii. Excessive Separation of functions 1. Morgan v US I. SecAg must hear adjudication by statute. He tells deputy to hear adjudication and report back. Deputy hears it, tells SecAg and SecAg signs the order. a. If deputy had made the decision himself this would be OK (presumption that subdelegation is permissible). b. Held: one who hears must decide. 2. Morgan v US IV: But: cannot probe the mental processes of the decisionmaker a. Though this seems inconsistent w/ Overton Park, remember that Overton Park was informal adjudication whereas Morgan was formal adjudication. iv. Bias 1. Pecuniary interest—it‘s a matter of degree a. Easy cases: i. Tumey v Ohio: judge‘s salary paid from fines he personally levied on persons convicted in his court ii. Ward v Village of Monroevill: mayor of village presided over trials, the fines from which were half of village‘s revenue b. Harder cases: i. Gibson v Berryhill, USSC, 1973. State board of optometry was only independent doctors, but they licensed mass mkt. doctors as well. Board brings action against big company; Company claims bias, since independent doctors want to drive the mass mkt. docs out of business. Held: Biased. 1. trade off between expertise and impartiality—the more expertise you get, the less impartial your decisionmaker. Maybe here, the importance of expertise isn‘t all that relevant—maybe mass mkt optometry is different than independent optometry 2. prejudgment a. Withrow v Larkin, USSC, 1975. Doctors on state board both conduct investigation and adjudicate hearings. Held: the combination of investigative and adjudicative functions doesn‘t by itself create unconstitutional risk of bias. i. Consistent w/ Gibson: In Gibson, the dist ct thought we had a case of agency capture; the board was captured by the interest of independent optometrists who had a direct conflict w/ corporate optometrists. In Withrow, by contrast, the examining board wasn‘t systematically biased against any particular subclass; the marginal gain to the regulators from putting a person out of business is small (much smaller than Gibson). b. Generally, you only need recusal when you‘ve prejudged this particular adjudication or have preknowledge of adjudicative facts. Prejudgment of policy issues traditionally thought ok. j. Publicity and Disclosure i. Freedom of Info Act 1. Partially codified in APA § 552. Requires fed‘l agencies to make ―promptly‖ available to ―any person‖ any written info in agencies‘ possession unless info w/in one of nine exceptions from compelled disclosure (disclosure for exempt materials is optional, but 3d parties might object to disclosure of exempt info). No standing requirement for requester. Any ―record‖ (which includes e-mails and electronic data as well as documents) must be disclosed. If denied, you get de novo review in fed‘l dist ct, and ―the burden is on the agency to sustain its action.‖ Sanction for noncompliance is contempt. Agency must answer FOIA complaint w/in 30 days. a. Kissinger v Reporters‘ Committee for Freedom of the Press, USSC, 1980: FOIA doesn‘t require agencies to obtain and disclose records no longer in their possession. b. Rushforth v Council of Economic Advisors, DC Cir, 1985: Prez and his advisors aren‘t an ―agency‖ for purposes of FOIA. But note that definition for ―agency‖ is broader in FOIA (§ 552(f)) than for general APA purposes (§ 551(1)). c. In considering claimed exempt status for info, the specific interests of the requester in disclosure shouldn‘t be considered; the ―only relevant public interest in the FOIA balancing test is the extent to which disclosure of the information sought would ‗shed light on the gov‘t‘s performance of its statutory duties‘ or otherwise let citizens know ‗what their government is up to.‘‖ Dept of Defense v FLRA, USSC, 1994 (holding that union‘s interest in promoting collective bargaining shouldn‘t be considered in request for gov‘t workers‘ addresses; disclosure denied). 2. Exemptions from disclosure: a. National Security Exemption: in theory, this provision, together w/ provision for in camera judicial review of withheld records, would permit de novo court review of nat‘l security classifications by the executive branch, but in practice, cts are reluctant to inspect docs in camera or to 2 nd -guess prez. b. Internal Personnel Rules and Practices Exemption: § 552(b)(2) exempts docts that are ―related solely to the internal personnel rules and practices of an agency.‖ i. Two primary purposes unlerlie exemption: (1) protection of employee privacy and (2) protection of agencies from harassment. ii. Exemption curtailed in Dept of the Air Force v Rose, USSC, 1976: student editors of NYU Law Review doing research on student discipline in military academies could obtain disclosure of case summaries of air force academy discipline proceedings w/ personal references and other identifying info deleted. The case summaries, edited to preserve anonymity, didn‘t fall w/in exemption (and also fell outside the privacy exemption, infra). iii. Enforcement manuals are w/in exemption, even though people argue that these manuals resemble the ―secret law‖ the FOIA was designed to open to public scrutiny. c. Exemption for Documents Governed by Statutes that Specifically Direct Nondisclosure d. Confidential Business Information Exemption: § 552(b)(4) provides that ―trade secrets or commercial and financial information obtained from a a person and privileged or confidential need not be disclosed to the public.‖ i. Purpose of FOIA is to ensure citizen scrutiny of fed‘l gov‘t, not open private businesses to public scrutiny ii. National Parks & Conservation Assn v Morton, DC Cir, 1974, held that info is ―confidential‖ if its disclosure would be likely to ―cause substantial harm to the competitive position‖ of the person who submitted it, or if disclosure is likely to ―impair the Government‘s ability to obtain necessary information in the future.‖ Others argue for a broader exemption, including all info which wouldn‘t be released to the public by the person from whom it was obtained. e. Exemptions for Privileged Agency Materials: § 552(b)(5) provides that agencies may withhold ―inter-agency or intra-agency memorandums or letters which would not be available by law to a party other than an agency in litigation with the agency.‖ i. Consider atty-client and work-product privilege. Also includes a ―deliberative process‖ privilege, unique to gov‘t, to shield from disclosure records pertaining to or reflecting the process of internal gov‘t decisionmaking—to shield frank discussions w/in gov‘t. ii. Agency communications w/in agency, w/ other agencies, and w/ neutral consultants are protected. But if agency communicates w/ consultant who is really advocating self- interestedly a position, then no exemption. Dept of Interior v Kalamath Water Users Protective Assn, USSC, 2001 (communications between Bureau of Indian Affairs and several tribes regarding the position the BIA should take on behalf of the tribes in the state proceedings concerning water allocation rights). f. Personal Privacy Exemption: § 552(b)(6) protects ―personnel and medical filess and similar files the disclosure of which would constitute a clearly unwarranted invasion of personal privacy.‖ g. Investigatory Records Exemption: § 552(b)(7) permits records to be withheld if disclosure would (1) interfere w/ enforcement proceedings, (2) deprive a person of a fair trial or an impartial adjudication, (3) create an unwarranted invasion of personal privacy, or disclose the identity of a confidential source, or (4) disclose investigative techniques and procedures or endanger the life or safety of law enforcement personnel. i. Procedural threshold to claim exemption: agencies must establish w/ specific evidence that the source spoke w/ an understanding that the communication would remain confidential h. Financial Institution Exemption: § 552(b)(8) permits nondisclosure of reports prepared by fed‘l agencies about operations of banks and financial institutions. Purpose is to ensure the stability of financial institutions (but this may be superfluous b/c of exemption 4) i. Geological Exploration Exemption: § 552(b)(9) exempts ―geological and geophysical information and data, including maps concerning wells.‖ Also probably superfluous b/c of exemption 4. 3. FOIA Procedures and Policies a. FOIA states that courts should review disclosure issues de novo (in departure from APA § 706) b. FLRA v US Dept of Defense, 10 th Cir, 1993: no Chevron deference w/ regards to FOIA/Privacy Act. 4. Reverse FOIA litigation a. 3d parties might try to prevent agency disclosure of info under FOIA. Your right to prevent disclosure where disclosure prohibited by a statute other than the FOIA is the APA (even if other statute doesn‘t give private right of action). b. Practical problem w/ these cases is notice. If agency doesn‘t notify person that someone is requesting info submitted by him, he can‘t really sue, now can he… ii. The Government in the Sunshine Act 1. Open-Meetings provisions: requires that every part of every meeting must be open to public unless w/in one of 10 specific exemptions. a. ―Meeting‖: includes deliberations of at least a quorum of members where the deliberations determine or result in the conduct of agency business (some degree of formality required) b. No meeting may be closed unless majority of membership votes to close it. Must do so week in advance and notify the public. Full transcript taken of closed meeting, and any portion of meeting that doesn‘t meet requirement will be disclosed. 2. Some exemptions parallel FOIA exemptions: meetings (1) vital to national defense or foreign policy; (2) concerned w/ internal personnel rules and practices of the agency; (3) specifically exempted from disclosure by another statute; (4) concern trade secrets and confidential commercial or financial info; (5) would constitute a clearly unwarranted invasion of personal privacy; (6) would disclose investigatory records compiled for law enforcement purposes under certain limited circumstances; (7) relate to bank or financial institution examination reports. Three new exemptions: meetings (8) involving accusing any person of a crime, or formally censuring any person; (9) would frustrate implementation of a proposed agency action if prematurely known; (10) concern the agency‘s participation in formal rulemaking or litigation. V. Access to Courts a. Reviewability i. Modern Presumption in favor of reviewability: 1. Abbott Laboratories v Gardner, USSC, 1967: presumption in favor of reviewability. APA gives judicial review to one ―suffering legal wrong because of agency action, or adversely affected or aggrieved by agency action within the meaning of a relevant statute.‖§ 702. Unless other statute precludes relief and the action isn‘t committed to agency discretion, then you get review. § 701(a). You get review not only of ―agency action made reviewable by statute,‖ but also review of ―final agency action for which there is no other adequate remedy in court.‖ § 704. Case held that the failure of FDCA to provide for judicial review of regulations didn‘t preclude review, even though the act did specifically provide review of other types of reg. 2. why have presumption? a. Crowell: Administrative adjudication is only OK b/c we know there is eventual Art III review i. Must have review to be true to vesting of judicial power in courts ii. Art III review is other ½ of Constitutional bargain that allows ALJ‘s iii. BUT, proves too much b/c then Cong. could never prevent review b. Principal—Agent Model i. Chevron showed deference justified as a global embodiment of Cong. Intent ii. Same idea here where we need a default rule and Cong. wants review ii. Two exceptions: APA‘s review provisions don‘t apply ―to the extent that (1) statutes preclude judicial review, or (2) agency action is committed to agency discretion by law.‖ § 701. iii. Statutory preclusion 1. Statutory preclusion can be explicit or implicit, but you start w/ presumption of reviewability. Legislative silence about availability of review normally not construed to show an intent to preclude review. Constitutional challenges always reviewable. 2. Block v Community Nutrition, USSC, 1984. By statute, Sec‘y of Agr. sets min prices milk handlers must pay to milk producers for milk. Milk for drinking gets highest price, milk for butter/chz/powder in lower class. P, a nonprofit/consumers, and a milk handler challenged Sec‘y‘s decision that higher price applied to reconstituted milk not used to manufacture butter/chz. Held: AMA Act precludes judicial review by consumers of Sec‘y determination that reconstituted milk for drinking gets higher price, b/c presumption in favor of reviewability by consumers overcome by b/c clear and convincing evidence of preclusion may be found in fairly discernible inferences drawn from statutory scheme as a whole a. Preclusion issue turns on whether Congress intended for class to be relied upon to challenge agency disregard of law. b. Here, since consumers shut out of initial administrative scrutiny, no basis for attributing to Congress the intent to have consumers challenge orders judicially. c. Plus, consumer suits would disrupt administrative system: easy device to evade administrative exhaustion requirement 3. Bowen v Mich Academy of Family Physicians, USSC, 1986. HHS rule cut reimbursement for non-bd certified docs. The Act provides for review by patients but not doctors. Held: review permitted; no expressio unius. a. Hard to reconcile w/ Block. Best way to handle this is to say that Block was really about standing, not reviewability. 4. Johnson v Robinson, USSC, 1974. Veteran denied benefits and wants review. VA statute expressly forbids review of agency decisions. Held: the no-review clause doesn‘t bar consideration of the constitutionality of the congressional decision not to provide benefits to conscientious objectors. iv. Committed to agency discretion by law 1. APA § 706 provides that a reviewing court ―shall … hold unlawful and set aside agency action … found to be … an abuse of discretion.‖ But § 701(a)(2) states that this doesn‘t apply to ―agency action ... committed to agency discretion by law.‖ 2. No Law to Apply: In Overton Park, USSC explains that § 701(a)(2) exception for action ―committed to agency discretion‖ is a ―very narrow exception.‖ The legislative history of APA indicates that it‘s applicable where ―statutes are drawn in such terms that in a given case there is no law to apply‖; i.e., no legal standard against which unlawfulness can be judged. In Overton Park, they found a legal standard in the no ―prudent or feasible alternative‖ language in the statute. But sometimes a standard will be so open-ended that any exercise of discretion qualifies as legitimate. a. This exception differs from statutory preclusion exception: statutory preclusion refers to congressional foreclosure; § 701(a)(2) refers to an absence of judicially cognizable standards. 3. Heckler v Chaney, USSC, 1985. Death row inmates challenge FDA‘s decision not to bring enforcement action against users of lethal injection chemicals for not making sure chemicals are ―safe‖ for execution. Held: With regard to refusing to undertake enforcement action, there is no presumption of reviewability. This is something traditionally committed to agency‘s absolute discretion. a. Lower courts have developed exceptions to this doctrine, deciding that Chaney is limited to enforcement decisions in particular cases: i. Chaney doesn‘t apply to guidelines/policies ii. To rulemaking (American Horse Protection Assn v Lyng) iii. To jurisdictional refusals iv. To ―abdication‖: where an agency has systematically abdicated enforcement in a relevant area. 4. Webster v Doe, USSC, 1988. CIA fires doe for being gay b/c Director of CIA deems it in the interest of the US. Held: not reviewable; organic statute‘s text and purpose confirm that firing decisions committed to Director‘s discretion. a. Scalia concurrence: i. Wants to use a multifactor test including judge-made APA law: § 701(a)(2) applies whenever the common law of judicial review of agency action says not to review something. This is a functional/historical test. 1. Chances for error 2. Judicial competence in particular area 3. Political question concerns b. Problems w/ ―no law to apply‖: i. To the extent that there‘s no law to apply— nothing to guide agency‘s discretion—then we think that maybe we‘ve got a nondelegation problem. ii. Alternatively, doesn‘t statute in Webster prohibit the director from openly and notoriously operating under a different standard? If that‘s true, then we‘re just talking evidence. b. Standing i. Data Processing, USSC, 1970. Comptroller ruled that banks may make data processing services available to other banks and to bank customers. P is a competitor. Held: P has standing; the organic statute‘s purpose is, in part, to protect competitors, so a competitor who is in fact injured can assert standing. APA grants standing to a person ―aggrieved by agency action w/in the meaning of a relevant statute.‖ This includes aesthetic, conservational, and recreational as well as economic values. ii. Constitutional Requirements (from ―case or controversy‖ requirement) 1. Injury in fact a. Concrete and particularized b. Actual or imminent rather than conjectural or hypothetical—but can injury be rephrased as opportunity injury? i. Lujan v Defenders of Wildlife, USSC, 1992. End Species Act vests enforcement w/ various secys who decided it applies to action taken in foreign nations, then limited its geographic scope to the US. Wildlife groups and individual Ps (zookeepers, etc.) sued, saying that revised group misinterpreted statute, permitting funding abroad of projects that threaten endangered species. § 7(a)(2) of ESA says that fed‘l agencies cannot endanger habitat of endangered species etc. Held: no standing; none of the Ps have shown injury in fact; congress can‘t confer standing on them w/out violating art III. ii. Injury: mere intent to return isn‘t enough. 1. However, we can rephrase the claim in the following way from an outcome injury to an opportunity injury (as in Northeastern Florida Chapter of Assoc General Contractors v Jacksonville: challenge to minority set aside, where injury was lack of opportunity to bid on equal footing): it‘s not that Ps have to buy a plane ticket, but the opportunity to see crocodiles is an independently valuable thing, and but for the illegal regulation, the chances of having an opportunity to see crocodiles will be greater. Lujan Ps don‘t want to see crocodiles, they just want the opportunity to see crocodiles. a. Difference between Lujan and Jacksonville: In Jacksonville, the contractors at least had bid on projects; in Lujan, didn‘t even have tickets to go see crocs. 2. Scalia tries grounding restrictions in Art II: Scalia remarks that there‘s a SoP problem: courts should only discuss rights of individuals who have suffered concrete harms, b/c where Congress gives courts power to decide rights of uninjured people, then Congress is usurping power of Prez, transferring power to Take Care from Prez to courts (acting thru Ps). a. But the Take Care clause only imposes upon Prez a duty to comply w/ statutes (rather than just shielding executive action from judicial review). When he violates a statute, he violates Take Care Clause b. Thus, the Art II argument does no work whatsoever, and nothing‘s left but Scalia‘s Art III args, which are just wrong (see supra discussion of opportunity injuries). c. This test is more about identifying what kind of Ps we want rather than what kinds of injuries are sufficient. Want to discourage ideological Ps. Unclear why. i. Only Ps w/ concrete/personal stake will litigate a case w/ sufficient adversary rigor; otherwise, risk of collusive/lethargic litigation ii. Limitations on standing serve democratic goals—ensures courts don‘t decide policy rather than political branches iii. People whose concrete interests at stake have standing; others are officious intermeddlers d. Recreational, Aesthetic, or Environmental Injury i. Mere interest in case isn‘t enough, but if person uses area and gov‘t action would harm it, then is injury. e. Risk as Injury i. To subject a person to immediate increased risk of harm is itself injury even if it is conjectural whether the harm would ever occur. Louisiana Environmental Action Network v EPA. f. Procedural Injury i. Lujan: suggests that procedural violation by itself could never satisfy injury requirement—the violation is a generalized grievance (assuming that the procedural injury isn‘t one specifically felt by a finite group of persons only, as opposed to one that affects everyone equally) 1. e.g., procedural violation in Lujan— failure to conduct interagency consultation—was internal to exec brach and didn‘t involve public. 2. Contrast this w/ a case where procedural violation is failure to offer public hearing, where Ps would have objected at hearing but couldn‘t. g. Informational Injury i. Legal rights test: check to see if purpose of law is to give P right to info: Akins. 1. FEC v Akins, USSC, 1998. Group of voters sued to challenge FEC‘s determination that jewish group wasn‘t a political committee and thus didn‘t need to make disclosures. Held: voters have standing b/c of information injury, and b/c their right to info comes from statute. a. uses older legal right conception of standing to replace normal injury in fact requirement and distinguish away taxpayer standing cases (Richardson) ii. If purpose of reg is to provide info for purely intragov‘tal not prepared for public consumption or subject to public comment, (even though available to public under FOIA), then probably no injury. 2. Causation a. Procedural Violations and Causation i. Lujan: suggests that procedural violations don‘t cause the injury. ii. However, you can complain about procedural violations if ultimate action would cause injury: 1. Agency fails to prepare environmental impact statement before building dam. Landowner whose land would be flooded can show causation. This is hypo from Lujan in which court demonstrated how a person could establish standing when there is a procedural violation—b/c there‘s also an actual injury caused by the agency decision to build the dam (not the procedural violation itself). b. Third Party Actions and Causation i. No causation if injury actually caused by third parties, not the agency. Simon v Eastern Ky Welfare Rights Org. 1. Simon v Eastern Ky Welfare Rights Org, USSC, 1976. Ps are several indigents and organizations representing them. Challenging IRS‘s revenue ruling modifying responsibilities of nonprofit hospitals qualifying as charitable orgs. Old rule required hospital to accept some patients in need of hospital services who cannot pay for them to the extent of its financial ability. New rule said that so long as emergency room accepted everyone, other services could accept only paying customers. Ps allege that new rule encourages hospitals to deny services to indigents, and that as indigents they are intended beneficiary of statute. Held: no standing; Ps failed to allege that injury was cause of agency action; plus, victory wouldn‘t have given them desired remedy. 2. If P had been able to show that historically, orgs always altered behavior to retain tax-exempt status, case would have come out other way. Where agency‘s action has determinative or coercive affect on third party, causation satisfied by a showing of that effect. Bennett v Spear. 3. Redressability: must show that favorable court decision will likely prevent/redress injury. a. Third Party Actions and Redressability i. Causation problems can also be viewed as redressability problems. 1. e.g., Simon (requiring IRS to withdraw hospital‘s tax-exempt status wouldn‘t relieve injury of lack of free care for poor) 2. e.g., Lujan (barring funding doesn‘t prevent third party from building project causing extinction) a. Note that if P could show that w/out fed‘l funding, project couldn‘t be built, that would be enough to satisfy the ―likely‖ standard. ii. Procedural Problems and Redressability 1. Correcting procedural violation doesn‘t necessarily stop injury from occurring. But normal rules of redressability don‘t apply re: procedural violations: 2. Lujan: ―The person who has been accorded a procedural right to protect his concrete interests can assert that right without meeting all the normal standards for redressability and immediacy.‖ iii. Prudential Requirements 1. Zone of interests test: § 702 of APA: P must fall w/in zone of interests intended to be protected by statute/reg in question a. It‘s for congressional resolution; need positive law to confer standing b. Look not at whole statute, but at the particular provision that is allegedly violated, not some other provision. c. Only need to be arguably in the zone of interests. Data Processing. 2. No generalized grievances (e.g., Richardson) – but see Akins. 3. Associational standing a. Members must have own standing b. Interests at stake must be germane to organization c. Individual member participation not required.