Concept
Roberts Court Limits
Removal of officers
Roberts Court Limits
Morrison v. Olson asked a functional question: does a removal restriction impede the President’s ability to perform his constitutional duty? For the independent counsel the Court answered no, pointing to the limited nature of her responsibilities and to Congress’s perceived need for her to function independently.1 The Roberts Court has displaced that balancing approach with a more formalist method that favors bright-line rules, and those rules constrain Congress’s ability to rein in the executive branch.2
Two decisions carry the change, and Chief Justice Roberts used the same technique in both. Writing for a five-Justice majority in Free Enterprise Fund and again in Seila Law, he disclaimed any intent to “revisit our prior decisions allowing certain limitations on the President’s removal power” and trained his focus on a “novel” aspect of the arrangement before him.3 In Free Enterprise Fund the novelty was two layers of for-cause protection. In Seila Law it was a single director shielded by for-cause protection.
Counting is easy to apply, and it can be blunt. All else equal, an official protected by two layers may enjoy more independence than one protected by a single layer, but all else is often not equal, because other variables also affect independence. Free Enterprise Fund pushed those variables aside and fixed on the number of layers.4 Seila Law fixed on the number of heads. The subsections below take each rule in turn and then collect the questions the Court has left open.
Two Layers of Protection
Free Enterprise Fund v. Public Company Accounting Oversight Board held that two layers of for-cause protection between the President and an officer are unconstitutional, because the dual layers impede the President’s ability to ensure that the laws are faithfully executed.5 The textual anchor is the Take Care Clause: the President “shall take Care that the Laws be faithfully executed.”6 The rule counts layers, and what it counts is protection against removal inside the executive branch.
The challenge began when the Board opened a formal investigation of Beckstead and Watts, a Nevada accounting firm registered with it. The firm and the Free Enterprise Fund, a nonprofit to which it belonged, sued for a declaratory judgment and an injunction barring the Board from exercising its powers. They argued that the Sarbanes-Oxley Act gave Board members wide-ranging executive power without subjecting them to presidential control.7 Two protections stood between the President and the Board. The SEC could remove a Board member only for willful violations of the Act, Board rules, or the securities laws, willful abuse of authority, or unreasonable failure to enforce compliance, as determined in a formal Commission order made on the record after notice and a hearing. The Act did not even let the Commission remove a member who cheated on his taxes, because that conduct was not among the listed grounds.8 The Commissioners, for their part, were removable by the President only for inefficiency, neglect of duty, or malfeasance in office, a protection the Court assumed for purposes of decision.9 The Chief Justice described the combination as one that “does not merely add to the Board’s independence, but transforms it.”10
The earlier removal cases involved a single layer. Only one level of protected tenure separated the President from the officer exercising executive power, and the President, or a subordinate he could remove at will, decided whether the officer’s conduct met the good-cause standard.11 Morrison fits the pattern: the statute gave the Attorney General, an officer directly responsible to the President and through whom the President could act, several means of supervising or controlling the counsel.12
The Sarbanes-Oxley Act did something quite different. It not only protected Board members from removal except for good cause but withdrew from the President any decision about whether good cause exists, and vested that decision in other tenured officers, the Commissioners, none of whom is subject to the President’s direct control.13 The result was “a Board that is not accountable to the President, and a President who is not responsible for the Board.”14
The Court identified two linked harms. The first is lost control. When the First Congress created the first executive departments, it settled on the view that the executive power includes a power to oversee executive officers through removal, and that because the Constitution had not expressly taken that power away, it remained with the President. Madison put the point by saying that if any power is in its nature executive, it is the power of appointing, overseeing, and controlling those who execute the laws.15 Impeachment does not displace that power. Article II provides for removal of civil officers on impeachment and conviction,16 but the First Congress rejected the view that impeachment is the only route, on the pragmatic ground that it made removal of incompetent or corrupt administrators too difficult, and it accepted that impeachment supplemented rather than displaced removal by less arduous means.17
The second harm is lost accountability. Without the ability to oversee the Board, or to attribute its failings to officials he can oversee, the President is no longer the judge of the Board’s conduct; he can neither ensure that the laws are faithfully executed nor be held responsible for a member’s breach of faith. Diffusion of power carries diffusion of accountability, and without a clear and effective chain of command the public cannot tell where the blame for a pernicious measure should fall.18
The Court also saw that the problem would grow. If the dispersion of responsibility were allowed to stand, Congress could multiply it by adding still more layers of good-cause tenure. Because Congress controls the salary, duties, and existence of executive offices, only presidential oversight can counter its influence, and multilevel tenure protections offered a blueprint for extensive expansion of legislative power.19 In pursuit of a “workable government,” Congress cannot reduce the Chief Magistrate to a “cajoler-in-chief.”20
The count tracks levels of protected tenure, not rungs of the hierarchy. An officer may answer to several supervisors and still fall within the single-layer cases, so long as the President or an at-will subordinate decides whether cause exists.21 Students err by treating every for-cause limit as invalid, or by counting supervisors instead of protections.
The Commission’s other powers did not cure the second layer. Altering the budget or powers of an agency as a whole is a problematic way to control an inferior officer, because the Commission cannot wield a free hand to supervise individual members if it must destroy the Board in order to fix it.22
The remedy followed the defect. Invalidating the removal restrictions leaves the Board removable by the Commission at will and leaves the President separated from Board members by only a single level of good-cause tenure, so the Commission is fully responsible for the Board’s actions, which are no less subject than the Commission’s own functions to presidential oversight.23 Congress remains free to restructure the Board further, for example by making its members removable by the President for good cause or at will, but that “editorial freedom—far more extensive than our holding today—belongs to the Legislature, not the Judiciary.”24
The removal holding also settled the Board members’ status as inferior officers. Their work was supervised and directed at some level by principal officers, the SEC commissioners, and because the removal provisions fell, the Act is properly read as giving the Commission full power to remove Board members at will, one of the defining characteristics of an inferior officer.25 The Court rejected three further challenges to the Board under the Appointments Clause.26
The dissent attacked the premise that a second layer matters. Neither the Necessary and Proper Clause nor the separation of powers is absolute in removal cases: depending on the nature of the office, its function, or its subject matter, Congress sometimes may limit the President’s authority to remove an officer.27 Administrative units come in many shapes and sizes, and the functional approach of the Court’s precedents asks how a removal provision actually affects presidential power, which is not always susceptible to elementary arithmetic.28
Applying that approach, the dissent tested the majority’s claim against scenarios in which the President and the Commission disagree about a Board member. Where the President wants to keep a member whom the Commission wants to dismiss, the second layer helps the President by hindering the Commission.29 Where the President wants a member removed and the Commission reasonably declines, the President remains “powerless to intervene” even after the lower layer is struck, because the Commissioners themselves are protected, as the majority assumed.30 The dissent conceded that such situations are logically possible, doubted that they matter, and concluded that because the double layer sometimes helps and sometimes hurts the President, its overall effect is at most indeterminate.31
The dissent also defended the Board on functional grounds. The Board’s adjudicatory responsibilities, its technical character, the need to attract experts, and the importance of showing the public that the job is nonpolitical justify assuring members that they need not fear for their jobs when they carry out their tasks competently, while the Commission remains the ultimate authority over Board policies and actions.32 And it warned about reach. It found 48 agencies whose heads are removable only for cause and 573 high-ranking officials in those agencies who are themselves removable only for cause, and it saw no way to avoid sweeping hundreds, perhaps thousands, of senior officials into the holding.33 The majority saw no reason to address whether any of those positions would fall and left the question for a future case.34 The Open Questions subsection takes up what that leaves unsettled.
Single Directors After Seila Law
Seila Law LLC v. Consumer Financial Protection Bureau held that Congress may not insulate from presidential removal a single individual who heads an independent agency and wields significant executive power. The Consumer Financial Protection Bureau, an independent regulatory agency charged with ensuring that consumer debt products are safe and transparent, was led by one Director removable only for inefficiency, neglect, or malfeasance.36 The Director served a five-year term37 and had “no boss, peers, or voters to report to,” yet wielded vast rulemaking, enforcement, and adjudicatory authority over a significant part of the economy.38 Restricting the removal of that “principal officer” unconstitutionally compromised the President’s power to supervise and remove someone who exercised executive power, and the Court was explicitly unwilling to extend congressional authority over removal that far.39
Seila Law proceeds from a general rule and two exceptions, and an answer should follow that order. The general rule, from Myers v. United States, allows the President to remove subordinates at will.40 The Court read Humphrey’s Executor and Morrison narrowly, as limited exceptions to the Myers principle: one for multimember expert agencies that do not wield substantial executive power, and one for inferior officers with limited duties and no policymaking or administrative authority.41
Humphrey’s Executor approved for-cause removal restrictions on “a multimember body of experts, balanced along partisan lines, that performed legislative and judicial functions and was said not to exercise any executive power.”42 The Federal Trade Commission, the Court had said there, “cannot in any proper sense be characterized as an arm or an eye of the executive”; it acted “in part quasi-legislatively and in part quasi-judicially,” and Congress had made its duties “free from executive control.”43 Congress could create such agencies and require them to act independently, because one who holds office only at another’s pleasure cannot be depended on to maintain an attitude of independence against that other’s will.44
Morrison v. Olson supplies the second exception, which Seila Law confined to inferior officers with limited duties and no policymaking or administrative authority.45
The Bureau fit neither exception. Humphrey’s did not apply because the Bureau was led by a single Director who qualified neither as a “body of experts” nor as non-partisan, and the Director was not a mere legislative or judicial aid in light of significant rulemaking, adjudicatory, and enforcement authority. Morrison did not apply because the Director was not an inferior officer with limited duties.46 The Court therefore treated the case as a request to extend its precedents to a “new configuration”—“an independent agency that wields significant executive power and is run by a single individual”—and was unwilling to take that step.47
Two reasons support the refusal to extend the exceptions, and each does separate work. The first is history. Lower-court judges, litigants, and amici had turned up “only a handful of isolated incidents in which Congress had provided good-cause tenure to principal officers who wield power alone rather than as members of a board or commission,” and a structure with one director at its head, insulated from presidential review, was described as “wholly unprecedented” and “an indication of a severe constitutional problem.”48
The second is accountability, and it explains why the number of heads matters. As the lone leader of a powerful body, the Director could act “unilaterally” with “no colleagues to persuade, and no boss or electorate looking over her shoulder.” That arrangement contravened a basic principle of checks and balances, which, with the significant exception of a President who is persistently watchdogged by and directly accountable to the citizenry as a whole, “scrupulously avoids concentrating power in the hands of any single individual.”49
The Bureau’s funding sharpened the concern. It did not rely on the usual appropriations process; it drew its money from the Federal Reserve, which is itself funded outside the appropriations process through bank assessments.50
The remedy followed the defect. The Court invalidated the removal restrictions and severed the Director’s for-cause protection from the statute, so the Director became removable by the President at will.51
A common error is to read Seila Law as barring for-cause protection for every lone official. The rule reaches a principal officer who alone heads an agency and wields significant executive power, so a stem that gives a lone official a superior, narrow duties, and no policymaking authority is asking about the inferior-officer exception.52
Collins v. Yellen applied the same reasoning to the Federal Housing Finance Agency, which is also led by a single Director.53 The Court called its result a “straightforward application of our reasoning in Seila Law,” notwithstanding the difference in the “relative importance of the regulatory and enforcement authority” between the two agencies.54 As Justice Sotomayor put it in a partial dissent, “all that matters is that [the FHFA] is an agency led by a single Director.”55 A narrower portfolio therefore does not rescue a single head. The structure controls.
Humphrey’s Executor has survived for now, but only as to multimember agencies. Seila Law and Collins dealt with agencies headed by one individual, unlike Humphrey’s, which upheld removal restrictions on the multimember FTC, so the two cases did not resolve “one of the fiercest (and oldest) fights in administrative law”: the continuing validity of the Humphrey’s exception to Myers’s general rule.56 Justices Thomas and Gorsuch would have gone further in Seila Law and overruled Humphrey’s entirely, and whether the Court as a whole will follow them remains to be seen.57
Critics press hardest on whether so categorical a rule is justified. The prevailing majority has displaced Morrison’s functional balancing with bright-line rules.58 The Free Enterprise Fund dissent had argued that administrative units come in many shapes and sizes, and that a rule can strengthen a President as easily as weaken him, since a rule that takes power from the President’s opponents may help him.59 A rule keyed to the number of heads does not ask which of those effects is present.
The third step carries the most uncertainty, because Justices Thomas and Gorsuch would overrule Humphrey’s and the Court has not said whether it will follow them. The next subsection takes up that question.65
Open Questions
Free Enterprise Fund fixed a rule for two layers and then marked off what it was not deciding. Several passages implied a categorical bar on multi-layer for-cause arrangements, but others left the door open for carve-outs: the Chief Justice suggested that the rule did not necessarily apply to the millions of governmental employees (as opposed to “officers”) in the federal civil service, the tens of thousands of commissioned officers in the U.S. military, or some 1,500 administrative law judges (ALJs) serving across federal agencies.67 Those suggestions raised follow-on questions about the decision’s precise import and scope, and Seila Law and Collins, which concerned single directors, did not answer them.68 Four matters therefore remain unsettled: whether Humphrey’s Executor survives, and whether the two-layer rule reaches ALJs, the civil service, or the military. An answer that treats any of them as decided misstates the law.
Humphrey’s Executor is the broadest open question. Seila Law described it as approving for-cause protection for a multimember expert body that performed legislative and judicial functions and was said to exercise no executive power, and the Court refused to extend that approval to a single director.69 Morrison had already moved away from Humphrey’s analysis: it recognized that the independent counsel’s duties were executive and upheld good-cause protection anyway.70 Lower courts meanwhile continue to apply Humphrey’s to commissions. In Consumers’ Research v. Consumer Product Safety Commission, the Fifth Circuit noted that Seila Law “cast doubt” on the constitutionality of agencies like the CPSC, yet observed that although the logic of Humphrey’s may have been overtaken, the decision “has not been overruled—at least not yet.”71
The reach of the two-layer rule to ALJs is the most pressing question. The Administrative Procedure Act requires an agency engaged in formal adjudication to use the agency itself, one of its members, or an ALJ to preside, and because the first two are impracticable, agencies almost invariably designate an ALJ.72 An ALJ is an inferior officer employed by the agency at which he presides, yet almost as independent of that agency as a federal district judge is of the Executive and Legislative branches. That independence has a history: throughout the 1930s Congress received complaints that hearing examiners favored the agencies where they presided, studies confirmed the pro-agency bias, and Congress responded by severely limiting the power of agencies to appoint biased decision makers or to subject them to any form of pressure.73
Removal of an ALJ is limited at two levels, by restrictions on the Merit Systems Protection Board’s power to remove ALJs and by similar restrictions on the removal of the Board’s own members.74 It would seem to follow that the multiple limits violate the Take Care Clause as Free Enterprise Fund interpreted it. That is far from clear, however, because the Court said in a footnote that it was not addressing the removal of ALJs, and a good argument can be made that assuring ALJs’ decisional independence furthers the goals of due process and is consistent with the Take Care Clause.75 The Free Enterprise Fund dissent asked the sharper question: would the Court’s potential exception for adjudicators apply to an ALJ who also has important administrative duties beyond pure adjudication?76
The case for protecting ALJs rests on neutrality. Politics should normally play no role in adjudication, because a person’s rights or duties should not hinge on partisan considerations, so when executive agencies adjudicate it may be critical that those who perform these duties not serve at the whim and pleasure of the President.77 Defenders add that the Take Care Clause is not offended, because the same statute lets the agency replace an ALJ’s initial decision with its own, which gives the presidentially appointed agency head complete power over the legal and policy content of every agency decision.78 The analogy to judges has a limit. Judges are protected by the Constitution’s command that they “shall hold their Office during good Behavior,” and the text has no parallel provision safeguarding the tenure of non-judicial officers.79 A challenger would add that, as the Court observed about the Commission and the Board, control over an agency’s powers is a poor substitute for control over the individual officer.80
Lower courts are split in spirit. Jarkesy v. SEC brought the two-layer problem for SEC ALJs before the Fifth Circuit, and the Supreme Court granted certiorari but declined to rule on the issue when it affirmed on other grounds. A federal district court later invalidated for-cause removal restrictions for ALJs in the Justice Department’s Office of the Chief Hearing Officer in Walmart Inc. v. King, because of the combination of restrictions on the Merit Systems Protection Board’s power to remove ALJs and similar restrictions on removal of the Board’s members.81 Neither ruling binds the Supreme Court.
The dissent in Free Enterprise Fund saw the practical stakes. The federal government relies on 1,584 ALJs in over 25 agencies, and the dissent asked whether every losing party before an ALJ now has grounds to appeal on the basis that the decision against him is unconstitutional. It also asked what happens in the meantime: whether the work of all these officials is to be put on hold while the courts of appeals decide, whether Congress must act to remove the for-cause provisions, and whether the President can then restore them by executive order.82
The suggested carve-out for the civil service rests on the line between employees and officers. The Court has defined an officer as an appointee exercising significant authority under the laws of the United States, and “lesser functionaries” fall outside the Appointments Clause.83 The dissent’s count of 573 senior officials who are removable only for cause shows why the line is contested rather than settled. Military officers present the sharper difficulty. The dissent said the majority might simply answer that the military is different, but it would have to explain how, since it is hard to see why the Constitution would give a President who is Commander in Chief less authority to remove inferior military officers than comparable civil officials.84 The majority left that question for a future case.
That uncertainty follows from the Court’s method. It displaced Morrison’s balancing approach with bright-line rules,87 and a rule keyed to layers and heads says nothing by itself about officials who differ in kind, such as adjudicators and military officers. Each open question above sits in that gap, which is why courts reason by analogy and why an answer must say which analogy it is drawing.