Concept
Control of Executive Officers
As chief executive, including the “take care” clause
Control of Executive Officers
Article II charges the President to “take Care that the Laws be faithfully executed” and, in the same sentence, to “Commission all the Officers of the United States.”1 The pairing fits the office. Section 1 vests the whole executive power in one person,2 but one person cannot enforce the federal code alone, so subordinates do the work. The previous section showed that the Take Care Clause is read to require meaningful presidential control over those subordinates. This section asks what that control consists of when a subordinate resists, and the answer runs through removal.
An instruction guides an officer only if the officer has something to lose by ignoring it, and the most direct thing an officer can lose is the job. Yet neither clause quoted above mentions removal. Apart from impeachment, the Constitution says nothing about how an executive officer’s tenure ends.3 Any presidential power to dismiss officers, and any limit Congress may place on that power, has been implied from the structure of Article II.4 Because the doctrine rests on implication, the cases have disagreed about how much Congress may do to shield an officer from presidential control.
Appointment has its own provisions in Section 2, so this unit concentrates on removal, where the Take Care Clause does most of its work. Four groups of decisions follow. Myers ties removal to the executive power itself. Bowsher asks what happens when Congress keeps a removal power for itself. Humphrey’s Executor, Perkins, Wiener, and Morrison v. Olson test when Congress may restrict the President’s own power to dismiss. Free Enterprise Fund and Seila Law mark where those restrictions stop. One question carries through all four: who other than the President controls the officer’s tenure, and how much executive power does the officer wield?
Myers and Removal at Will
Article II, Section 1 vests “the executive Power” in the President, and Myers v. United States asked whether that power includes control over who stays in office.5 The Constitution has no removal clause for executive officers. Section 4 provides for removal by impeachment, and Chief Justice Taft observed that it is the only express provision on removals and that the Convention never discussed the subject.6 Any presidential power to remove is therefore implied. It has long been traced to the vesting sentence together with the duty in Section 3 to take care that the laws be faithfully executed,7 and Madison thought it “absolutely necessary” that the President be able to remove appointees.8
The statute in Myers provided that first-, second-, and third-class postmasters were to be appointed, and could be removed, by the President with the Senate’s advice and consent, for a term of four years.9 A first-class postmaster was removed without that consent, and the question was whether the President has the exclusive power to remove executive officers whom the President appointed with the Senate’s advice and consent.10 By a vote of six to three, the Court held the consent requirement invalid.11
Taft’s opinion rests on four connected arguments, and the first starts from the vesting sentence. If appointments and removals of executive subordinates are not exercises of executive power, he asked, what are they? They are plainly neither legislative nor judicial.12 The British Crown held both powers, so the framers would naturally have read “executive power” to include both.13 A President who must act through subordinates, and who is charged to see that the laws are faithfully executed, must be able to choose those subordinates and to remove those for whom the President cannot continue to be responsible.14
The implication carries weight because of accountability. A President answerable for how the laws are executed would otherwise answer for officers the President could neither correct nor replace. Madison made the point in the First Congress: if the President alone holds the removal power, the chain of dependence is preserved, with the lowest, middle, and highest officers depending on the President and the President on the community, whereas a Senate share in removal would “abolish at once that great principle of unity and responsibility in the Executive department.”15 Taft supplied the practical version. Requiring the President to file charges and submit them to the Senate might make impossible the unity and coordination that effective executive administration needs.16
The second argument ties removal to appointment. Section 2 lets the President appoint, with the Senate’s advice and consent, the officers of the United States whose appointments the Constitution does not otherwise provide for, and it lets Congress vest the appointment of inferior officers in the President alone, in the courts of law, or in the heads of departments.17 Taft treated removal as an incident of that appointment power, because officials who select their executive subordinates need the power to remove them in order to meet their own responsibility.18 The incident attaches to appointment and not to consent. In Taft’s words, “[t]he power of removal is incident to the power of appointment, not to the power of advising and consenting to appointment.”19 The Senate’s check on appointments therefore brings no share of the removal power with it.
The third argument reads the Senate’s role as a limit. The words of Section 2 that follow the general grant in Section 1 are either an enumeration of specific executive functions or limits on that grant, and limits should not be enlarged beyond the words used.20 Taft also explained why the two checks differ in practice. When the Senate rejects a nominee, the President ordinarily has an ample field of other capable candidates. A Senate veto over removal is a much greater limitation and a much more serious blending of legislative with executive functions, so it is not to be implied.21
The fourth argument is history. The First Congress took up the question within two years of the Convention, and its leaders included members of the Convention. Taft read its vote as a legislative declaration that the power to remove officers appointed with the Senate’s consent vested in the President alone, a meaning no one doubted until the Johnson impeachment trial of 1868.22 He cautioned that a congressional conclusion on a constitutional question is not conclusive. He gave the decision weight because he agreed with its reasons, because it came so soon after the Convention on a question of primary importance, and because it would shape the later laws organizing the government.23 History supports the first three arguments as evidence and does not replace them.
Taft’s conclusion states the result. Article II grants the President “the general administrative control of those executing the laws, including the power of appointment and removal of executive officers,” a conclusion confirmed by the duty to take care that the laws be faithfully executed.24 Congress may provide for removals of inferior officers only after it has vested their appointment in some authority other than the President with the Senate’s consent, and only on condition that it does so.25 Buckley v. Valeo later summarized the holding as a ruling that Congress could not by statute divest the President of the power to remove an officer in the Executive Branch whom the President was initially authorized to appoint.26
That condition marks the other pole of the doctrine. Taft’s own account of the postal service shows how the two poles work. Postmasters were long appointed by the Postmaster General because Congress so provided under the excepting clause, but once Congress required certain classes to be appointed by the President with the Senate’s consent, they became subject to removal by the President alone, and any contrary legislation had to fall.27 United States v. Perkins illustrates the other side of the line, because Congress may restrict the removal of inferior officers whose appointment it has vested in a department head.28 An older formulation about executing a treaty states the same default and exception. Congress may prescribe the mode of execution and may devolve execution on others, but “till this be done, it seems the duty of the Executive department to execute the contract by any means it possesses.”29 The two cases differ on one fact, which is who made the appointment.
Taft marked one limit on his own rule. Some executive officers perform quasi-judicial duties, as members of executive tribunals whose decisions after hearing affect individuals do, and in a particular case the President cannot properly influence or control those decisions. Even then the President may treat a decision, once rendered, as a reason for removal if the officer’s discretion has not been wisely exercised.30 The reservation matters because the Court soon read Myers narrowly, as deciding only that the President could remove a first-class postmaster without the Senate’s advice and consent, and treated Taft’s broader statements as beyond the point decided.31 The later subsections follow that narrowing.
Step two carries the contrast between the postmaster the President could remove and the inferior officer Congress could protect. Step five is where this subsection hands off to the next ones, which ask what Congress may do when it neither takes the removal power for itself nor requires the Senate’s consent.
Congress Cannot Remove Officers
Article II, Section 4 supplies the only removal role the Constitution gives Congress: the President, the Vice President, and all civil officers “shall be removed from Office on Impeachment for, and Conviction of, Treason, Bribery, or other high Crimes and Misdemeanors.”33 Impeachment by the House and trial by the Senate can rest only on those grounds, so it is a deliberate process reserved for serious wrongdoing.34 When the House impeaches an executive officer for bribery and the Senate convicts, the removal is lawful because the Constitution prescribes that route. Bowsher v. Synar governs every other congressional role, and its rule is that Congress cannot reserve for itself the power to remove an officer charged with executing the laws except by impeachment.35 Myers barred Congress from conditioning removal on the Senate’s consent. Bowsher takes the next step and asks what happens when Congress keeps the removal power for itself.
The statute in Bowsher was the Gramm-Rudman-Hollings Act, which aimed at a balanced budget by setting maximum annual deficits and, when the deficit was projected to exceed the maximum, requiring across-the-board spending cuts. The Office of Management and Budget and the Congressional Budget Office each estimated the deficit and calculated the necessary reductions. They submitted a joint report to the Comptroller General, who reviewed their estimates and reported his own conclusions to the President, and the President had to put the Comptroller General’s reductions into effect unless Congress met the deficit goal by other means. The Comptroller General headed the General Accounting Office, which the Budget and Accounting Act of 1921 created. The President nominated him from a list recommended by congressional leaders, but only Congress could remove him, by impeachment or by joint resolution on grounds the 1921 Act specified.36
The Court reasoned first from the structure of Article II. After an officer is appointed and confirmed, the Constitution provides for removal by Congress only through impeachment and conviction on the grounds Section 4 lists, and a direct congressional role in removing officers who execute the laws beyond that one is, in the Court’s words, “inconsistent with separation of powers.”37 History points the same way. In the First Congress, Madison’s position prevailed and a congressional role in removal was rejected, and that “Decision of 1789” provides “contemporaneous and weighty evidence” of the Constitution’s meaning because many Members of the First Congress had taken part in framing it.38 Myers had already refused to read the excepting clause as letting Congress draw to itself, or to either branch of it, the power to remove or a right to participate in removal, because that would infringe the separation of governmental powers.39
The second reason is practical, and it carries the holding. An officer takes direction from whoever can end the officer’s tenure, so a removal power held by Congress is in effect a power to supervise. The Court quoted the district court: “Once an officer is appointed, it is only the authority that can remove him, and not the authority that appointed him, that he must fear and, in the performance of his functions, obey.”40 From that premise the Court concluded that vesting the execution of the laws in an officer answerable only to Congress would, in practical terms, reserve in Congress control over the execution of the laws.41 The structure of the Constitution does not permit Congress to execute the laws, so Congress cannot grant to an officer under its control a power it does not possess.42
The Court tied that reasoning to INS v. Chadha, which struck down the legislative veto, the device by which Congress reverses executive action without passing a new law through both Houses and presenting it to the President.43 An officer controlled by Congress would hand Congress a veto in another form, because Congress could remove, or threaten to remove, the officer for executing the laws in any way Congress found unsatisfactory.44 The threat alone does the work. Congress would never need to enact a statute changing the budget reductions. It would only need to make plain that an unwelcome calculation could cost the Comptroller General the job. The Act’s worthy aim of restraining persistent excessive spending did not save the mechanism, because a failure of political will does not justify an unconstitutional remedy.45
The dissenting opinions attacked the majority’s rule as too rigid. They would have asked a functional question, whether a limit on removal prevents the Executive Branch from accomplishing its constitutionally assigned functions, and they argued that the Comptroller General’s practical independence made the threat of congressional control remote.46 The two sides disagree about what counts as the harm. The majority treats the removal power itself as the problem, because an officer removable only through Congress is under Congress’s control whether or not Congress ever carries out the threat. The dissent asks how the office actually works. Because the majority’s holding states the governing rule, an exam answer follows it, and the dissent serves as the counterargument that a rule keyed to formal removal terms can overlook real independence.
The rule also has a limit. Bowsher bars a congressional role in removal. It does not by itself decide whether Congress may limit the grounds on which the President or another executive official may remove an officer, and the Court has permitted some such limits.47 In Morrison v. Olson, the Court stressed that the removal power there stayed with the Executive, which could terminate the counsel for good cause.48 The line that results separates two kinds of statutes. A statute that gives Congress, or an arm of Congress, power to make or block a removal decision is always invalid. A statute that leaves removal in the Executive Branch and limits only the grounds is sometimes valid.
Step three carries the holding and step four marks its edge. The Director in the pipeline statute fails at step three. The Commissioner in the rail statute reaches step four, where the question becomes whether the limit on the President’s removal power interferes with the President’s ability to execute the laws, which the next subsection takes up.
Permissible Good-Cause Limits
Article II vests “the executive Power” in the President and, in Section 3, directs the President to “take Care that the Laws be faithfully executed.”50 Myers and Bowsher stop Congress from sharing in the removal of an officer who executes the laws, whether by requiring the Senate’s consent or by keeping the removal power for itself. A narrower question arises when Congress leaves removal inside the Executive Branch and limits only the reasons for which it may be used, most often by allowing removal for cause alone. The Court has permitted some such limits, though not all.51 A limit on grounds can survive where a congressional role in removal cannot because the Take Care Clause asks whether the President can still see that the laws are faithfully executed. A cause standard fits that duty so long as someone in the executive chain keeps a real means of correcting an officer who fails. The cases that uphold such limits are Humphrey’s Executor, Perkins, and Morrison, and Wiener extends the same logic to a statute that is silent on removal.
Humphrey's Executor v. United States came first. The statute gave Federal Trade Commission commissioners fixed terms and allowed removal during the term only for “inefficiency, neglect of duty, or malfeasance in office.” The President did not claim that the commissioner he dismissed was guilty of any of these offenses, because he wanted to change the policy direction of the agency. The Court held the removal-for-cause provision a constitutionally proper limit on the President’s removal power.52
The Court’s reasoning began by reading Myers narrowly. The “narrow point actually decided” in Myers, Justice Sutherland explained for a unanimous Court, was only that the President could remove a first-class postmaster without the Senate’s advice and consent, and Taft’s broader statements went beyond that point and were not entitled to stare decisis effect.53 The Commission, in the Court’s view, was a different kind of body from a postmaster’s office. Morrison later said that the quasi-legislative and quasi-judicial labels in Humphrey’s Executor and Wiener largely reflected a judgment that it was not essential to the President’s proper execution of his Article II powers that those agencies be headed by individuals removable at will.54 Independence was the point of tenure, since an official who holds office only during another’s pleasure “cannot be depended upon to maintain an attitude of independence against the latter’s will,” and that is why Congress could forbid removal except for cause.55 Myers itself had left room for this result, because Taft acknowledged quasi-judicial duties that the President cannot properly influence in a particular case.56
Perkins supplies the rule for inferior officers. When Congress vests the appointment of inferior officers in the heads of departments, it may restrict the power to remove them, and Taft’s own conclusion in Myers left the door open for exactly that case.57 The Court has since allowed Congress to vest the removal power in persons other than the President and to limit the circumstances in which certain executive officials may be removed, although not every restriction is valid.58 Congress’s choice of the appointing authority is what makes the limit available.
Wiener v. United States tested what a statute’s silence means. The statute creating the War Claims Commission said nothing about removing commissioners, yet the Court prevented a commissioner from being dismissed without cause. Because the commissioners were expected to “adjudicate according to law,” the Court inferred a congressional desire to protect them from arbitrary removal.59 Adjudication of individual disputes is often thought to call for independent, apolitical judgment, and that thought supplied the reason for reading protection into a silent statute.60
The first dismissal fails for the reason given in Wiener. Congress’s silence in a statute that creates a body meant to adjudicate impartially can be read to protect its members from arbitrary removal, and a wish to install the President’s own appointees is not a ground that protection allows.61 The second dismissal rests on a failure of performance, the kind of reason a cause standard permits, so it likely stands. The two removals involve the same Board, the same statute, and the same silence, and they differ only in the reason given. That difference is how a cause standard works in practice, because it governs why an officer may be removed and leaves the power itself in the executive.
Morrison v. Olson moved the test from the officer’s label to the effect of the limit. The independent counsel statute provided for a counsel appointed by a special court and removable by the Attorney General for good cause.62 Unlike Myers and Bowsher, the statute gave Congress no role in removal, because the power stayed with the Executive Branch. The Court said the answer did not turn on classification. Its analysis of the officials’ functions was not irrelevant, but “the real question is whether the removal restrictions are of such a nature that they impede the President’s ability to perform his constitutional duty,” and the functions of the officials must be analyzed in that light.63
Applied to the independent counsel, that test came out in favor of the limit. The statute did not strip the removal power from the Executive. Because the counsel could be terminated for good cause, the Executive, through the Attorney General, retained ample authority to ensure that the counsel was competently performing statutory duties, and the Attorney General could remove the counsel for misconduct.64 The Court did not think the limit deprived the President of enough control to interfere impermissibly with the duty to ensure faithful execution of the laws.65
The mechanism explains why a cause standard can satisfy the Take Care Clause. The clause makes the President answerable for faithful execution, and faithful execution is threatened when an officer neglects the job, abuses it, or cannot perform it. A cause standard lets an executive official act on those failures. What it takes away is removal for reasons unrelated to performance, such as a wish to replace the officer with someone more congenial. Congress may also provide for judicial review of a removal decision to confirm that the statutory criteria are met, because review adds no burden to the President’s executive authority.66 The boundary lies at complete deprivation. The Court stressed that the statute did not completely strip the removal power from the President, which would leave no means of ensuring faithful execution.67
Justice Scalia dissented, and his objection shows what the majority’s test gives up. He observed that the line between purely executive functions and quasi-legislative or quasi-judicial ones had often been criticized as unclear and even irrational, but he argued that it at least identified officers whose functions were entirely within the President’s control.68 If the removal of a prosecutor, “the virtual embodiment of the power to ‘take care that the laws be faithfully executed,’” can be restricted, he asked, what officer’s removal cannot? He called the result “an open invitation for Congress to experiment.”69 The majority’s rule governs, so an exam answer applies the impediment test, and the dissent supplies the counterargument that the test is elastic.
Later decisions read these cases more narrowly than Morrison’s open-ended question suggests. Seila Law describes two exceptions to the President’s general power to remove those who wield executive power on the President’s behalf. The first is Humphrey’s Executor, which allowed for-cause protection for an independent agency headed by a multimember body of nonpartisan experts with quasi-legislative or quasi-judicial authority. The second covers inferior officers with limited duties, the category that Perkins and Morrison illustrate.70 A good-cause problem should therefore be run with Morrison’s functional question and these two categories in mind.
Step three does the work in most problems. It separates a statute that merely limits the reasons for removal from one that leaves no executive official able to act on dereliction.
Layered and Single-Head Limits
Article II vests “the executive Power” in a single President and charges that President to “take Care that the Laws be faithfully executed.”72 The previous subsection showed that Congress may restrict the grounds of removal when an executive official can still act on an officer’s neglect or misconduct. Two lines of decision mark where that permission ends. In Free Enterprise Fund v. Public Company Accounting Oversight Board, the Court held that two levels of for-cause protection between the President and an officer who executes the laws impair the President’s oversight of faithful execution, because Article II “makes a single President responsible for the actions of the Executive Branch.”73 Seila Law and Collins then applied related reasoning to agencies run by one director. Both limits rest on the same worry, that an officer who wields executive power sits where the President cannot reach.
The Sarbanes-Oxley Act created an accounting oversight Board whose members the SEC could remove only for good cause, and the President could remove SEC Commissioners only for cause. The question was whether those two layers could be stacked. The Court said they could not, observing that the Act was “highly unusual in committing substantial executive authority to officers protected by two layers of good-cause removal.”74
The answer turns on who decides whether cause exists. With one layer of tenure between the President and the officer, the President, or an official the President can remove at will, decides whether the officer’s conduct justifies removal. With two layers, that decision passes to officials the President cannot remove merely for disagreeing. A President who believes a Board member is neglecting the job then holds no lever at all, unless the commissioners’ own refusal to act is so unreasonable that it amounts to neglect or malfeasance on their part. That is why the Court said the President “cannot delegate ultimate responsibility or the active obligation to supervise that goes with it,”75 and why Congress cannot provide that an inferior officer can be removed only for cause by an officer who can be removed by the President only for cause.76
Behind the mechanism lies a point about accountability. The Court has described management of the Executive Branch as a task for which “imperative reasons requir[e] an unrestricted power” in the President to remove the most important subordinates in their most important duties.77 The President can always choose to restrain himself in dealing with subordinates, but he cannot bind his successors by diminishing their powers, and he cannot escape responsibility for his choices by pretending they are not his own.78 No one doubts Congress’s power to create a vast and varied federal bureaucracy. The majority asked where, in such a structure, the role is for oversight by an elected President, and it added that convenience and efficiency are not the hallmarks of democratic government.79
Justice Breyer dissented. He reasoned that neither the Necessary and Proper Clause nor the separation of powers is absolute in removal cases, so that Congress may sometimes limit removal depending on the nature, function, or subject matter of the office, and he listed thousands of inferior officers whose insulation from removal the decision seemed to put in jeopardy.80 The majority answered that the Sarbanes-Oxley Act was highly unusual in committing substantial executive authority to officers protected by two layers of good-cause removal.81 It also declined to use the occasion to invalidate independent agencies such as the SEC.82 The majority’s holding is the governing rule, and the dissent supplies the counterargument that its logic may reach beyond the facts.
The remedy shows exactly where the line falls. The Court did not strike down the Act. It severed the unconstitutional removal provisions from the rest of the statute,83 which converted the SEC’s power over Board members into at-will removal authority.84 The result is the configuration the earlier cases approved: a Commission removable by the President only for cause, supervising Board members whom the Commission can dismiss whenever it likes, so that the President can hold the Commission responsible for the Board as for the rest of its work. A single layer is where the statute survives, and two layers are where it fails.
Layering concerns how many tenure shields stand between the President and an officer who exercises executive power. A second line of cases concerns agencies headed by one person. In Seila Law LLC v. Consumer Financial Protection Bureau, the agency was the Consumer Financial Protection Bureau, led by a single Director. The Court began from the premise that, as a general rule, members of agencies are subject to the President’s unlimited removal power so that the President can “take Care that the Laws be faithfully executed,” and it recognized only two exceptions.85 The first is Humphrey’s Executor, under which an independent agency headed by a multimember board with quasi-legislative or quasi-judicial authority could be isolated from the President with for-cause protection.86 The second covers inferior officers with limited duties.87
The Consumer Financial Protection Bureau’s Director fit neither exception. A single Director is neither a body of experts nor nonpartisan, and the Director held significant rulemaking, adjudicatory, and enforcement authority, so the office was neither a mere legislative or judicial aid nor an inferior position with limited duties.88 The Court invalidated the removal restrictions, which made the Director removable by the President at will.89 The logic can be traced link by link. The Humphrey’s exception rests on a body that does not wield substantial executive power, so shielding it costs the President nothing Article II protects. An agency with substantial executive power removes that premise. A multimember body at least spreads its authority among several members, while a single director holds all of it for a fixed period, and neither the President nor anyone answerable to the President can remove the director over a disagreement about how the power is used.
The common thread with Free Enterprise Fund is an officer who exercises significant executive power and sits beyond meaningful presidential control, whether because two layers of tenure intervene or because one person leads the agency alone. Collins v. Yellen applied Seila Law to an agency led by a single Director, the Federal Housing Finance Agency. Justice Sotomayor, writing separately, described the controlling feature this way: “[A]ll that matters is that [the FHFA] is an agency led by a single Director.”90 The description comes from a partial dissent, but it indicates that the single-head feature did most of the work.
Because both agencies had one head, neither case resolved whether the Humphrey’s exception for multimember agencies survives. A federal court of appeals considering the multimember Consumer Product Safety Commission noted that Seila Law “cast doubt” on agencies like the Commission and that the logic of Humphrey’s may have been overtaken, but it observed that the decision has not been overruled, “at least not yet.”91 That is a lower-court decision, so it shows how courts are handling a multimember agency and states no holding of the Supreme Court.
Administrative law judges present a third question the Supreme Court has not resolved. A court of appeals held that two layers of insulation improperly impede the President’s power to remove ALJs based on how they exercise their statutory discretion in deciding cases, and the Supreme Court granted certiorari but affirmed on other grounds without ruling on the issue.92 The contrary argument is that Congress recognized that the decisional independence of ALJs could conflict with the President’s duty to take care, and removed that risk by empowering the agency to 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 all agency decisions.93 An essay on ALJs should present the unsettled question and both arguments, because no Supreme Court holding answers it.
The first two statutes fail on different facts, the number of layers and the number of heads, and the third shows that removing one layer ends the layering problem. In the first two, a court would likely sever the protection and leave the body in operation, as the Court did in both decisions.94
Step two separates the Fuel Review Panel from the Pipeline Review Panel, because the President’s ability to act on a judgment of dereliction survives a single layer and disappears with a second. Step four separates the single-head cases from the multimember agency in the court of appeals decision, and step five is where an answer should stop asserting and start weighing, since the Supreme Court has not resolved those questions.