Concept
The Take Care Clause
As chief executive, including the “take care” clause
The Take Care Clause
Article II, Section 3 ends its list of presidential responsibilities with the clause this section studies: the President “shall take Care that the Laws be faithfully executed, and shall Commission all the Officers of the United States.”1 The same sentence tells the President to report on the state of the Union and recommend measures to Congress, allows the President to convene the houses on extraordinary occasions, and requires the President to receive ambassadors and other public ministers.2
The clause differs in kind from the Vesting Clause. Section 1 grants power (“The executive Power shall be vested in a President of the United States of America”),3 and the dispute over how much that grant contains ran through the previous section. Section 3 imposes a duty, and the duty runs to laws that other actors have made. A claim of presidential power that rests on the Take Care Clause must therefore be squared with the content of the laws the President is bound to carry out.
The same words also support supervision. The President cannot personally carry out every federal statute, and Section 3 pairs the duty to see that the laws are executed with the power to commission the officers through whom they are executed.4 Because Article II vests executive power in the President and imposes this duty, agency execution of the laws is subject to presidential supervision, though how far that supervision goes is disputed.5 The clause thus restrains the President by tying the office to laws Congress has written, and it empowers the President by justifying oversight of those who carry the laws out.
Three questions organize what follows. What does the duty require, beginning with the rule that the President executes law and does not make it? How far does “the Laws” reach: to statutes and treaties alone, or to duties that flow from the Constitution itself? And how much control over subordinate officers does the clause imply?
Execute, Not Make, Law
The verb in the clause carries its limit. To execute a law is to carry out something that already exists, so the clause assumes a lawmaker other than the President.6 The Court enforced that assumption in Youngstown, where the Government defended the seizure of the steel mills by pointing to several constitutional provisions that grant executive power to the President. Justice Black, for the majority, answered that “the President's power to see that the laws are faithfully executed refutes the idea that he is to be a lawmaker.”7
Black's reasoning has two steps. First, the Constitution limits the President's part in lawmaking to recommending laws the President thinks wise and vetoing laws the President thinks bad, and it is “neither silent nor equivocal about who shall make laws which the President is to execute.”8 Second, the seizure order looked like a statute. Its preamble gave reasons why certain policies should be adopted, proclaimed those policies as rules of conduct, and authorized an official to issue further rules and regulations to carry them into execution.9 The difference, in Black's words, was that the order “does not direct that a congressional policy be executed in a manner prescribed by Congress—it directs that a presidential policy be executed in a manner prescribed by the President.”10 That sentence supplies the working question for any directive: whose policy does it carry out?
Congress had also spoken. It had refused to adopt seizure as a way of settling labor disputes and had provided instead for mediation, conciliation, investigation by boards of inquiry, and public reports, with temporary injunctions available for cooling-off periods. If all of this failed, the unions were free to strike after a secret vote.11 The order therefore replaced a congressional choice with a presidential one.
Justice Douglas, concurring, framed the same limit in terms of constitutional structure. The Court could not sanction the seizures, he wrote, “without reading Article II as giving the President not only the power to execute the laws but to make some,” a step that “would most assuredly alter the pattern of the Constitution.”12 A later opinion compressed the holding of Youngstown into one sentence: the President may not execute and exercise legislative authority belonging only to Congress.13
Justice Frankfurter, concurring, relied on a description of the duty that Justice Holmes had given in Myers: “The duty of the President to see that the laws be executed is a duty that does not go beyond the laws or require him to achieve more than Congress sees fit to leave within his power.”14 The sentence has two edges. As a limit, it denies the President any claim that the duty itself justifies an act no law authorizes. As a measure of responsibility, it holds the President answerable for achieving only what the laws, and the means Congress supplies, allow.
Justice Jackson explained why the limit matters to individuals. He matched the President's authority to execute the laws against the Fifth Amendment's guarantee that no person be deprived of life, liberty, or property without due process of law, and observed: “One gives a governmental authority that reaches so far as there is law, the other gives a private right that authority shall go no farther.”15 The mechanism is direct. When the executive acts against a person or that person's property, the person can ask what law authorizes the act. If the answer is a statute, the clause supplies the authority and the statute's terms measure it. If the answer is only the President's own preference, the clause supplies nothing, and the private right applies in full. One formulation of the rule describes the President as bound to execute the laws “by the means and in the manner which the laws themselves prescribe.”16
Medellín v. Texas applied the limit where no statute was in play at all. The President directed state courts to give effect to a judgment of the International Court of Justice, although the treaties underlying the judgment were non-self-executing. A non-self-executing treaty is one ratified on the understanding that it will not have domestic effect of its own force, and that understanding, the Court said, precludes the claim that Congress has implicitly authorized the President, acting alone, to achieve the same result.17 Such a treaty becomes domestic law only in the way any other law does, and the governing principle is that the power to make the necessary laws is in Congress and the power to execute them is in the President.18 The Government did not rely on the Take Care Clause for the memorandum, and the Court called that a wise concession, because the clause allows the President to execute the laws, not make them.
The two decisions fail the President for different reasons. In Youngstown a law on the subject existed and pointed the other way. In Medellín no domestic law existed for the courts to apply, and the memorandum would have created it. In both cases the directive was itself the source of the duty it imposed, and the Take Care Clause had no statute to attach to.
A hypothetical shows the same line in a domestic setting. Suppose a President, impatient with a stalled bill, orders that no employer may require overtime from a worker and directs the Attorney General to seek injunctions against violators. No statute imposes the ban. The order is the only source of the duty, so the President has made the rule and ordered its enforcement. A textual objection says that Article I vests all legislative power in Congress and so leaves none for the executive. A structural objection says that the President would be aggrandizing the executive by taking a power that belongs in the legislature, and Madison, quoting Montesquieu in Federalist No. 47, supplied the principle behind both: there can be no liberty where the legislative and executive powers are united in the same person or body of magistrates.19 If Congress had enacted the identical ban, the same instruction to the Attorney General would be unobjectionable, because enforcing an enacted law is what the clause contemplates. The enforcement instruction is the same in both versions; what differs is whether a statute stands behind the duty enforced. An answer that defends the order by citing the Take Care Clause assumes the very law that is missing.
None of this strips the clause of constructive force. A President charged with seeing that the laws are executed may instruct the officers who execute them, so a directive telling agencies how to carry out an existing statute falls within the clause. It stays within the clause because the duty it creates falls on subordinates and the policy it carries out belongs to Congress. The statute behind the directive also fixes its reach, since the Executive's administrative activity cannot go beyond the limits of the statute that created it,20 and courts may review executive action for compliance with the law.21 How far the supervisory side extends is the subject of the third subsection below.
Step two does the real work. A response that stops at the observation that the President is charged with executing the laws has assumed the law whose existence the problem puts in doubt, and the next subsection shows how Congress's position then fixes the strength of the President's remaining claim.
Reach of “The Laws”
The clause directs the President to see that “the Laws” be faithfully executed and does not say which laws it has in mind.23 The previous subsection showed that the President executes laws and does not make them. The question now is how large the body of law is that the President must execute, and whether the duty carries an implied authority to act when no statute supplies the means. A narrow answer confines “the Laws” to acts of Congress and treaties, read according to their express terms. A broad answer adds duties that grow out of the Constitution itself, out of international relations, and out of the nature of the government. Three authorities support the broad reading: a decision about guarding a federal judge, an episode in a railroad strike, and John Marshall's argument in the Jonathan Robbins affair.
The first is re Neagle (also known as Cunningham v. Neagle). A federal officer was guarding a Justice of the Supreme Court who was riding circuit, and no specific statute authorized the President to assign such a guard. The Court held that the officer had acted in line of duty and, in holding that no statute was necessary, framed the question this way: “Is this duty limited to the enforcement of acts of Congress or of treaties of the United States according to their express terms, or does it include the rights, duties and obligations growing out of the Constitution itself, our international relations, and all the protection implied by the nature of the government under the Constitution?”24
A statute-only reading fails on those facts for a practical reason. Congress cannot list in advance every threat to the government's ability to function, and a duty that stopped at express statutory terms would leave the President watching a federal judge be attacked until Congress anticipated the danger. Treating protection of the federal courts' work as part of faithful execution lets the President answer a threat that no statute foresaw. Counting the Constitution among “the Laws” has a second consequence: the duty reaches constitutional law as the Supreme Court interprets it, together with the statutes Congress passes to protect civil rights and civil liberties, so the executive's role in protecting those rights grows out of the congressional protection of individual liberty.25
The railroad-strike episode extends the idea to a statute-based setting. In the account Chief Justice Vinson gave in his Youngstown dissent, no statute authorized the President's action and the Governor of Illinois disclaimed any need for help, but the President's concern was that the federal laws on interstate commerce and the mails be continuously and faithfully executed. His agents obtained the injunction that the Court upheld in In re Debs, and the Court scrutinized each step the President took and gave his conduct full approval.26
The Robbins affair involved a treaty. When the House of Representatives debated the surrender of a prisoner under an extradition treaty, no statute prescribed the method of surrender, and John Marshall, then a member of the House, answered: “The treaty, which is a law, enjoins the performance of a particular object. The person who is to perform this object is marked out by the Constitution, since the person is named who conducts the foreign intercourse, and is to take care that the laws be faithfully executed. . . . Congress, unquestionably, may prescribe the mode, and Congress may devolve on others the whole execution of the contract; but, till this be done, it seems the duty of the Executive department to execute the contract by any means it possesses.”27 The Myers majority pointed to Marshall's argument as one the Court had approved in Fong Yue Ting v. United States, and in the same passage it added that where the discretion to be exercised is the President's own, cabinet officers must do the President's will.28 That second point belongs to the next subsection.
In all three authorities a body of law already existed (the operation of the federal courts, the statutes on commerce and the mails, a treaty), and what the President supplied was the means of carrying it out. Marshall's own sentence builds in the limit, since Congress may prescribe the mode and the duty to use available means lasts only until it does. Justice Story reasoned the same way in his dissent in Brown v. United States. No act of the legislature defined the mode of warfare, and he concluded that the executive should be governed by the law of nations as applied to a state of war, with the legislature free to narrow and limit the rights of war, and the executive holding all the rights of modern warfare “until such limit is assigned.”29 Both statements describe a default rule that fills a gap in the means and that Congress can displace. That feature explains why the broad reading is less alarming than it first sounds, and why Congress's position decides how far it carries.
Chief Justice Vinson tried to stretch the broad reading to the steel seizure. He argued that the President had acted only to preserve the legislative programs from destruction until Congress could act, and he stressed that the President had told Congress he would abide by whatever it did.30 The majority rejected the seizure, and the contrast with Neagle shows where the stretch gave way. In Neagle the President protected the working of the federal judiciary and no statute stood in the way. In Youngstown the President took private property to advance his own view of how to head off a labor dispute, in a field where, as the previous subsection described, Congress had chosen other procedures.31 A protective step for an existing federal function passes the test, while a measure that reaches private property to carry out a presidential policy does not.
The narrower view runs through Black's and Jackson's opinions in Youngstown and through Medellín. It accepts that statutes and treaties are “the Laws” and that the President chooses the means within them, and it resists using the clause to carry out the President's own policy. Marshall began from a treaty that was itself a law enjoining a particular object, so the President supplied only the means. The judgment in Medellín was not domestic law, so there was nothing for the clause to execute until Congress acted. Both lines of authority start from existing law. They part over whether the President's judgment about what protects or advances that law can authorize action against private parties when Congress has taken a different course.
The historical edge of the argument lies in the Civil War. Lincoln relied in part on the Take Care Clause when, at the outbreak of the war, he suspended habeas corpus, blockaded southern ports, and called for soldiers without congressional permission. His opponents insisted that the framers were committed to limiting the powers of every branch, and that no single person should decide whether to suspend habeas corpus or send troops abroad.32 Lincoln replied that the laws were being resisted in nearly one-third of the states and asked whether all the laws but one were to go unexecuted and the Government itself go to pieces lest that one be violated. Chief Justice Taney, sitting on circuit in Ex parte Merryman, read Article II in the opponents' spirit: its grants are in “precise and plain words,” its short term and narrow limits show the framers' jealousy of executive power, and they withheld from the President many powers of the English executive and conferred only those deemed essential.33 He held that only Congress could suspend the writ and, unable to enforce his order, sent a copy of his opinion to the President so that the President, in fulfilment of the obligation to take care that the laws be faithfully executed, could decide how to see the court's process respected. Taney thus used the clause against the detention, because the laws the President must execute include the protections the military order had set aside. Congress eventually ratified the suspension of the writ by legislation enacted on March 3, 1863, nearly two years after Lincoln first suspended it.34
A narrow power to protect federal personnel, property, and instrumentalities survives even for those who reject any general presidential emergency power. The argument is that debate over presidential emergency power has obscured a narrower, inherent power to protect and defend the personnel, property, and instrumentalities of the United States from harm; the Executive may, for example, enforce the government's own contract or property rights without statutory authority.35 The Neagle guard fits that description. The line falls there for two reasons. A reading that stopped at express statutory terms would leave the government unable to protect its own officers and functions against threats no statute anticipated. A reading that treated any asserted necessity as a source of power would have no boundary except the President's own judgment of what the necessity requires. The protective power avoids both problems by asking what is being protected and whether Congress has spoken to the problem.
Presidential Control of Execution
The clause creates a duty to see that the laws are executed, which differs from a duty to execute each of them personally, and that difference is the basis for presidential control. No President can carry out every federal statute alone, and Section 3 pairs the duty with the power to commission the officers of the United States.37 Because Article II vests executive power in the President and imposes this duty, agency execution of the laws is subject to presidential supervision, though how far that supervision goes is disputed.38 Section 1 points the same way, since it vests the executive power in a single President.39
The dispute is about what “executive power” contains. Corwin called it a “term of uncertain content,” and the uncertainty is sharpest in debates over executive control of administrative agencies. One version of the unitary executive theory gives the President almost complete discretion in exercising executive power within the executive branch itself, and broader versions identify certain actions as executive in character whether they occur inside the branch or beyond it.40
The most direct form of control is instruction. In Myers v. United States the Court wrote that the highest and most important duties of the President's subordinates are those in which they act for him, so that they exercise his discretion and not their own.41 Where the discretion to be exercised is the President's, in determining the national public interest and directing subordinates to protect it, cabinet officers must do the President's will.42 The Myers majority also quoted earlier language describing the executive departments as aids that represent the President in a thousand acts to which his personal attention can hardly be called, which is how he fulfills the duty “expressed in the phrase that ‘he shall take care that the laws be faithfully executed.’”43
Myers itself marked the limit of instruction: the degree of guidance the President may exercise over executive officers varies with the character of their service as prescribed in the law under which they act.44 At one end, the law makes the officer's discretion the President's, and direction is complete. At the other, Congress fixes the officer's duty, and the law controls. Kendall v. United States shows that second end. Congress had directed the payment of an arbitrated award, and the Postmaster General refused to pay it on the ground that only the President could direct him. The Court rejected the idea that every officer of the executive department stands under the exclusive direction of the President, and it held that a duty Congress imposes on an officer comes from the law.
A hypothetical applies the line. Suppose a statute directs a department head to pay any claim that an arbitrator has approved, and the President instructs the department head to withhold payment on one approved claim. The instruction does not guide discretion the statute leaves open. It directs a violation of a duty the statute imposes, and the Take Care Clause, whose standard is faithful execution, cannot authorize it. Now suppose the President directs every executive agency to conduct a cost-benefit analysis of each proposed rule before it takes effect. The directive reaches officers inside the executive branch, instructs them in applying statutes Congress already enacted, and creates no binding duty for private parties, so it falls on the supervisory side of the clause. The live dispute in such cases is whether a particular directive asks officers to do something a statute forbids, and a directive that asks only for action to the extent the law permits guides discretion and does not override a duty.
Control over execution also has a structural side. Printz v. United States held that Congress could not require local law enforcement officers to run background checks on handgun buyers under the Brady Act. Most of the opinion rests on federalism, but a shorter passage speaks to the separation of powers among the federal branches. The Constitution, the Court said, does not leave to speculation who is to administer the laws Congress enacts: the President takes care that they be faithfully executed, personally and through officers he appoints. The Brady Act transferred that responsibility to thousands of state officers who would carry out the program without meaningful presidential control, if such control is possible at all without the power to appoint and remove. The Court tied the point to the Framers' insistence on unity in the executive, adopted to secure vigor and accountability. That unity “would be shattered, and the power of the President would be subject to reduction, if Congress could act as effectively without the President as with him, by simply requiring state officers to execute its laws.”45
The mechanism is easy to trace. A statute needs someone to enforce it, and if Congress may choose enforcers whom the President cannot appoint, instruct, or remove, then Congress can complete the whole cycle of making and executing a law while the President's duty to take care has nothing to attach to. The cost-benefit directive reaches officers the President nominates, instructs, and can hold to account, while the Brady Act reached officers over whom the President had no such hold. Both arrangements involve the execution of federal law, so the difference lies entirely in the President's relationship to the officer who executes it. Two cautions keep the point in proportion. The passage sits inside a federalism decision and was made in passing, so it is better cited as a statement of structure than as a freestanding holding about the Take Care Clause. And the same logic raises a harder question about officers whom the President cannot freely remove.
That question leads into removal, where the Court has declined to treat either principle as absolute. The Necessary and Proper Clause does not let Congress free all executive officials from dismissal at the President's will, and the separation of powers does not give the President an absolute authority to remove any and all of them. Depending on the nature of the office, its function, or its subject matter, Congress sometimes may limit the President's removal authority.46 The Court has accordingly permitted Congress to vest the removal power in others and to limit the circumstances under which certain officials may be removed, but not every restriction is valid.47 When a statute does restrict removal, Congress can enforce the restriction by providing for judicial review of the removal decision, and because review only ensures that an officer is removed in accordance with the statute, it does not put any additional burden on the President's exercise of executive authority.48
The Court gave the supervisory reading its clearest modern application in Free Enterprise Fund v. Public Company Accounting Oversight Board. The members of the Board, which regulated the accounting industry, were removable by the Securities and Exchange Commission only for cause, and the Commissioners were themselves protected by good-cause tenure, so two layers of for-cause protection separated the President from the Board. The Court held that this multilevel protection is contrary to Article II's vesting of the executive power in the President, because the President cannot “take Care that the Laws be faithfully executed” if he cannot oversee the faithfulness of the officers who execute them. The reasoning runs link by link. Deciding whether an officer is neglecting or misusing the job is the core of oversight, and the statute committed that decision to another officer, who might disagree with the President and whom the President could not remove merely for disagreeing. The Court also saw that the arrangement could be multiplied: “If Congress can shelter the bureaucracy behind two layers of good-cause tenure, why not a third?” The Government was unwilling at oral argument to concede that even five layers would be too many.49
The concern is accountability. When the President cannot control those who execute the laws, the public cannot tell on whom the blame for a bad measure should fall. A President answerable to a national electorate is also well placed to centralize and coordinate the regulatory process, and less likely than a single-mission agency to be myopic.50 The supervisory reading thus leads straight to the question Myers framed: whether the President has the exclusive power to remove the executive officers whom he has appointed.51 The clause supplies the reason for control without settling how much control the Constitution demands over each kind of officer, and the later sections on appointment and removal take up that question.
Step two separates Printz from the cost-benefit directive, and step three separates Myers from Kendall. The fourth step is where this subsection hands off to the next major section, because the Court's statements about oversight in Printz and in Free Enterprise Fund both end by pointing to appointment and removal as the tools that make supervision real.