Concept
Youngstown and Presidential Power
As chief executive, including the “take care” clause
Youngstown and Presidential Power
Article II names the President’s powers without saying how they combine with what Congress has done. The Vesting Clause provides that “[t]he executive Power shall be vested in a President,” and the Take Care Clause directs the President to “take Care that the Laws be faithfully executed.”1 The earlier sections showed how far each clause can be read in either direction, but neither tells a court how heavily to weigh a presidential claim when Congress has already authorized the act, forbidden it, or said nothing.
Youngstown Sheet & Tube Co. v. Sawyer, the Steel Seizure Case, supplies the method. The Court held that the President could not seize the nation’s steel mills to prevent a strike, and Justice Black’s opinion rested on the point that the President executes the laws and does not make them. The reasoning that courts and students use most, however, comes from Justice Jackson’s concurrence. It sorts presidential action by the position Congress has taken, because, as Jackson put it, “[p]residential powers are not fixed but fluctuate, depending upon their disjunction or conjunction with those of Congress.”2 Later cases have followed the analysis, although it is better understood as a spectrum than as three sealed boxes.3
| Category | What Congress has done | Strength of the President’s claim |
|---|---|---|
| First | Authorized the act, expressly or impliedly | Maximum: the President’s own powers plus all that Congress can delegate4 |
| Second | Neither granted nor denied authority | The President’s independent powers alone, with a zone of twilight at the margin5 |
| Third | Taken a position incompatible with the act, expressly or impliedly | Lowest ebb: the President’s own powers minus any powers Congress holds over the subject6 |
The table holds the President’s own Article II powers constant and varies only what Congress adds or takes away. On any problem about domestic presidential power, ask first what Congress has done. The subsections below begin with the case, take the three situations in turn, and end with the weight that long executive practice carries in reading Article II.
The Steel Seizure Case
The Government defended the seizure of the steel mills with Article II alone. No express constitutional language grants the power, so the Government asked the Court to imply it from the aggregate of the President’s powers, relying on the Vesting Clause, the Take Care Clause, and the Commander in Chief Clause.7 The mill owners answered that the seizure order was lawmaking, a function the Constitution confides to Congress. The Government replied that the President had found the order necessary to avert a national catastrophe, and that he held “inherent power” to meet the emergency.8
The facts are short. President Truman issued Executive Order 10340, which directed the Secretary of Commerce to take possession of the steel mills, and the companies obeyed under protest and sued the Secretary.9 The district court issued a preliminary injunction on April 30, the court of appeals stayed it the same day, and the Supreme Court granted certiorari on May 3 and announced its decision on June 2, 1952.10 The Court held that the seizure order could not stand.11
Black’s opinion asked first whether any statute authorized the seizure, and none did. Two statutes, the Selective Service Act of 1948 and the Defense Production Act of 1950, authorize the President to take property under stated conditions. The Government admitted, though, that those conditions were not met and that the order did not rest on either statute, and it called the seizure provisions of the 1950 Act “much too cumbersome, involved, and time-consuming for the crisis which was at hand.”12 That admission left the order standing on the President’s own constitutional powers. An order issued under one of the statutes, with its conditions satisfied, would have drawn on Congress’s authority as well, which is the situation Jackson’s first category describes.
Congress had also considered seizure and withheld it. When it took up the Taft-Hartley Act in 1947, it rejected an amendment that would have authorized governmental seizures in emergencies, and the plan it adopted did not provide for seizure under any circumstances.13 Apparently Congress thought that seizure, like compulsory arbitration, would interfere with collective bargaining.14 An order that used the method Congress had refused put the President’s judgment in place of a choice the legislature had already made, which is why the order looks like lawmaking and not like execution.
The Commander in Chief Clause could not carry the order either. Black set aside the Government’s decisions about military commanders directing the fighting in a theater of war as not controlling, and he declined to hold that the Commander in Chief has the ultimate power as such to take possession of private property in order to keep labor disputes from stopping production.15 He put the conclusion in a sentence: “This is a job for the Nation’s lawmakers, not for its military authorities.”16
The Vesting and Take Care Clauses fared no better. The President’s power to see that the laws are faithfully executed refutes the idea that he is a lawmaker. The Constitution limits his part in lawmaking to recommending laws he thinks wise and vetoing laws he thinks bad, and it is neither silent nor equivocal about who makes the laws he must execute.17 The order itself read like a statute: its preamble stated reasons for a policy, it proclaimed rules of conduct, and it authorized an official to issue further regulations. A President who announces his own policy and the rules for carrying it out has made law and has not executed Congress’s. Congress’s power to adopt such policies, to authorize the taking of private property for public use, and to regulate relations between employers and employees was, Black wrote, beyond question, and the Constitution does not subject it to presidential or military supervision.18
Two lines of argument support the result. A textual argument says that, by vesting “all” legislative power in Congress, Article I precludes the executive from exercising the lawmaking power. A structural argument says that the President has sought to aggrandize the executive branch by usurping legislative power, and it draws on Madison’s warning, quoting Montesquieu, that “[t]here can be no liberty where the legislative and executive powers are united in the same person, or body of magistrates.”19
Black gave the emergency no weight as a source of power. The Founders entrusted the lawmaking power to Congress alone in both good and bad times, he wrote, and a review of the historical events and the fears of power behind that choice would only confirm the holding.20 If a grave occasion could supply missing authority, the President would decide for himself when a crisis was serious enough, and Congress’s check would last only until he found an urgent reason to bypass it.
Chief Justice Vinson, joined by Justices Reed and Minton, dissented on the ground of emergency. Much of the argument against the President, he said, was directed at straw men: the case did not involve a President acting solely on his own notions of the public welfare, and it raised no question of unlimited executive power, because the President had closed the door to that claim by telling Congress that he would abide by whatever action it took.21 On the dissent’s account, the President acted to preserve the legislative programs for military procurement and price stabilization from destruction until Congress could act, and the possession was temporary and subject to congressional direction.22 Vinson added that Congress and the courts had responded to such executive initiative with consistent approval, and that the facts showed no arbitrary action or dictatorial usurpation of congressional power.23 He also noted the diversity of views in the six opinions of the majority, which is one reason the concurrences, Jackson’s above all, matter as much as Black’s opinion for later problems.24
Black’s opinion and the concurrences differ in method, and essays on the case reward noticing it. Black’s opinion is generally read as the more formal approach: it asks whether the claimed power comes from a statute or a provision of the Constitution, and it reads the executive power as the power to carry out laws that Congress has made. Frankfurter and Jackson took a more flexible, functional approach, which commentators defend on the ground that a modern regulatory government, able to regulate complex industries, could not operate under a rigidly formalist separation of powers.25 The formal approach offers a bright line and pays for it in flexibility. The functional approach lets the answer vary with what Congress has done, and that variable is the subject of the next three subsections. Neither approach, however, saved the seizure.
A complete answer on the steel seizure therefore assembles four facts. No statute authorized the seizure, and the two statutes that permit takings were conceded not to apply. Congress had considered a seizure remedy in Taft-Hartley and withheld it. No clause of Article II names the function, and the Commander in Chief Clause does not reach the taking of private property to settle a labor dispute. And the order itself announced policy and rules in the President’s name, which is the work of a legislature.
Maximum Authority
Justice Jackson’s first category covers the President who acts with Congress behind him: “When the President acts pursuant to an express or implied authorization of Congress, his authority is at its maximum, for it includes all that he possesses in his own right plus all that Congress can delegate.”27 The President’s own powers, such as the vesting and take-care powers of Article II, form the base, and Congress adds whatever the Constitution lets it hand over, so the total is the most the executive can claim.28 The usual separation-of-powers objection also loses its target. That objection says the President has taken a function that belongs to another branch, but here the branch that holds the lawmaking function has approved the action. Jackson therefore wrote that a seizure executed under an Act of Congress would be “supported by the strongest of presumptions and the widest latitude of judicial interpretation,” with the burden of persuasion resting “heavily upon any who might attack it.”29
The same reasoning explains what a loss in this category means. Because the President is using his own powers and every power Congress could delegate, a holding that the act is unconstitutional leaves no federal power to supply it. Jackson put the point this way: the result “usually means that the Federal Government as an undivided whole lacks power.”30 A challenger who wins here therefore usually relies on a limit that binds Congress as well as the President, such as an individual right, and cannot rest on a complaint about which branch acted.
Authorization need not be express, because Congress seldom lists every step an executive must take. A dissent in Hamdan v. Rumsfeld drew on earlier decisions to argue that Congress “cannot anticipate and legislate with regard to every possible action the President may find it necessary to take,” and that its failure to do so does not, “especially . . . in the areas of foreign policy and national security,” imply congressional disapproval.31 Because that reasoning comes from a dissent, it shows how the first-category argument is made and does not state the Court’s holding.
Ex parte Quirin shows a President in the first category. The Court described the President’s order creating a military commission as an exercise of “the authority conferred upon him by Congress, and also such authority as the Constitution itself gives the Commander in Chief, to direct the performance of those functions which may constitutionally be performed by the military arm of the nation in time of war.”32 From that combination the Court drew its standard of review: a detention and trial the President orders in the declared exercise of his powers as Commander in Chief in time of war and grave public danger are not to be set aside “without the clear conviction that they are in conflict with the Constitution or laws of Congress constitutionally enacted.”33 That deference is the first category’s presumption at work.
An authorization carries the President’s action only as far as its terms reach, and Hamdan shows the limit. Applying Jackson’s framework, the dissent read the Authorization for Use of Military Force, which permits “all necessary and appropriate force” against those responsible for the September 11 attacks, to bring the military commission within Congress’s grant, so that the President’s choice of a commission was “entitled to a heavy measure of deference.”34 The Court disagreed. It concluded that the commission departed from limits Congress had set on military commissions, and an order that exceeds the terms of a grant cannot draw on the grant. It leaves the first category, and the President must defend it on his own powers alone. Both cases involve a military commission and a statute that mentions commissions. What differs is whether the President’s order fit the limits Congress attached, and that fit is the variable the first category turns on.
The first category also marks a boundary for precedent. Jackson eliminated the steel seizure from it because the Government conceded that no congressional authorization existed, and that concession removed “the support of the many precedents and declarations which were made in relation, and must be confined, to this category.”35 An answer that leans on Quirin to defend an act Congress never authorized has borrowed authority from the wrong category. The categories are also better read as a spectrum, because congressional support can be strong or faint and a court must decide how much of it the facts show.36
Step three does the decisive work. Quirin passes it because the order drew on authority Congress had conferred, and Hamdan fails it because the commission departed from the limits Congress set. Step two comes first for a reason, since an order with no statute behind it never reaches the first category, as the steel seizure never did.
The Zone of Twilight
Justice Jackson’s second category asks what Article II can accomplish when Congress has neither authorized nor forbidden the act: “When the President acts in absence of either a congressional grant or denial of authority, he can only rely upon his own independent powers, but there is a zone of twilight in which he and Congress may have concurrent authority, or in which its distribution is uncertain.”38 The sentence has two halves. The first limits the President’s claim, and the second admits that the limit blurs at the edges.
The limit is that the President brings no borrowed authority. In the first category his own powers were joined by everything Congress could delegate, but here Congress has delegated nothing, so the claim stands or falls on Article II alone. An argument that the Vesting Clause is a bulk grant, or that the Take Care Clause lets the President pursue a policy of his own, must succeed on its own terms.39 The base is not empty. Jackson refused to narrow the powers Article II does mention by a “niggardly construction,” and he gave them the scope that “reasonable, practical implications” afford in place of “a doctrinaire textualism.”40 Even so, nothing from Congress adds to the President’s footing.
The zone of twilight is the second half of the sentence. “[C]ongressional inertia, indifference or quiescence may sometimes, at least as a practical matter, enable, if not invite, measures on independent presidential responsibility,” and any actual test of power “is likely to depend on the imperatives of events and contemporary imponderables rather than on abstract theories of law.”41 Two reasons support so loose a standard. The first is structural: the Constitution “diffuses power the better to secure liberty” but also contemplates that practice will integrate the dispersed powers into a workable government, and the Founders did not expect the separation of powers to be airtight, so the branches often share functions.42 The second is practical. Congress cannot anticipate every situation the President may face, and a dissent in Hamdan quoted the observation that Congress’s failure to legislate for every contingency does not, especially in foreign policy and national security, imply congressional disapproval.43 That argument comes from a dissent in a foreign-affairs setting, so it shows how the generous reading of silence is made and does not state a rule for every domestic problem. A rule that denied the President all initiative until Congress spoke would leave the government unable to respond to anything new, since statutes are written in advance and events are not.
Because the standard turns on events and context, the second category poses a question and supplies no formula: what does Congress’s silence mean? Silence may reflect inattention, a willingness to leave a hard decision to the President, or a deliberate choice to withhold the power. The first two readings invite presidential action, and the third forbids it. The same absence of a statute can therefore land a case in the second category or, as in the steel seizure, in the third, where Jackson treated Congress’s silence and its refusal to grant seizure authority as deliberate restraint.44 Whether that was a plausible reading of silence is open to question, and the question is a good test of the second category’s limits.
The evidence that separates these readings is the surrounding legislation and Congress’s conduct toward similar executive acts, and in a close case acquiescence matters most. A “systematic, unbroken, executive practice, long pursued to the knowledge of the Congress and never before questioned” can “raise a presumption that the [action] had been [taken] in pursuance of its consent.”45 The reason is evidentiary. One instance of silence fits inattention as well as consent, but a long run of identical executive acts, each known to a Congress able to stop them, is hard to explain as inattention, so the silence begins to look like a choice. The last subsection of this chapter develops that argument as a gloss on the executive power.
Medellín v. Texas shows that the absence of a statute does not by itself put an act in the twilight zone. The President issued a memorandum directing state courts to reopen final criminal judgments in order to give effect to an international tribunal’s judgment, and the treaties behind the judgment were non-self-executing, meaning that the ratifying Senate understood them not to create domestic law without implementing legislation. The Court held that this character “not only refutes the notion that the ratifying parties vested the President with the authority to unilaterally make treaty obligations binding on domestic courts, but also implicitly prohibits him from doing so,” which placed the President’s reliance on the treaties in Jackson’s third category, “not the first or even the second.”46 The missing implementing legislation looks like silence, yet the Senate’s understanding at ratification expressed Congress’s implied will, so the field was not open.
The Government’s other argument relied on the independent authority recognized in the claims-settlement cases. Those cases rest on the view that a systematic, unbroken practice, long pursued to Congress’s knowledge and never questioned, can raise a presumption of consent, and the Court described the authority they support as “narrow and strictly limited”: the authority to settle international claims disputes by executive agreement.47 The memorandum had no such record behind it. The Court noted that the Government itself had called it “unprecedented action,” that the Government could identify no earlier presidential directive to state courts, and that this directive reached “deep into the heart of the state’s police powers” by compelling state courts to reopen final criminal judgments.48 The settlements and the memorandum shared a missing statute. What separated them was a long, known practice of one kind of act that made Congress’s silence look like consent, set against a first-of-its-kind order that gave Congress nothing to acquiesce in.
The steel seizure shows a silence that was not silence at all. Jackson first eliminated the seizure from the first category because the Government conceded that no congressional authorization existed.49 He then asked whether it could fit the second category and found that it could not: Congress, he wrote, “has not left seizure of private property an open field but has covered it by three statutory policies inconsistent with this seizure.”50 The record he relied on includes the takings statutes the Government conceded it had not satisfied and the seizure amendment Congress rejected in Taft-Hartley.51 The mechanism follows from what makes silence an invitation. Silence invites presidential action when it leaves a gap, but where Congress has built several routes to an end and the President takes another, its failure to list his route reads as selection and leaves no gap to fill. The seizure was therefore left “to be justified only by the severe tests under the third grouping,” which the next subsection takes up.52
Step two decides the steel seizure and the treaty-based argument in Medellín, which lost because Congress’s will pointed the other way. Step four decides the claims-settlement practice, which succeeded on a record of acquiescence, and the gloss argument in Medellín, which failed for want of one.
Lowest Ebb
The third category describes a President who acts against Congress. In Jackson’s words, the President’s power is then “at its lowest ebb,” because the course he takes is “incompatible with the expressed or implied will of Congress.”54 The formula is a subtraction. In the first category the President’s powers were added to everything Congress could delegate, and in the second they stood alone, but here Congress’s powers over the subject come off the total. The act succeeds only if something remains: Jackson said the steel seizure “can be supported only by any remainder of executive power after subtraction of such powers as Congress may have over the subject,” and that “we can sustain the President only by holding that seizure of such strike-bound industries is within his domain and beyond control by Congress.”55
Two reasons explain why the President’s position is weakest here. The first is the rank of a statute. A dissent in Zivotofsky v. Kerry argued that Congress may make laws necessary and proper for carrying the President’s powers into execution, that the President must take care that Congress’s legislation be faithfully executed, and that Acts of Congress made in pursuance of the Constitution are the supreme law of the land while acts of the President, apart from treaties, are not.56 That is a dissenting view of the case, but the structure it describes follows from the constitutional text. An order cannot outrank the statute it contradicts, and the duty to take care gives the President no authority to pick which valid statutes to execute.
The second reason concerns what the President must prove. Because Congress’s powers over the subject are subtracted, the President wins only if the subject lies within his domain and beyond Congress’s control, which means an exclusive power. Jackson noted that the Court’s first review of such seizures came in “the least favorable of possible constitutional postures” for that reason.57 A holding that a power is exclusive takes Congress out of the field, so it needs a firm footing in Article II. The route remains open. In Zivotofsky v. Kerry the Court relied on the President’s recognition power to hold parts of a statute unconstitutional, over a separate opinion’s objection that the Court had distorted that power.58 If the Constitution gives the President a power over the subject that Congress cannot reach, Congress has no power there to subtract, and the President’s position no longer sits at its lowest ebb.
The steel seizure shows the category at work. The previous subsection explained why the order could not be defended in the second category, since Congress had covered seizure with statutory policies the President passed over. The same record places the order in the third. Congress specifically rejected a proposal to empower the President to seize any plant, mine, or facility in which a threatened stoppage would, in his judgment, imperil public health or security, and it directed the President, if a strike was not settled within the 80-day injunction period, to submit a report with recommendations to Congress.59 The Act does not say in terms that the President may not seize, yet Frankfurter read Congress as having withheld the power “as though it had said so in so many words,” which he put as a message to the President: “You may not seize. Please report to us and ask for seizure power if you think it is needed in a specific situation.”60
That reading explains why the category speaks of an expressed or implied will. A legislature that weighs a method, rejects it, and builds a different procedure for the same problem has stated its position as plainly as a prohibition would. Burton reached the same conclusion: Congress had prescribed specific procedures, exclusive of seizure, and had reserved to itself the right to decide where and when seizure would be authorized, so the President’s order “invaded the jurisdiction of Congress.”61 Burton also left open the President’s power to meet an “imminent invasion or threatened attack,” which the steel dispute was not.62 The Court has applied the same logic outside seizure. In New York Times Co. v. United States, the President lacked authority to seek a judicial prior restraint against publication of classified information where Congress, in creating a system for classifying documents, had declined to give the executive that remedy.63
Clark’s concurrence isolates the variable Congress controls. In Little v. Barreme the President’s instructions ordered a broader seizure than a statute allowed, and the Court treated the statute as having prescribed how the law was to be carried into execution and as excluding the broader seizure.64 Clark drew a two-part rule: where Congress has laid down specific procedures for the type of crisis the President faces, he must follow them, but in the absence of such action by Congress his independent power to act depends upon the gravity of the situation.65 The gravity of an occasion therefore matters only where Congress has not prescribed its own method, and in Youngstown the emergency was the very one for which Congress had chosen procedures. Jackson’s closing reflection points the same way: power to legislate for emergencies belongs in the hands of Congress, though only Congress itself can prevent power from slipping through its fingers.66
The Medellín memorandum, discussed in the previous subsection, belongs here as well. The non-self-executing character of the treaties “implicitly prohibits” the President from creating domestic law on his own, so his claim, so far as it rested on the treaties, fell in the third category.67 Set beside Youngstown, Medellín shows that Congress’s will can be implied from something other than a statute aimed at the President’s act. There the opposing will came from the understanding behind the Senate’s ratification, and in Youngstown it came from a rejected proposal and a set of statutory procedures. In each case the President replaced a choice that Congress had made.
Step four decides most third-category problems, and the usual outcome is that the President loses, because a claim of exclusive power must be made out from Article II itself. The gravity of the occasion does not fill that step where Congress has prescribed its own method, since Clark’s rule applies the gravity test only in the absence of congressional action. Long practice does not fill it either, for Frankfurter, whose opinion gives executive practice its weight, added that deeply embedded ways of conducting government cannot supplant the Constitution or legislation.68 The next subsection takes up that argument and the limits Frankfurter placed on it.
Historical Practice as Gloss
Article II vests “the executive Power” in the President without listing what the phrase contains, and the gap invites an argument from history.70 Justice Frankfurter gave that argument its standard form in his Youngstown concurrence: “a systematic, unbroken, executive practice, long pursued to the knowledge of the Congress and never before questioned, engaged in by Presidents who have also sworn to uphold the Constitution, making as it were such exercise of power part of the structure of our government, may be treated as a gloss on ‘executive Power’ vested in the President by § 1 of Art. II.”71 A gloss is an explanatory note written beside a text. In this setting it means a settled understanding of what a constitutional phrase includes, built by repeated use and announced in no single opinion.
Frankfurter’s reason for giving practice that role ran through the structure of the Constitution. The Constitution is a framework for government, he said, so the way the framework has consistently operated fairly establishes that it has operated according to its true nature.72 He called it an “inadmissibly narrow conception of American constitutional law” to confine it to its words and disregard “the gloss which life has written upon them.”73 The point bears directly on the Vesting Clause debate earlier in this chapter. A sentence that names a power without defining it leaves room for practice to define it, and a gloss is how that happens.
The formula works as evidence about Congress’s silence. Congress can end an executive practice by legislating, so an act that Presidents repeat and Congress leaves alone begins to look like an act Congress accepts. Each element of the formula guards against a different way that inference can fail. A practice that is systematic and unbroken is uniform enough to count as one practice, and a long one has given Congress time to respond. The requirement that Congress knew of the practice matters because a legislature cannot consent by silence to something it never saw. The requirement that no one questioned the practice excludes a record in which an earlier objection shows the matter was contested. The reference to Presidents who swore to uphold the Constitution makes each repetition a separate judgment about constitutional authority and more than a convenience. When all of those conditions hold, the silence is hard to explain as inattention, and that is the sense in which practice counts as evidence of consent.
The practice that satisfied Frankfurter’s standard was the one in United States v. Midwest Oil Co. He counted it: over eighty years, in 252 instances, and by Presidents “learned and unlearned in the law,” lands that Congress had opened for entry were temporarily withdrawn so that Congress could deal with such withdrawals.74 The number and span make the practice systematic, the changing cast of Presidents makes it a series of separate judgments, and the withdrawals were temporary protection of property belonging to the United States, which left Congress free to act.75 The dissent reproduces a passage that describes the field in which such practice arises: “just as there are fields which are peculiar to Congress and fields which are peculiar to the Executive, so there are fields which are common to both, in the sense that the Executive may move within them until they shall have been occupied by legislative action,” because “the President is the active agent, not of Congress, but of the Nation.”76
Frankfurter held the Government’s seizure precedents to the same standard, and they fell short. He found that the list of executive seizures in circumstances comparable to the steel dispute “reduces to three” in the six months from June to December 1941, and that those “three isolated instances” could not add up, in number, scope, duration, or contemporaneous legal justification, to the kind of executive construction on which a gloss rests. They also came with no long-continued acquiescence by Congress.77 He added that no remotely comparable practice could be vouched for seizure of property at a time when the country was not at war in the only constitutional way in which it can be at war, and he thought it irrelevant to reopen the controversy over some acts of Lincoln during the Civil War.78 Set beside the oil-land withdrawals, the seizures lack every feature that made that record persuasive. Two hundred fifty-two instances over eighty years stand against three in six months, temporary protection of government-owned land stands against the taking of private industrial plants, and a practice Congress could watch and answer stands against acts with no acquiescence at all.
Chief Justice Vinson read the history the other way, and his account shows what the opposing view counts as practice. His dissent cites the 1941 occupation of Iceland as one of at least 125 incidents in which Presidents ordered the armed forces to act or hold positions abroad without congressional authorization and without a declaration of war.79 It also recounts later plant seizures, notes that seizures after the Smith-Connally Act rested on that Act as well as on the President’s power under the Constitution, and describes a dispute over whether the Act reached Montgomery Ward, a retail department store and mail-order concern.80 The list loses to Frankfurter’s test for two reasons. It gathers acts of many kinds under one heading, emergency action to protect the national defense, and the seizures after the Smith-Connally Act drew on statutory authority and so show what Congress authorized. Frankfurter asked a narrower question: whether comparable seizures, made without statutory authority, formed a long and unchallenged line.
The dissent’s account also shows how ambiguous congressional silence can be. In 1941 Senator Connally proposed an amendment authorizing the President to seize any plant where an interruption of production would unduly impede the defense effort. Its proponents did not suggest it would add to the President’s powers, opponents called it unnecessary because the President already had the power, and Congress did not approve it at that session.81 A legislature that declines to enact a seizure authorization may be withholding the power, which is how Frankfurter read Congress’s later handling of seizure in the Taft-Hartley Act.82 Silence supports a gloss only when it can be read one way.
Frankfurter placed a limit on the gloss in the same passage that states it: “Deeply embedded traditional ways of conducting government cannot supplant the Constitution or legislation, but they give meaning to the words of a text or supply them.”83 The limit follows from the job practice does. Practice is evidence of what an open phrase such as “executive Power” includes, and evidence adds nothing where a text or a statute already answers the question. If a long executive habit could override an enacted statute, Presidents could amend the law by repeating a violation often enough. The steel seizure shows the limit at work. Frankfurter wrote that the President could not have issued the order had Congress explicitly negated the authority in formal legislation, and he read the Taft-Hartley Act as telling the President, in effect, “You may not seize.”84 An exam fact pattern that recites a long tradition and asks whether it authorizes the act should therefore be checked first for a statute or constitutional provision that speaks directly to the act. A tradition can fill a gap, and it cannot repeal a rule that Congress or the Constitution wrote.
The Court has since used the gloss in this disciplined form. In Medellín v. Texas the Government defended a presidential memorandum directing state courts to reopen criminal judgments by relying on the claims-settlement cases, which rest on the view that a systematic, unbroken practice, long pursued to Congress’s knowledge and never before questioned, can raise a presumption that the action was taken with Congress’s consent.85 The Court accepted the principle and refused the application. It described the authority those cases support as “narrow and strictly limited,” the authority to settle international claims disputes by executive agreement, and it found no comparable record for the memorandum, which the Government itself had called “unprecedented action” and for which it could identify no earlier presidential directive to state courts.86 The Court’s counting resembles Frankfurter’s and differs from the dissent’s inventory. It defined the practice by the specific kind of act and did not let a long record of claims settlements stand for the broader proposition that Presidents act freely in foreign affairs. Because the Court spoke of a presumption, a qualifying practice strengthens the President’s claim without ending the inquiry, and its weight still depends on what Congress has done, as Jackson’s categories show.
Step four removes the seizures that rested on the Smith-Connally Act, which is why Vinson’s longer list shrinks under Frankfurter’s test. Step one explains the result in Medellín, where settling claims by agreement and directing state courts are different acts even though both sit under the heading of foreign affairs. Step five comes last because a record that passes the first four steps still yields only an interpretation of an open phrase, and a statute that speaks to the act controls.