Interpretation & Debate

Article II, Section 2: Treaty Power and Appointments

Matters of Debate

Common Interpretation

John O. McGinnis headshot
by John O. McGinnis

George C. Dix Professor in Constitutional Law at Northwestern University's Pritzker School of Law

Peter M. Shane headshot
by Peter M. Shane

Jacob E. Davis and Jacob E. Davis II Chair in Law at The Ohio State University Moritz College of Law

The Constitution provides, in the second paragraph of Article II, Section 2, that “the President shall have Power, by and with the Advice and Consent of the Senate to make Treaties, provided two thirds of the Senators present concur.” Thus, treaty making is a power shared between the President and the Senate. In general, the weight of practice has been to confine the Senate’s authority to that of disapproval or approval, with approval including the power to attach conditions or reservations to the treaty. 

For instance, the authority to negotiate treaties has been assigned to the President alone as part of a general authority to control diplomatic communications.  Thus, since the early Republic, the Clause has not been interpreted to give the Senate a constitutionally mandated role in advising the President before the conclusion of the treaty.

Also of substantial vintage is the practice by which the Senate puts reservations on treaties, in which it modifies or excludes the legal effect of the treaty. The President then has the choice, as with all treaties to which the Senate has assented, to ratify the treaty or not, as he sees fit.

The question of whether the President may terminate treaties without Senate consent is more contested. In 1978, President Carter gave notice to Taiwan of the termination of our mutual defense treaty. The U.S. Court of Appeals for the District of Columbia held that the President did have authority to terminate the treaty, but the Supreme Court in Goldwater v. Carter (1979), vacated the judgment without reaching the merits. The treaty termination in Goldwater accorded with the terms of the treaty itself. A presidential decision to terminate a treaty in violation of its terms would raise additional questions under the Supremacy Clause, which makes treaties, along with statutes and the Constitution itself, the “supreme Law of the Land.”

There remains the question of how the Treaty Clause comports with the rest of the system of enumerated and separated powers. Missouri v. Holland (1920) suggests that the Treaty Clause permits treaties to be made on subjects that would go beyond the powers otherwise enumerated for the federal government in the Constitution. In Reid v. Covert (1957), however, the Court held that treaties may not violate the individual rights provisions of the Constitution.  And in Bond v. United States (2014) the Court read the Chemical Weapons Convention Implementation Act narrowly, observing that a treaty’s implementing statute cannot, without a clear statement, intrude upon ‘the delicate balance of federalism’ and convert the national government’s limited powers into a general police power.”

Whether the Treaty Clause is exclusive—or merely first among several tools—for striking significant international bargains remains contested. In fact, the majority of U.S. pacts with other nations are not formal “treaties,” but are sometimes adopted pursuant to statutory authority and sometimes by the President acting unilaterally. The Supreme Court has endorsed unilateral executive agreements by the President in some limited circumstances. For instance, in United States v. Belmont (1937), the Court upheld an agreement to settle property claims of the government and U.S. citizens in the context of diplomatic recognition of the Soviet Union. In Dames & Moore v. Regan (1981), the Court upheld President Carter’s agreement with Iran, again concerning property claims of citizens, in the context of releasing U.S. diplomats held hostage by Iran. The Court has never made clear the exact scope of permissible unilateral agreements, but they appear to include at least one-shot claim settlements and agreements attendant to diplomatic recognition though no case suggests a sole‑executive agreement can supersede an Act of Congress. 

With so-called congressional-executive agreements, Congress has also on occasion enacted legislation that authorizes agreements with other nations. For instance, trade agreements, like the North America Free Trade Agreement (NAFTA), have often been enacted by statute. In contrast, the Senate balked when President Carter floated the idea of approving SALT II by ordinary statute instead of by the Constitution’s two‑thirds treaty route for the Strategic Arms Limitation Talks II (SALT II) treaty. It is sometimes argued in favor of the substantial interchangeability of treaties with so-called congressional-executive agreements that Congress enjoys enumerated powers that touch on foreign affairs, like the authority to regulate commerce with foreign nations. But, unlike legislation, international agreements establish binding agreements with foreign nations, potentially setting up entanglements that mere legislation does not.

Since Chief Justice John Marshall’s opinion in Foster & Elam v. Neilson (1829), the Supreme Court has distinguished between treaties that are now called self-executing and treaties that are non-self-executing. Self-executing treaties have domestic force in U.S. courts without further legislation. Non-self-executing treaties require additional legislation before the treaty has such domestic force. In Medellín v. Texas (2008), the Court suggested there may be a presumption against finding treaties self-executing unless the treaty text in which the Senate concurred clearly indicated its self-executing status.

  1. Appointments

The remainder of Section 2  of Article II deals with the subject of official appointments. With regard to diplomatic officials, judges and other officers of the United States, Article II lays out four modes of appointment. The default option pairs presidential nomination with Senate confirmation. With regard to “inferior officers,” Congress may, within its discretion, vest their appointment “in the President alone, in the courts of law, or in the heads of departments.” The Court has yet to mark a clear boundary between methods of appointment: inferior officers may be chosen inside the executive hierarchy or, where no ‘incongruity’ arises, by the judiciary. Morrison v. Olson (1988).

Buckley v. Valeo (1976) confirms that the Article II variations are Congress’s sole options in providing for the appointment of officers of the United States. The text, however, raises the questions: Who counts as an “officer” of the United States, as opposed to a mere employee? And what characterizes an officer’s status as “inferior,” as opposed to “superior” or “principal?”

The Court’s definition of “officer” in Buckley entails a degree of circularity. In general, “any appointee exercising significant authority pursuant to the laws of the United States” is an “officer of the United States.” By contrast, a federal employee is not an “officer” if performing “duties only in aid of those functions that Congress may carry out by itself, or in an area sufficiently removed from the administration and enforcement of the public law as to permit their being performed by persons not ‘Officers of the United States.’” A later case, INS v. Chadha (1983), may implicitly have given the Buckley formulation more substance. Chadha held that the enactment of legislation is Congress’s only permissible means of taking action that has the “purposes and effect of altering the legal rights, duties and relations of persons . . . outside the legislative branch.” Importing Chadha’s holding into the Buckley holding implies that, at a minimum, any administrator Congress vests with authority to alter the legal rights, duties and relations of persons outside the legislative branch would have to be an “officer,” and not an employee, of the United States because that officer would be performing a function forbidden to Congress acting alone.

Freytag v. Commissioner of Internal Revenue, 501 U.S. 868 (1991), applied the Buckley formulation to hold that special trial judges of the U.S. Tax Court were “officers of the United States.” The Court’s treatment of the issue cited a number of characteristics without indicating the weight that might attach to each. The Court observed: “The office of special trial judge is ‘established by Law,’ Art. II, § 2, cl. 2, and the duties, salary, and means of appointment for that office are specified by statute. These characteristics distinguish special trial judges from special masters, who are hired by Article III courts on a temporary, episodic basis, whose positions are not established by law, and whose duties and functions are not delineated in a statute. Furthermore, special trial judges perform more than ministerial tasks.”

Distinguishing inferior from principal officers has also sometimes proved puzzling. Morrison v. Olson, which upheld the judicial appointment of independent counsel under the Ethics in Government Act of 1978, applied a balancing test focused on the breadth of the officer’s mandate, length of tenure, and limited independent policymaking. A later decision, however, provided an additional or perhaps substitute bright-line test, defining “inferior officers” as “officers whose work is directed and supervised at some level by others who were appointed by Presidential nomination with the advice and consent of the Senate.” Edmond v. United States (1997).

Later cases have appeared to clarify that it would be inconsistent with “inferior officer” status that an officer be empowered to make significant binding decisions on behalf of the executive branch. Thus, for example, in United States v. Arthrex, Inc., 594 U.S. 1 (2021), the Court found that Administrative Patent Judges serving in the Commerce Department’s Patent and Trademark Office were not “inferior” because, in certain cases, they could make effectively final decisions concerning whether existing patents satisfy the novelty and nonobviousness requirements for inventions. Because their mode of appointment—appointment by the Secretary of Commerce—would be valid only if APJ’s are “inferior,” the Court read into the relevant statute authority for the Director of the PTO to review and, as appropriate, set aside decisions of the APJ’s, thus confirming their subordinate rank. The Court underscored the point in Lucia v. SEC (2018), holding that administrative law judges who enter final adjudicatory orders wield ‘significant authority’ and therefore must be appointed—or at least directly controlled—by a presidentially accountable officer; where meaningful supervision is absent, such officials cross the constitutional line from inferior to principal status. The Court likewise found members of the U. S. Preventive Services Task Force within  Department of Health and Human Services (HHS) to be inferior because the Court determined them to be removable at will and subject to review and overruling by the Secretary of HHS. Kennedy v. Braidwood Management, Inc., 606 U.S. ___ (2025).

The Recess Appointments Clause reflects an era when Congress rode home for months, so the President could keep offices filled during the long recess. Over the ensuing decades—and extending to modern times when Congress itself sits nearly year-round—the somewhat awkward wording of the Clause seemed to pose two issues that the Supreme Court decided for the first time in 2014. First, does the power of recess appointments extend to vacancies that initially occurred while the Senate was not in recess? Second, may a period of Senate adjournment trigger the President’s recess appointment power even if that period of adjournment occurs during a Senate session, rather than between the adjournment of one session sine die and the convening of the next? Finding the text ambiguous, the Court answered both questions affirmatively, provided that the relevant “intra-session” recess lasted ten days or longer. (As a result, in the particular case, the Court ruled against the President, because the relevant recess was too short.) The majority rested its analysis on what it took to be a relatively consistent pattern of behavior by Congress and the executive branch, effectively ratifying the President’s power as thus construed. But by holding that even pro‑forma sessions interrupt a ‘recess,’ Noel Canning may render the Clause largely vestigial in our age of near‑continuous Senate sittings: only the rare adjournment of ten days or more now opens the constitutional window for a recess appointment. NLRB v. Noel Canning (2014).

Perhaps the greatest source of controversy regarding the Appointments Clause, however, surrounds its implications, if any, for the removal of federal officers. The Supreme Court has held that Congress may not condition the removal of a federal official on Senate “advice and consent,” Myers v. United States (1926), and, indeed, may not reserve for itself any direct role in the removal of officers other than through impeachment, Bowsher v. Synar  (1986).

Those cases do not determine, however, whether Congress may limit the President’s own removal power, for example, by conditioning an officer’s removal on some level of “good cause.” The Supreme Court first gave an affirmative answer to that question in Humphrey’s Executor v. United States (1935), which upheld a statutory limitation on the President’s discretion in discharging members of the Federal Trade Commission to cases of “inefficiency, neglect of duty, or malfeasance in office.” Morrison v. Olson reaffirmed the permissibility of creating federal administrators protected from at-will presidential discharge, so long any restrictions on removal do “not impermissibly interfere with the President’s exercise of his constitutionally appointed functions.” Although this formulation falls short of a bright-line test for identifying those officers for whom presidents must have at-will removal authority, the doctrine at least implies that presidents must have some degree of removal power for all officers. That is, presidents must be able at least to secure an officer’s discharge for good cause, lest the President not be able to take care that the laws be faithfully executed.

The Roberts Court, however, has taken significant steps toward expanding the President’s right to remove executive officers at will. In 2010, the Court held that officers of the United States may not be shielded from presidential removal by multiple layers of restrictions on removal. Thus, inferior officers appointed by heads of departments who are not themselves removable at will by the President must be removable at will by the officers who appoint them. Free Enterprise Fund v. Public Co. Accounting Oversight Board (2010).

A decade later, the Court characterized Humphrey’s Executor and Morrison v. Olson as two exceptions from what it inferred from Myers v. United States as the proper understanding of Article II, namely, that presidents are constitutionally entitled to remove without cause officers who exercise executive power on the president’s behalf. In Seila Law v. Consumer Finance Protection Bureau (2020), the Court applied that understanding to set aside statutory tenure protection for the head of any agency governed by a single director, such as the CFPB. The following term, the Court likewise invalidated statutory tenure protection for the administrator of the Federal Housing Finance Agency, notwithstanding that the regulatory reach of the FHFA was far narrower than that of the CFPB.

In Seila Law, the Court limited Morrison v. Olson to the approval of “for cause” tenure protection “for inferior officers with limited duties and no policymaking or administrative authority.” It read Humphrey’s Executor as providing an exception from removal at will only for multimember bodies of experts, balanced along partisan lines, that perform quasi-legislative and quasi-judicial functions. It is currently uncertain whether the Court will allow that exception to remain or will instead overrule Humphrey’s Executor. Under President Donald Trump, the Justice Department has argued that an exemption for multimember agencies cannot survive the logic of Seila Law, and the President determined to test that proposition by firing a number of independent officers notwithstanding statutes purporting to protect them from at-will removal.

Two such officials whom the President fired succeeded temporarily in securing reinstatement based on lower court decisions continuing to follow Humphrey’s Executor. Nonetheless, in a 6-3 per curiam decision, Trump v. Wilcox (2025), the Supreme Court granted the Administration a stay of the reinstatement order while the dispute continues to work its way up to the Supreme Court on the merits. The majority opinion purports not to decide whether the multimember agencies at issue (and thus their principal officers) fall within the Humphrey’s Executor exception to what the Court how holds is an implicit Article II grant of at-will removal power to the President. Yet the Court predicted government success on the question whether those agencies exercise “considerable executive power,” which seems to suggest a predisposition to uphold the contested discharges. If the Court were to sustain the firings at issue, it would likely proceed either to read Humphrey’s Executor to apply only to agencies with no significant executive authority or overrule it altogether.

On the other hand, the Court appeared to recognize anxiety in financial markets that overruling Humphrey’s Executor or narrowing it to its facts would put at risk the independence of the Federal Reserve’s Board of Governors or other members of the Federal Open Market Committee. The Court tried, however, to allay such anxiety, stating: “The Federal Reserve is a uniquely structured, quasi- private entity that follows in the distinct historical tradition of the First and Second Banks of the United States.” Whether the majority persists in that view following what is sure to be full briefing and oral argument remains to be seen.

Delphic Article II

Peter M. Shane headshot
by Peter M. Shane

Jacob E. Davis and Jacob E. Davis II Chair in Law at The Ohio State University Moritz College of Law

Article II of the U.S. Constitution is plainly critical to establishing two fundamental institutional relationships: the President's relationship with Congress and the President's relationship to the remainder of the executive establishment, which we would now call "the bureaucracy." Despite the text's seeming specificity on some key points—e.g., the President's role in the appointments process—the Constitution's silences and the ambiguity of the text in other respects have fueled spirited arguments through the centuries for very different concepts of the American presidency. To paraphrase Justice Robert Jackson, Americans may "be surprised at the poverty of really useful and unambiguous authority applicable to concrete problems of executive power as they actually present themselves." Youngstown Sheet & Tube Co. v. Sawyer (1952).

With regard to the legislative-executive relationship, the Washington Administration set institutional precedents that have been followed with such consistency over the centuries that they now dominate our understanding of Article II. To the uninitiated reader, the Treaty Clause might be thought to imply that treaties represent the sole permissible instrument for formalizing the nation's international obligations, or that the Senate, because of its "advice and consent" role, would be a full partner with presidents in the negotiation of treaties. Neither is the case. The Washington and Adams Administrations used executive agreements, without Senate consent, both in arranging for the international delivery of mail and in settling claims arising from the seizure of a U.S. ship by a Dutch privateer. Such agreements, sometimes pursued unilaterally and sometimes with statutory authority, now far outnumber treaties as instruments of international commitment. As for actual treaties, when the Senate failed to provide Washington prompt advice concerning the negotiation of peace between Georgia and the Creek Indians, he established the now-uniform practice of presenting to the Senate for its consent only treaties that have already been completed.

The first Congress and the Washington Administration also began filling in some of the constitutional silences regarding their respective powers. Congress first asserted its unstated power to investigate the executive branch by establishing a special committee to look into the bloody defeat of the U.S. Army by a confederation of Indian tribes in the Northwest Territory. Washington, for his part, provided the committee with those executive branch documents it sought to inform its investigation, but only after determining with his cabinet that the disclosure decision was discretionary on his part and that presidents might constitutionally withhold information that ought, in the public interest, not be disclosed. He later implemented his view by withholding from the House of Representatives documents it sought in connection with negotiations over the Jay Treaty. This laid a foundation for future claims of executive privilege, a phrase nowhere found in Article II.

Text, even aided by history, however, shines less light on constitutional requirements for the President's relationship to those other instrumentalities of government that Congress creates but which are not part of the federal judiciary—that is, to the plethora of "departments," "agencies," "administrations," "boards," and "commissions" comprised within the executive branch. Recent decades have seen much ardent advocacy on behalf of the so- called "Unitary Executive Theory"—specifically, the view that Article II, by vesting law execution power in the President, forbids Congress from extending any such authority to individuals or entities not subject to presidential control. Adherents to the unitary executive reading of Article II insist that the Constitution guarantees the President plenary powers, which Congress may not limit, both to discharge unelected executive administrators at will and to direct how those officials shall exercise any and all discretionary authority that they possess under law.

In holdings prior to the Roberts era, the Supreme Court had not embraced Unitary Executive Theory. It came closest in Myers v. United States (1926), which held that Congress could not require the President to secure Senate consent before firing a Senate-confirmed postmaster in the executive branch. Chief Justice (and former President) William Howard Taft, writing for the majority, argued in extensive dicta that it was necessary for the President’s capacity to take care that the laws be faithfully executed that all presidentially appointed officers serve at the President’s pleasure. Technically, however, excluding the Senate from the removal process did not resolve the question whether Congress could simply limit the permissible grounds for removal by presidents acting on their own.

Over the next decades, however, Taft’s dicta had minimum impact because the Court answered that last question in the affirmative. The Supreme Court’s unanimous 1935 decision in Humphrey’s Executor v. United States—rendered by an intellectually heavyweight Court with a philosophically diverse membership—upheld statutory tenure protections for members of the Federal Trade Commission. The FTC performed “executive functions,” in the Court’s view, only in connection with its primary “quasi-legislative” and “quasi-judicial” missions—that is, administrative adjudication and rulemaking. Myers, the Court stated, applied only to “purely executive officers,” a class it did not define further.

Rooted in Humphrey’s Executor, the Court’s account of executive power grounded a presidency woven into a system of checks and balances, under which presidents owe significant accountability to both of the other branches. Congress might choose, as it has, to make most principal officers in the executive branch subject to discharge at the will of the President. But constitutionally, specifying the conditions of officer tenure would remain Congress’s choice. Congress could (and did) vest functions in tenure-protected officers.

Likewise, Congress could limit the “unitariness” of presidential control by vesting the appointment of inferior officers—even inferior executive officers—in courts of law. There is no textual bar to interbranch appointments and, contrary to some originalist claims, the boundary between executive and judicial power was highly uncertain in 1787. As the Supreme Court recognized in Ex Parte Siebold (1883): “It is no doubt usual and proper to vest the appointment of inferior officers in that department of the government, executive or judicial, or in that particular executive department to which the duties of such officers appertain. But there is no absolute requirement to this effect in the Constitution, and if there were, it would be difficult in many cases to determine to which department an office properly belonged.”

The Court’s 7-1 decision in Morrison v. Olson (1988), reaffirmed these points. Citing Humphrey’s Executor with approval, the Court upheld the constitutionality of the independent counsel provisions of the post-Watergate Ethics in Government Act, including the statutory provisions barring the counsel’s removal at the will of the President. The Court also rejected the constitutional challenge to the judicial appointment of prosecutors, citing Ex Parte Siebold.

As for courts, the judicial role vis-à-vis the executive in a system of checks and balances is simply stated. Judicial review of administrative action to insure compliance with statutory requirements of both substance and procedure is essential to the rule of law under a democratic constitution.

Key Roberts Court opinions, however, have pointed in a very different direction. A majority now insists the president “alone composes a branch of government” and thus holds the “entirety” of executive power. It describes the civil service as wielding executive power “on the president’s behalf.” The Court contends that what gives the bureaucracy “legitimacy and accountability” is that Americans get to vote for president, to whom bureaucrats are tied through “a clear and effective chain of command.” A majority has even declared presidential powers of removal and supervision to be “conclusive and preclusive,” thus beyond Congress’s power to regulate and largely beyond the judiciary’s power to review. No one would now be surprised if a divided Roberts Court were to overrule Humphrey’s Executor or construe it so narrowly as to be nearly meaningless.

In my view, the Roberts Court’s narrative of the presidency is triply wrongheaded. It muddles constitutional text. It flouts constitutional history. It is willfully blind to the risks of authoritarianism. The very idea that the Constitution creates a one-person branch of government is provably untrue just by reading the Constitution, which explicitly refers to “executive departments.” Moreover, these departments have “duties,” most of which are to be set forth in statutes. Fulfilling statutory duties is the job of these agencies, which, in doing their work, act not on behalf of the president, but chiefly on behalf of Congress. The president’s role in this scheme is one of supervision, not command. He is charged to “take care that the laws be faithfully executed.” The Constitution underscores the president’s supervisory position by providing he may “require the opinion, in writing, of the principal officer in each of the executive departments, upon any subject relating to the duties of their respective offices.” This is not at all a one-person branch of government, and its design is not the prerogative of the president, but of Congress.

Likewise, the idea that vesting the president with “the executive power” means “all” the executive power is also not found in the Article II text, which does not contain the word “all.” Where the word “all” does appear is in the Constitution’s vesting in Congress the power to “make all laws which shall be necessary and proper for carrying into execution [Congress’s] powers, and all other powers vested by this Constitution in the government of the United States, or in any department or officer thereof.” Far from signaling a wide swath of “conclusive and preclusive” executive authority beyond accountability to the other branches, the text suggests a sweeping legislative power to prescribe how executive power is to be exercised.

The president’s constitutional role properly described thus does not require at-will removal power, except in the cases of those few officials who directly assist the president in fulfilling specific Article II roles. For all others—the overwhelming majority of government officers and employees—presidents need only the power to discharge persons who have failed to faithfully execute the law, thus providing “good cause” for their removal. This means that presidents, even when implicitly authorized by statute to fire officials at will, should be conscious that their power to do so is typically at the sufferance of Congress.

History is no more on the side of the Roberts Court than is the text. The Article II Vesting and Take Care Clauses, viewed either independently or together, did not originally have the semantic implications that unitary executive theorists imagine. These kinds of clauses were prevalent in early state constitutions that also established relationships between governors, as chief executives of the states, and state agencies. Rather than giving governors unitary executive control over state administration, these constitutions nearly all split supervision of the bureaucracy among the different branches of government—the governor, the legislature, and, in some states, the courts. Originalist defenders of a unitary executive reading of the federal Constitution often dismiss the interpretive significance of pre-1787 state constitutions on the ground that these early texts paid only lip service to separation of powers principles, while presenting the Framers chiefly with examples of government structure to avoid. The problem with this stance is that state constitutions written in the first decades after 1789 persisted in using the same clauses, by that time found also in Article II, to describe state governments in which governors continued to lack unitary control. Close study of the state constitutions and state administrative practice under them thus belie any "unitary executive" reading of Article II that purports to be based on contemporary understandings of the text alone.

Nor is the argument borne out by a history of institutional practice. The First Congress's handiwork regarding the structure of the initial administrative departments is inconsistent with the idea that the Framers intended a unitary executive. Congress accommodated presidential control at different levels, from seemingly complete, as with the Department of State, to essentially non-existent, as with the boards and commissions authorized to oversee the Mint, to buy back debt of the United States, and to rule on patent applications.

Unitary executive advocates may point to a variety of presidential statements over the years asserting the existence of a comprehensive presidential supervisory authority. But again to quote Justice Jackson, who wrote in 1952 about constitutional debates on the scope of presidential power: "A century and a half of partisan debate and scholarly speculation yields no net result but only supplies more or less apt quotations from respected sources on each side of any question." Youngstown Sheet Tube v. Sawyer (1952). Unitarian arguments based on presidential statements simply cannot overcome Congress's conspicuous eclecticism from its first session forward in fashioning different administrative structures with different lines of accountability to different sources of supervision.

The argument for the unitary presidency makes the mistake of anachronism. The managerial presidency extolled in the late eighteenth century was just not conceptualized in the policy terms now understood by modern presidentialists. Even if the original presidential office had been intended to be unitary in some administrative sense, the President's originally designed managerial powers cannot logically add up to the contemporary version of unitary power urged upon us by twenty-first century presidentialists, who interpret the Constitution as putting the President personally in charge of the exercise of any or all policy making discretion that Congress may delegate to anyone within the executive branch.

But going beyond even text and history, what is alarming about the Roberts Court’s reinterpretation of Article II is how far it can go in enabling a kind of administrative despotism. As a matter of principle, anyone concerned with preserving robust constitutional checks and balances should be disturbed by the Court’s expansive understanding of the president’s unilateral powers, regardless of that president’s policy agenda. Should those powers be pushed, however, to enable an authoritarian agenda, the risk to democracy and the rule of law is patent. Justice Joseph Story, whose classic 1833 Commentaries on the Constitution of the United States opposed reading into Article II a power of at-will presidential removal power, wrote:

[I]t is utterly impossible not to feel, that, if this unlimited power of removal does exist, it may be made, in the hands of a bold and designing man, of high ambition and feeble principles, an instrument of the worst oppression and most vindictive vengeance . . . It would convert all the officers of the country into the mere tools and creatures of the President. A dependence so servile on one individual would deter men of high and honorable minds from engaging in the public service. And if, contrary to expectation, such men should be brought into office, they would be reduced to the necessity of sacrificing every principle of independence to the will of the chief magistrate, or of exposing themselves to the disgrace of being removed from office. .

There is available a better view of Article II, which is fully reconcilable with the text and truer to our history. It holds that outside those particular subjects that are independently within the President's inherent powers, such as issuing pardons or making treaties, the degree of policy control the President may exercise over subordinate officers is up to Congress. Congress is limited, in turn, only by the Constitution's constraints on the scope of national legislative authority and the President's entitlement to dismiss officers of the United States who are breaking the law or negligent in the execution of their duties.

For this reason, there is an intimate connection between the President's relationship with Congress and the President's relationship to the remainder of the executive establishment. Specifically, the latter is significantly determined by the former. The Constitution gives Congress the political discretion to defer substantially to the pleas of the executive for highly centralized control over administrative agencies, but only if Congress chooses to do so. The bare framework of Article II leaves presidents with the task of persuading Congress that authorizing such control over any particular agency is in the public interest—a judgment of policy, not constitutional interpretation. It is regrettable that such is not the view of the Roberts Court.

An Originalist Reading of Article II, Section 2

John O. McGinnis headshot
by John O. McGinnis

George C. Dix Professor in Constitutional Law at Northwestern University's Pritzker School of Law

The practice and jurisprudence of the Treaty and Appointments Clauses err when they depart, as they too often do, from the original meaning of the Constitution. The original meaning is the meaning that would have been most likely embraced by a reasonable person at the time of the Framing. Because the Constitution is written in the language of the law, the original meaning is constituted by the text in its historical and legal context. Courts are obligated to use the interpretive methods at the time of enactment to find the better-supported meaning, even if an ambiguous text can yield more than one meaning.

The results of an originalist reading of these Clauses would at times favor the President, but at other times disfavor him, but they would more generally promote accountability. They would also create more bright-line rules and limit the discretion of the Supreme Court to make decisions according to opaque balancing tests that maximize its own power.

Appointments Clause. The Appointments Clause must be read against the background of "the executive power" granted to the President. That authority included the traditional powers of an executive, not simply enumerated powers as those specified in Article I. Article II then qualifies that understanding by expressly giving some of the executive's traditional powers to Congress. In the Appointments Clause, the Senate is given the power to advise and consent to nominations. Because the Constitution does not change the executive's power to dismiss subordinate officers, the President retains that unqualified power, as it was part of the traditional executive authority. This view reflects the majority view of the First Congress after a deliberate debate when they insulated the President's authority over the Secretary of State. See Saikrishnah Prakash, New Light on the Decision of 1789, 91 Cornell L. Rev. 1012 (2006). Recent historical work further fortifies this reading, underscoring that the “executive Power” at the Founding encompassed a removal authority necessary to maintain accountability. See Aditya Bamzai & Saikrishnah Prakash, The Executive Removal, 136 Harv. L. Rev. 1756 (2023).

The contrary decisions of the Court were both wrong and unclear. In Morrison v. Olson (1988), for instance, the Court did not offer a rule for determining when Congress could insulate the President's power, but made instead the question depend on such factors as the scope and authority of the office at issue. This aggrandized the Court's power and unsettled an established framework for government. By contrast, the Court’s more recent path—Seila Law (2020) and Collins (2021)—rejects novel single-director removal protections and returns to first principles of Article II supervision, suggesting that the Supreme Court may be converging on the original meaning embraced here.

Similarly, the Court was wrong to permit courts to appoint executive officials so long as there is no "'incongruity' between the functions normally performed by the courts and the performance of their duty to appoint." Morrison v. Olson (1988). It is true that the Appointments Clause allows "courts of law" to appoint "inferior officers." But just as the President's authority under the Appointments Clause must read against the background of Article II, so the courts' authority must be read against the background of Article III that defines their own powers. There the judicial power is defined as "extending to cases." The authority of courts of law in appointments matters is thus more naturally read as ancillary to their defined powers. Moreover, permitting interbranch appointments presents two unpalatable choices. One possibility would be to allow Congress to authorize the judiciary to appoint all inferior officers in the executive branch, thereby depriving the President of the personnel necessary to implement policy and hollowing out his constitutionally assigned responsibility to take care that the laws be faithfully executed. The other possibility would be to let the judiciary decide for itself how far such appointment power should extend, without any constitutional standard to guide its discretion. Either path would transfer control over the executive branch to a body insulated from electoral accountability. Accordingly, courts of law can appoint the officers ancillary to their own work of deciding cases, like law clerks and bailiffs, but not executive officials.

Similarly, Morrison's balancing test for what is an inferior officer wrongly focused on the breadth of the officer's mandate, length of tenure, and limited independent policy making. The appropriate test for inferior officer flows directly from the term's obvious meaning: such an officer must be subordinate to a principal officer; one who has been confirmed by the Senate. It also provides a bright line rule. Happily, the Court may be moving to embrace this test. See Edmond v. United States (1997).

The Court has also failed to follow the original meaning of the Recess Appointments Clause. (For an excellent discussion of the original meaning, see Michael B. Rappaport, The Original Meaning of the Recess Appointments Clause, 52 UCLA L. Rev. 1487 (2004)). First, the power of recess appointments extends only to vacancies that initially arose while the Senate was not in recess. This "arise interpretation" is much better supported than an interpretation that makes the Clause applicable to vacancies that exist whenever there is a recess. The phrase "happened during the recess" naturally implies an event that occurred during the recess. Indeed, not reading the Clause in this way deprives the word "happened" of any independent function. The "arise" interpretation was also the meaning of the Clause embraced even by the executive in the early Republic.

Second, the term "recess" applies only to intrasession recesses. That conclusion flows from the use of the terms adjournment and recess, the former of which in the Constitution seems to be used to refer to intrasession and the latter of which to intersession recesses. In contrast, the Supreme Court's functional rule of ten days cannot be found or inferred anywhere from the text. Moreover, the Court's suggestion in NLRB v. Noel Canning (2014) that its judge-made rule may not even apply in extraordinary circumstances, once again arrogates power to itself. Happily, while Noel Canning was not formally correct, its practical effect has cabined the Clause (especially via the modern Senate’s typical decision to hold pro-forma sessions even when Senators have left town), leaving little modern role for functional expansion.

The Treaty Clause. Just as the President can fire executive officials pursuant to executive power, the President can terminate treaties according to their terms, because that traditional executive power was not limited by the Treaty Clause. However, he cannot terminate treaties in violation of their terms because the Supremacy Clause makes treaties the supreme law of the land.

The Supreme Court is correct that President and the Senate can make treaties beyond the enumerated powers. The Treaty Clause is an executive power in Article II, and does not come with the limitations of Article I. Moreover, as Alexander Hamilton noted, its abuse is carefully guarded by a substantial supermajority rule, one that does not apply to legislation.

While most of the Court's decisions upholding executive agreements are not incorrect, the practice of executive agreements needs to be more clearly circumscribed. The high hurdle posed by advice and consent under a supermajority rule was meant to prevent foreign entanglements. Thus, purely executive agreements should be permitted only when they are one-time agreements, like prisoner exchanges or claim settlements, or when they are based solely on independent presidential authority, like the authority to recognize foreign nation states. See Michael B. Ramsey, The Constitution's Text in Foreign Affairs 191-217 (2007). Empirical work since 2020 documents persistent opacity and under-reporting in executive-agreement practice, strengthening the case for tightening the category and restoring Senate accountability. See Jack Goldsmith et al., The Failed Transparency Regime for Executive Agreements, An Empirical and Normative Analysis, 134 Harv. L. Rev. 629 (2020)

For similar reasons, the notion that Congress and the President together can strike international deals so long as they make a congressional-executive agreement is wrong and would deprive the Treaty Clause of much of its force. Perhaps the practice in some areas of congressional-executive agreements, like trade agreements, is so settled that it should not be reversed. But practice has never embraced the complete interchangeability of treaties and executive agreements, and such interchangeability cannot be squared with the Constitution's express requirements for making treaties.

Matters of Debate