Just as the Roberts Court has been expanding presidential authority to its historic maximum, recent legal scholarship has shown that the Founders intended, to paraphrase Justice Jackson’s famous Youngstown concurrence, a much lower ebb or at least an ambiguous twilight about “executive power,” in contrast to originalists’ unsupported certainties.
The Constitution is silent about whether the president has the power to remove executive officers. A century ago, the Supreme Court inferred a more limited removal power from Article II, and then it left in place Congress’s power to create independent regulatory commissions and adjudicative officers. The Roberts Court has gone far further, concluding that Article II implies an “indefeasible” removal power (i.e., Congress cannot set conditions on removal), first in Free Enterprise vs. PCAOB (invalidating what appeared to be two layers of “good cause” protections from removal), and then in Seila Law vs. CFPB (invalidating the protections for a single head of an agency with significant executive power). In the wake of these new interpretations, federal courts are questioning or invalidating the protections for Securities Exchange Commission (SEC) administrative law judges, the SEC itself, the Federal Trade Commission, and the National Labor Relations Board. On the campaign trail, former President Trump has promised “aggressive” presidential removal to “dismantle the deep state” and fire many civil servants (“rogue bureaucrats”), asserting presidential power to order “Schedule F” and to implement the Heritage Foundation’s “Project 2025.”
These assumptions about a sweeping removal power are part of the “unitary executive theory,” which argues that Article II gives the president broad implied powers that cannot be checked and balanced by the other branches. These assumptions about removal often serve as a foundation for other expansions of presidential power. For example, during oral argument on presidential immunity this April, Justice Kavanaugh went on a long soliloquy about how time had vindicated Justice Scalia’s solo dissent in Morrison v. Olson, which argued that independent counsels should be unconstitutional. Kavanaugh relied on that background to argue for broadening presidential immunity, and a majority seemed to agree.
The Supreme Court has inferred a removal power from Article II’s Vesting clause (“The executive Power shall be vested in a President”) and the Take Care clause (the president “shall take Care that the Laws be faithfully executed”). The Roberts Court has concluded that the “executive power” historically implied a removal power, and once “executive power” is “vested” in the president, it is completely and exclusively vested in the president. And the Roberts Court extrapolates from the president’s duty to “take care” that the president also must have an unlimited power to remove significant executive officers in order to fulfill that duty.
The main historical evidence for these claims is the “Decision of 1789,” in which the First Congress supposedly interpreted Article II along those lines. Legal historians recently have dismantled this mythic tale, finding indecision and contradictory evidence against these interpretations, from Jonathan Gienapp’s award-winning The Second Creation, Christine Kexel Chabot’s series of articles on the First Congress, and my article The Indecisions of 1789. Andrew Kent, Ethan Leib, and I offered historical counterevidence to the assumption that the Take Care clause would imply unchecked presidential powers like indefeasible removal. We found a centuries-long legal history of the term “faithful execution” as a legal limit on executive discretion, so that Article II executive powers were not unconditional, but instead constrained by proto-fiduciary duties of good faith, care, and loyalty to the public interest. It would be incongruous for the duties imposed by the power-limiting Take Care clause to create unlimited implied powers.
My historical research on the executive vesting clause found that the word “vested” did not imply exclusive, unconditional, or legislatively indefeasible removal power, contrary to some originalists’ claims. Jodi Short and I have also argued that the Roberts Court’s theory of “presidential superiority,” that the presidency was designed to represent the nation by “direct” elections, is deeply ahistorical. In this publication, Chabot astutely described the Roberts Court’s methods as “Article II Vibes.”
Nevertheless, originalists have responded that, so long as the Constitution’s other Article II power—“executive power”—implied removal, Article II gives it to the president and no one else. That response begs the antecedent question: Did “executive power” imply removal? From our twenty-first-century perspective (after the rise of the “imperial presidency”) we have assumed that the answer must have been yes. It is hard for us to imagine “executive power” in government or in the private sector without conjuring up the phrase “you’re fired.” But the seventeenth and eighteenth centuries were very different from ours in terms of communication, technology, decentralization, and necessary political compromises.
Daniel Birk found a long history of the English Parliament protecting executive officers from royal removal at pleasure. Jane Manners and Lev Menand followed with one of the best deep histories on the origins of the modern conditions (requirements of “inefficiency, neglect of duty, or malfeasance”): those requirements were actually “permissions” for removal because so many English offices had been completely unremovable.
Unitary executive theorists have responded, but without any evidence that holds up to inspection. In The Executive Power of Removal, Aditya Bamzai and Saikrishna Prakash attempted to resurrect the originalist argument. Their article has already been sufficiently refuted by Andrea Scoseria Katz and Noah Rosenblum in Removal Rehashed. I have also detailed their repeated use of historical evidence out of context in a long Appendix to “Indecisions of 1789” and in my amicus brief in SEC v. Jarkesy. They also do not respond to the critiques that Gienapp and I raised at a conference about their earlier work. One additional glaring problem is how non-originalist the unitary executive theorists’ response is. They offered very little evidence from the Convention and Ratification, and none of that evidence supports their argument for an indefeasible removal power. Instead, they retreated forward to a heavy reliance on mid-nineteenth-century practice and backward to European history and a “British Backdrop” of removal. They assert, without evidence, that the Framers intended “executive power” to refer to a European tradition of executive removal, especially by the English king. Steven Calabresi similarly relies on British colonial practice in his Jarkesy amicus brief and his forthcoming book. Ordinarily, we interpret the Constitution (especially the Bill of Rights) as a rejection of royalism and English practices. But Bamzai, Prakash, and Calabresi rely on royalism and British colonial rule. When is English history a model and when is it an anti-model? I have called this convenient historical cherry-picking “Heads I Win, Tails You Lose Originalism.”
Interestingly, at least one unitary executive scholar has rejected the others’ approach. In his book The President Who Would Not Be King, Michael McConnell argued that relying merely on royal traditions was, well, royalist, and therefore inconsistent with our republican rule-of-law tradition. Such an open-ended grab bag from centuries of royal uses (and abuses) would be potentially lawless and especially vulnerable to abuse during emergencies. McConnell’s response attempted a more originalist revival of removal power—ostensibly based on concrete legal sources. Specifically, he argued, “executive power” implied the royal prerogative powers, removal was a royal prerogative. But the evidence does not support these claims. First, Andrew Kent, Julian Davis Mortenson, and I have shown that the Framers did not think “executive power” implied the royal prerogative; and second, I have shown that there is no historical evidence that removal was a royal prerogative or even a general royal power.
These findings may seem too surprising to be believed. Even if it should seem obvious that royalism is no model for Article II, surely one reason, among many, that the Framers rejected European royalism is that absolute monarchs ruled absolutely, and we assume their absolute rule included removal. However, European history is much more complicated—and much more interesting—than those simple assumptions.
Building on others’ work in my forthcoming article “Venality and Functionality: A Strangely Practical History of Officers’ Independence and Limited Executive Power,” I dug into the history of offices to understand why English legal sources do not describe a general removal power. It turns out that English law protected more offices as unremovable “freehold property” than I had ever imagined, and for more surprising and fascinating reasons—a logic of political and technological necessities that lasted up to and beyond the American Founding. This functional logic was called “venality”: the buying and selling of profitable offices, as a more practical way to incentivize “faithful execution” in an era before modern technology, communication, and measurement.
Legal scholars have overlooked the European venality system and its persistence through nineteenth-century England. Economic historian Douglas Allen wrote the definitive book on the mixed decentralized system of offices, finding that it involved buying-selling markets in offices and in patronage that “maximized the value of the kingdom.” Nicholas Parrillo’s outstanding book Against the Profit Motive shows that, before the “salary revolution,” most public officers were compensated by fees and bounties, incentivizing them to do their jobs. Thus, these offices were profitable. And, because they were profitable, the central government could find a profitable market by selling them to the highest bidder. The sale of office relied on willingness to pay as a signal of literacy, expertise, self-assessment of competence, and willingness to work. At the same time, the officer who bought the office would require some legal protection for the investment. English law offered the strongest protection of all the European regimes for venality: the “freehold property” right in office.
G.E. Aylmer, the leading historian of early modern English administration, summarized the English system with six characteristics:
This system protected not only remote or mid-level executive officers. English historians have documented the unremovability of many powerful executive officers through the late eighteenth century, especially high English Treasury offices and even “department heads” in the cabinet. Even though this system had its critics, it was regarded as a necessity that could be reformed but not abolished. In Spirit of the Laws, Montesquieu defended venality as part of a well-balanced constitutional monarchy, and he rejected “immediate displacement” as a tool of “despotic government.” William Blackstone, Edmund Burke, and Jeremy Bentham each defended this system consistently with anti-corruption checks on the monarchy and Parliament and with republican values of checks and balances.
This legal system of sale-of-office and offices-as-freehold property lasted until the late nineteenth century in England, and it also shaped the colonial experience and the Founding. Some critics of my argument suggest that the Framers surely rejected this system. This response is a telling reminder of the “Heads I Win, Tails You Lose” originalism problem: the unitary executive theory relies on assumptions about English practice when they support presidential power, but when the historical record challenges those assumptions, suddenly it’s tails, you lose. Whether the Framers approved or disapproved of it, this mixed system of offices-as-property and patronage was the “British Backdrop” and not the Roberts Court’s (or Bamzai and Prakash’s) assumption of a general or default royal removal power.
To guard against cherry-picking, originalists should follow a rule that English practice by itself is not probative unless the documentary record shows the Framers endorsing it or following it. On that note, the evidence suggests at least a continuity of the offices-as-property legal norms among the Founding generation. The Constitution refers to “offices of profit” three times, and the Constitution did not abolish this system. Second, the Opinions Clause has always been a textual problem for the unitary executive theory, because if Article II implied a presidential power to remove department heads, it would not need to specify a power merely to ask for department heads’ opinions. The discussion of the Opinions Clause in the Ratification debates and the First Congress reinforces this background understanding that “department heads” could be independent. Third, scholars have struggled for years to explain why Chief Justice John Marshall concluded that William Marbury was “not removable” from his non-Article III office of Justice of the Peace, with just a five-year tenure. Manners and Menand’s findings, plus this story of venality, provide a more coherent explanation. Fourth, the debates in the First Congress reflected the persistence of offices-as-property, and early Congresses adopted a similar system of “sureties,” financial bonds for faithful performance, roughly similar to a financial investment. The sale of offices-as-property may seem strange and corrupt today, but it was a practical foundation for the nation-state, modern administration, and colonial expansion.
If the historical evidence has been clearly against the unitary executive theory and the Roberts Court’s assumptions about Article II’s original public meaning, why have unitary executive theorists and jurists remained so wedded to this historical interpretation? And why are conservatives who are otherwise committed to the decentralization of separation of powers and of federalism so wedded to centralized presidential power, even when both parties have been winning the presidency?
Nikolas Bowie and Daphna Renan observe that the unitary executive is part of the rise of separation-of-powers formalism, which enabled the rise of judicial supremacy and juristocratic power over the other branches from the Taft Court to the Roberts Court. A more specific explanation from Stephen Skowronek, John A. Dearborn, and Desmond King is that, in our era of congressional paralysis, a strong presidency is the conservatives’ only vehicle for fighting back against the New Deal administrative state’s constant growth. Ashraf Ahmed, Lev Menand, and Noah Rosenblum add that presidentialism serves the interests of de-regulation and big business, giving presidents the power to serve their special interest supporters by cutting through “red tape” and a pro-regulation bureaucracy.
I have offered another explanation: the unitary executive theorists tend to be cultural conservatives more than economic conservatives, and they seem to perceive the “deep state” and the “fourth branch” as a danger to their cultural and social values. The presidency has a structural tilt towards traditionalism and populism: both parties’ presidential primaries and the Electoral College produce a gravitational pull away from elite cultural and social values toward the more culturally conservative median voters of New Hampshire, Pennsylvania, and Wisconsin. Thus, both Democratic and Republican presidents have political incentives to curtail the bureaucracy’s socially progressive, pluralist, or secularist tendencies.
No matter which theory is right, it is increasingly clear that the unitary executive theory represents a legitimacy crisis for originalists and for the Roberts Court. History shows, in the very least, that unitary theorists have not met their evidentiary burden. Future litigants should caution the Roberts Court to exercise more restraint and leave in place the long-standing understanding of the removal power. If the Roberts Court grasps that the tide has turned, it should acknowledge that its theory of presidential superiority has ebbed.
