Since the first presidency of Donald Trump, public and academic interest in the theory of the unitary executive has increased exponentially. One of the most important aspects of the unitary theory is its political weaponization of the president’s removal power as a means of asserting control over independent regulatory agencies. Whether such a use of the removal power is constitutionally legitimate has, however, remained contested, both in academic discourse and Supreme Court jurisprudence. Two somewhat recent Supreme Court cases, Seila Law LLC v. Consumer Financial Protection Bureau and Collins et al. v. Yellen, have been touted as watershed cases in which the court fully embraces the unitary theory’s version of the removal power. As such, scholars have argued that these cases provide the final legal justification for the unitary theory and have consequently put it forward as the paradigmatic understanding of executive power.Footnote 1
In this article, I seek to determine to what degree Seila Law and Collins genuinely indicate a final legal sanctioning of the unitary theory, and thus precipitate a serious and significant disruption in both the political status quo and the stability of the constitutional system of separated powers. After a rigorous analysis of the trajectory of the court’s approach to the removal power, I find that there are four potential outcomes for the upcoming clashes between the Trump administration and several recently removed independent agency heads.Footnote 2 First, the court could abide by the logic of its decisions in Seila Law and Collins, preserving Humphrey’s Executor v. United States in full and dismissing the Trump administration’s challenges to judicial limitation on presidential removals. Second, the court could sanction the president’s recent removals, but do so in a way similar to its decisions in Seila Law and Collins—namely by separating the agencies in question from the protections put in place by Humphrey’s while retaining the idea of agency independence in general. Third, the court could overturn Humphrey’s entirely, eliminating removal protection for any and all independent agencies. Finally, the court could overturn Humphrey’s but carve out special exemptions for agencies that have historically been seen as independent from the president—most notably, the Federal Reserve.
Each of these potential outcomes have important constitutional implications, especially for the role of the Supreme Court in the Constitution’s separation-of-powers system. The first two options would essentially preserve the status quo, by allowing the court to continue adjudicating important interbranch disputes over the separation of powers. Either outcome would be historically consistent with the court’s approach to the removal power, as it would preserve its power to check the president and retain its constitutional authority, while also enabling it to provide limited policy victories for the governing coalition with which it is affiliated.Footnote 3 The latter two outcomes would effectively exclude the court from adjudicating removal cases in the future. Doing so would eliminate the court’s ability to fulfill an important aspect of its role within the separation-of-powers system, creating a situation that is both historically anomalous and, arguably, contrary to its constitutional incentives.
This article aims to combine the insights of legal scholars and political scientists on the court’s removal-power jurisprudence to form a more complete view of the current situation regarding the president’s authority over public servants. Understanding the various possibilities for the removal power is vitally important for determining the constitutional implications of the Trump administration’s use of the unitary theory. By examining the development of the court’s approach to the removal power from both a legal and political perspective, I stake out a robust set of possibilities for the future of both the Supreme Court in the constitutional system and the status of the unitary executive theory.
The Unitary Theory and Its Relation to Removal-Power Decisions
At bottom, the theory of the unitary executive asserts that presidents must be in complete and total control of all actions performed by executive officers (Calabresi and Christopher Yoo Reference Calabresi and Yoo2008, 30–38; Dodds Reference Dodds2020; Krent Reference Krent2008; Postell Reference Postell2017, 30–38). In effect, this means that if a president is involved in the appointment of any official, the executive must be granted full discretion in their removal. The unitary executive theory is particularly controversial in current scholarship, with many accusing it of being an existential threat to democracy and democratic government (Barber and Fleming Reference Barber, Fleming and Fleming2011; Driesen Reference Driesen2021; Nelson Reference Nelson2010; Shane Reference Shane2022).Footnote 4 Others more moderately identify the unitary theory as both a problem for effective constitutional governance, and a contradictory claim for conservative “originalists” to take (Chabot Reference Chabot2022; Crouch, Rozell, and Sollenberger Reference Crouch, Rozell and Sollenberger2017; Reference Crouch, Rozell and Sollenberger2020; Mohan Reference Mohan2020; Rozell and Sollenberger Reference Rozell, Sollenberger, Kelley and Shields2013; Shane Reference Shane2010; Skowronek Reference Skowronek2009; Skowronek, Dearborn, and King Reference Skowronek, Dearborn and King2022; Zug Reference Zug2023). Still others have offered critiques of the theory’s historical origins, arguing that neither precedent nor tradition countenance this theory of executive power (Alvis, Bailey, and Taylor Reference Alvis, Bailey and Taylor2013; Bailey Reference Bailey2008; Birk Reference Birk2021; Kastenberg Reference Kastenberg2022; Katz and Rosenblum Reference Katz and Rosenblum2023; Murray Reference Murray2019; Post Reference Post2020). Despite the general alarm and disdain for the unitary theory in the academy, some scholars have suggested that “the unitary executive theory will become the prevailing view of presidential power” (Barilleaux and Maxwell Reference Barilleaux and Maxwell2017, 34). Indeed, given the broad powers that the theory grants to presidents regardless of their partisan affiliation, it seems rather unlikely that it will ever be repudiated by any president (Howell Reference Howell2015, 6–8; Nelson Reference Nelson2010, 182). Consequently, a number of articles have been written suggesting various means of subverting the use of this kind of executive power (Cook Reference Cook2020, 310; Driesen Reference Driesen2021, 140; Fisher Reference Fisher2017; Murphy Reference Murphy2021, 444–45; Nelson Reference Nelson2010, 4; Yalof Reference Yalof and Nelson2010, 458–59).
This scholarship is certainly necessary for understanding the development of the unitary theory, as well as its potential dangers and consequences. Thorough discussion of the historical origins and normative effects of the unitary theory is invaluable for crafting legal arguments against it, which is necessary should one believe it actually poses a threat to democracy. Much of this scholarship assumes—without a proper analysis of the court’s jurisprudence, however—that cases like Seila Law and Collins represent substantial changes in the court’s attitude toward the removal power and thus the unitary theory. Consequently, they hypothesize about imminent threats to democracy on essentially faulty premises. This scholarship overestimates the salience of legal arguments in real political disputes about the unitary theory. Thus, many of these normative arguments assume conditions that do not actually exist. They conflate presidential action that has been historically permitted by the court with novel unilateral powers and thus fail to provide convincing reasons for circumscribing presidential authority as it currently exists.
With the advent of Donald Trump’s second presidential term, the salience of the unitary theory for contemporary politics has only grown. Indeed, Trump has made particular use of the unitary theory’s emphasis on the removal power to justify his assault against the federal bureaucracy (Waldman Reference Waldman2025). The fact that the unitary theory has—in a sense—begun to seep from the realm of theory to that of practice increases the demand for serious scholarship about its effects exponentially. Indeed, as presidential administrations change, the unitary theory changes as well, with different administrations embracing and emphasizing different aspects of it at different times.Footnote 5 The Trump administration is unique insofar as it seems to embrace the unitary theory in full, applying its logic in areas beyond public administration and into even more controversial areas such as budgetary impoundment (Pennington Reference Pennington2025; Sunstein Reference Sunstein2025). Consequently, it is especially necessary to determine whether the theory has come to characterize the court’s interpretation of law. For if it has, then Sonia Sotomayor’s warning in her dissent for Trump v. United States may very well be true: “[I]n every use of official power, the president is now a King above the law.”Footnote 6
The Foundations of the Court’s Approach to the Removal Power
There are two cases that set the stage for the court’s contemporary removal-power rulings. The first is Myers v. United States. In his controlling opinion for that case, Chief Justice William Howard Taft provides one of the foundational arguments in favor of the president’s unfettered removal power. In his controlling opinion, Taft insists that the First Congress unambiguously granted the president the unilateral power to remove any and all officials appointed by him with the advice and consent of the senate in the “decision of 1789.”Footnote 7 Taft engages in a lengthy analysis of the debate in the First Congress about the removal issue and concludes that Congress has the authority to cede its involvement in the appointment process, but not the authority to interpose itself should the president choose to remove an official.Footnote 8 Put simply, Taft argues that Congress’s ability to allow unilateral appointments by the president does not grant it the concurrent power to insert itself in the removal process. Indeed, Taft is explicit here, arguing that “had it been intended to give to Congress power to regulate or control removals … it would have been included among the specifically enumerated legislative powers in Article I, or in the specified limitations on the executive power in Article II.”Footnote 9 Taft further argues that the president’s capacity to unilaterally remove officials is crucial to the effective independent operation of the executive branch.Footnote 10
While Myers endorsed the president’s need for unilateral control over executive officials, the court significantly moderated its stance just nine years later in Humphrey’s Executor v. United States (Yalof Reference Yalof and Nelson2010, 457). In doing so, the court established important parameters for what it would consistently hold to be the proper exercise of the removal power and essentially repudiated the expansive interpretation of the removal power provided in Myers. The growth of the modern administrative state, as well as the contemporary standards for evaluating the scope and extent of the president’s removal powers, are all direct results of the effects of Humphrey’s. As such, the continued acceptance of Humphrey’s as the foundation of the court’s removal-power jurisprudence serves as the most significant obstacle for judicial affirmation of the unitary theory (Crane Reference Crane2015, 1837–40; Sunstein and Vermeule Reference Sunstein and Vermeule2021, 84).
While the curtailment of Humphrey’s may increase the president’s ability to use the removal power, the only way to reestablish the expansive interpretation of the removal power necessary for the unitary theory—and thus provide for its full legal justification—would be to fully overturn this case. Indeed, a January 2024 decision by the US Court of Appeals, while ultimately upholding and applying Humphrey’s, has indicated that perhaps the time is ripe for the Supreme Court to review and ultimately overturn Humphrey’s, declaring that recent cases like Seila Law and Collins have “overtaken” the “logic of Humphrey’s” without fully overturning the decision, thus leaving lower courts and potential plaintiffs in a strange spot.Footnote 11 This ruling—despite its apparent dissatisfaction with the law as it is—indicates that Humphrey’s still has real salience, especially in insulating administrative officials from removal despite the degree to which Seila Law and Collins may genuinely limit its scope.
Humphrey’s involved the removal for purely political reasons of William Humphrey from his position on the Federal Trade Commission. In Myers, Taft tentatively identified only a small exception to the president’s removal power and claimed that the unimpeded power to remove was crucial to the performance of the president’s constitutional duties (Burns Reference Burns2021, 118–25). In Humphrey’s, however, Justice George Sutherland overturned the dicta in Myers that implied a largely unrestricted presidential removal power, despite the fact that he joined Taft’s majority opinion in Myers. Footnote 12 Sutherland’s argument relies on the fundamental difference between the facts of the two cases.Footnote 13 Specifically, Sutherland claims that there is a fundamental distinction between officials performing inherently executive functions—as was the case in Myers—and those performing quasi-legislative or quasi-judicial functions, as he claimed was the case in Humphrey’s. Footnote 14 Because Humphrey was a member of the Federal Trade Commission, a body that Sutherland declares “cannot in any proper sense be characterized as an arm or eye of the executive,” congressional curtailment of the president’s removal power was fully constitutional.Footnote 15
The court’s decision in Humphrey’s had two significant results. First, it qualified and severely limited the scope of the removal power, confining it to officers clearly exercising “executive” authority who were directly answerable to the president. Second, it implicitly legitimated the existence of bureaucratic agencies carrying out what Sutherland referred to in his opinion as “quasi-legislative and quasi-judicial” functions and argued that the proper performance of their tasks demanded independence from presidential oversight and control.Footnote 16 Overall, Humphrey’s applied important caveats to the decision in Myers, limiting the president’s removal power beyond the scope of what proponents of the unitary executive would find acceptable. Most notably, Humphrey’s has hamstrung the president’s ability to exercise direct control over the policies of congressionally created independent regulatory commissions (IRCs), regardless of the president’s involvement in appointing these commissions’ members. Alvis, Bailey, and Taylor (Reference Alvis, Bailey and Taylor2013) note, however, that this significant departure from Myers rests on uncertain ground. They contend that “the only argument Sutherland could find to support his contention for the independence of the [Federal Trade Commission] here was Congress’s assertion that regulatory commissions were different [than normal executive agencies]” (Reference Alvis, Bailey and Taylor2013, 163). According to Alvis and his coauthors, Sutherland’s assertion that quasi-legislative and quasi-judicial powers are fundamentally different from executive powers is merely an assertion without serious constitutional weight.
Other scholars have gone farther, criticizing the decision in Humphrey’s as “messy, confusing, neologism-based, [and] indefensible” (Sunstein and Vermeule Reference Sunstein and Vermeule2021, 84; see also Entin Reference Entin1987, 744–49; Grossman and Sandoloski Reference Grossman and Sandoloski2021, 218–19). Indeed, Sutherland himself admits that the Federal Trade Commission performs an essentially executive function insofar as it “is an administrative body created by Congress to carry into effect legislative policies.”Footnote 17 In other words, the legal grounding in Humphrey’s is dubious insofar as the opinion tacitly admits that IRCs perform essentially executive functions but insists on categorizing these powers under a different name for the sake of prohibiting executive interference. At the same time, Jordan Cash (Reference Cash2020, 61–62) has argued that Sutherland’s decision in Humphrey’s was based in a “strong sense of how separation of powers should function” as well as a clear idea of “what constitutes executive power.” Thus, Cash provides evidence for taking Humphrey’s seriously as a constitutionally legitimate outline of presidential removal, and consequently gives legitimacy to the scholarly dispute about the unitary theory and the president’s use of the removal power.
Despite the contestation about the legitimacy of the constitutional reasoning in Humphrey’s, it was a unanimous decision that undoubtedly exercised significant influence over future jurisprudence. Indeed, Humphrey’s was reaffirmed and even perhaps expanded 23 years later in Wiener v. United States, a case in which the court unanimously decided that the Constitution did not grant President Dwight Eisenhower the authority to remove members of the War Claims Commission “merely because he wanted his own appointees on such a Commission.”Footnote 18 As Justice Felix Frankfurter argued in Wiener, the claim that Myers acknowledged a constitutional authority “to remove officials, no matter what the relation of the executive to the discharge of their duties and no matter what restrictions Congress may have imposed regarding the nature of their tenure” was “short lived.”Footnote 19 The result of Humphrey’s was to circumscribe the limits of the president’s removal power, and to legitimate the existence of IRCs insulated by statue from presidential removal.Footnote 20
Case Studies in the Court’s Approach to the Removal Power
Given that a genuine judicial affirmation of the unitary executive theory would require Humphrey’s to be overturned, a functional analysis of the development of the court’s approach to the removal power demands an examination of the most important cases in which there existed a real possibility of overturning Humphrey’s. Furthermore, conservative proponents of the unitary theory have long engaged in a concerted effort to decrease legal protections for administrative agencies and to “restore control” of the bureaucracy to the executive, often taking explicit aim at Humphrey’s as an obstacle to the full implementation of the theory, and thus their preferred legal outcomes.Footnote 21
The following cases have been selected for two reasons. First, they are often presented as foundational cases for legitimating or at least partially advancing the expansive removal power demanded by the unitary theory, especially in the time frame immediately after they were decided. They are thus “crucial” in the sense that they represent widely agreed-upon critical junctures in the court’s approach to the removal power (George and Bennett Reference George and Bennett2005, 20; Gerring Reference Gerring2007, 89, table 5.1). While the literature is in general disagreement about what theory of executive power these cases confirm or deny, their significance in this particular area remains uncontested.
Second, except for Bowsher v. Synar, each of these cases was decided by a largely conservative court—or at the very least a court composed of justices appointed by Republican presidents who favored an expansive removal power. This is especially true in the court’s three most recent removal cases. The court currently contains six justices who have deep ties to organizations like the Federalist Society—an institution designed to facilitate conservative policies through the court while also developing and honing a particularly conservative strain of constitutional interpretation (Hollis-Brusky Reference Hollis-Brusky2015, 147–64).Footnote 22 Furthermore, of the five justices appointed by President George W. Bush and President Trump, four have held significant positions in the executive branch under administrations that embraced the unitary theory (McMahon Reference McMahon2024, 187, table 9.3). Indeed, Chief Justice John Roberts—often thought of as a moderate in the Trump era—penned an internal memo during his time in the executive branch in which he claimed that “the time is ripe to reconsider the Constitutional anomaly of independent agencies,” evincing his clear support for more unilateral executive authority.Footnote 23 Roberts appears to have confirmed his past views in this area with his controversial decision in Trump v. United States, in which he outlines a rather expansive view of the president’s removal power in the Department of Justice.Footnote 24 Put simply, the court—as considered in these case studies—has demonstrated both an affiliation with unitarian actors, and a willingness to support unitarian interpretations of the Constitution.
These two factors suggest that the following five decisions discussed below represent a series of increasingly “most likely” cases for overturning Humphrey’s and confirming the unitary theory.Footnote 25 Despite these ideal circumstances, I find that the court has—surprisingly—been cautious about overturning the precedent set in Humphrey’s and confirming the unitary theory. To be sure, the court has been consistently willing to loosen the removal restrictions created by Humphrey’s. At the same time, however, it has also remained unwilling to overturn the case in its entirety, signaling a reluctance to fully confirm the unitary theory. Furthermore, while the development of the court’s jurisprudence has resulted in looser restrictions on the president’s removal power, the president is still denied the ability to unilaterally remove many bureaucrats from office—with even the Trump-era Roberts court remaining consistent in this area. Thus, while the second Trump administration has some evidence suggesting that the Roberts court might support an attempt to overturn Humphrey’s—and thus has pursued such outcomes in Trump v. Wilcox and Bessent v. Harris—it remains unlikely that the court will fully affirm the unitary theory.
Bowsher v. Synar and Morrison v. Olson
Two cases in the late 1980s represented the first substantive opportunities for overturning Humphrey’s. What makes them unique in this developmental analysis, however, is the fact that, rather than challenging the precedent set in Humphrey’s, both cases actually affirm it. Nevertheless, these cases remain important because they provide the basis for subsequent arguments in favor of a more expansive removal power, either through intricacies in the actual opinion, or noteworthy dissents. These two cases are examined together because of their proximity in time, and because neither contains enough significant information to constitute its own individual study.
In Bowsher v. Synar, the court ruled, in an opinion written by Chief Justice Warren Burger, that the powers delegated by Congress to the comptroller general violated the separation of powers.Footnote 26 Burger argued that “because Congress has retained removal authority over the Comptroller General, he may not be entrusted with executive powers,” and that despite this, Congress had delegated essentially executive powers to the comptroller general, a result that Burger found impermissible under the Constitution.Footnote 27 In Morrison v. Olson, the court held, in an opinion written by Chief Justice William Rehnquist, that the appointment of an independent counsel by a specially established court did not violate the separation of powers, despite the counsel’s exercise of executive power.Footnote 28
Despite the refusal of the court to reconsider Humphrey’s in both Bowsher and Morrison, both cases remain significant. First, scholars have argued that Bowsher opened the possibility of challenging the constitutionality of IRCs as such even if they considered such an outcome unlikely (Abikoff Reference Abikoff1987; Alvis, Bailey, and Taylor Reference Alvis, Bailey and Taylor2013; Gifford Reference Gifford1986; Hardie Reference Hardie1987; Verkuil Reference Verkuil1986).Footnote 29 Second, Justice Antonin Scalia’s dissent in Morrison would become a foundational text for future proponents of the unitary theory both on the court and off (Alvis, Bailey, and Taylor Reference Alvis, Bailey and Taylor2013, 203–7). In this dissent, Scalia indicates his support for the fundamental claim of the unitary theory by arguing that “it is not for [the court] to determine … how much of the purely executive powers of government must be within the full control of the President. The Constitution prescribes that they all are.”Footnote 30 At the same time, even Scalia’s spirited dissent fundamentally reaffirmed the basic principle in Humphrey’s—namely that the president only had authority to remove officials exercising clearly executive power. As such, neither decision represents in itself a substantive attack on Humphrey’s. At the same time, they serve as launching points for several broad developments in the court’s separation-of-powers jurisprudence.
Free Enterprise Fund v. Public Company Accounting Oversight Board
While Bowsher and Scalia’s Morrison make the legal affirmation of the unitary theory possible, the court seemed to shy away from a full-throated embrace of the unitary theory in its 2010 decision of Free Enterprise Fund v. Public Company Accounting Oversight Board. While this case certainly represented one of the court’s first significant reductions of the protections provided in Humphrey’s, it undeniably stopped short of overturning these restrictions in their entirety (Alvis, Bailey, and Taylor Reference Alvis, Bailey and Taylor2013, 208). According to Chief Justice Roberts, who authored the controlling opinion in Free Enterprise Fund, the structure of the Public Company Accounting Oversight Board (PCAOB) created a “dual for-cause limitation” that “withdraws from the President any decision on whether that good cause exists.”Footnote 31 As a result, the court declared the structure of the PCAOB unconstitutional. Roberts argued that the ramifications of finding the PCAOB’s structure constitutional would be to render bureaucratic agencies “immune from Presidential oversight, even as they exercised power in the people’s name.”Footnote 32 Such a structure, according to Roberts, “subverts the President’s ability to ensure that the laws are faithfully executed—as well as the public’s ability to pass judgment on his efforts.”Footnote 33 Interestingly, Roberts’s argument follows some of the key claims of the unitary theory. It emphasizes the importance of presidential responsibility, and the need to ensure that bureaucracies are accountable in a substantive way to elected officials.
Despite Roberts’s claims about the importance of agency accountability to the president however, he goes out of his way to preserve the precedents set by the court’s previous removal-power decisions—especially those established in Humphrey’s. Indeed, throughout the case, Roberts consistently limits his discussion to the constitutionality of removing “executive officers” only, implicitly accepting the distinction made in Humphrey’s and maintained since. Consequently, proponents of the unitary theory, writing in immediate response to the case, argued that Free Enterprise Fund “failed to establish any bright line for the constitutional role of independent agencies under the president’s control” (Stapler Reference Stapler2011, 715; see also Bader Reference Bader2009, 271). Others have argued that the decision is “critically deficient” in laying out a coherent path forward for subsequent removal-power jurisprudence (Jiang Reference Jiang2012, 732). At the same time, critics of the unitary theory argue that Free Enterprise Fund establishes the basic framework for the abolition of IRCs altogether, arguing that the decision “makes clear that there may be a gap between the court’s precedents [in Humphrey’s and other cases] and the Constitution” (Rao Reference Rao2011, 2543). Yet such claims seem largely unfounded. Robert’s decision in Free Enterprise Fund is far more concerned with congressional overreach than it is with defining the constitutional status of IRCs. Even critics of Free Enterprise Fund acknowledge that the ruling does not challenge the precedents established in Humphrey’s, and no meaningful concurrences exist that suggest a willingness to reevaluate these claims, even if critics are right to acknowledge that the court nods in the direction of Humphrey’s constitutional incoherence (Bellia Reference Bellia2012, 1376; Rao Reference Rao2011, 2543). Put simply, Free Enterprise Fund is a case that does not actually expand the president’s removal power but rather restricts the degree to which Congress can insulate bureaucratic agencies from presidential control. Insofar as Free Enterprise Fund deals with the removal power, it is focused primarily on placing limitations on congressional power in the interbranch conflict associated with bureaucracy, not on expanding or restricting the president’s ability to remove bureaucratic officials.
Seila Law LLC v. Consumer Financial Protection Bureau, Collins v. Yellen, and Consumers’ Research v. Consumer Product Safety Commission
It is thus clear that no court decision after Humphrey’s has attempted to circumvent or overturn the framework that case established. Presidents remain permitted to remove officials exercising purely executive power for any reason. Officials exercising quasi-legislative and quasi-judicial powers are, however, still protected from at-will removal. At the same time, all three of the landmark removal-power cases since Humphrey’s have contained the seeds for future reevaluations of the court’s jurisprudence, and even a genuine reversal of Humphrey’s. While Free Enterprise Fund did not challenge Humphrey’s directly, it seemed in many ways to set the stage for that case’s demise by subtly acknowledging the incoherence of the distinction between executive officials and officials exercising quasi-legislative and quasi-judicial power. Yet despite these felicitous circumstances, and the years of precedent that paved the way for a repudiation of Humphrey’s, the court failed to finally do so in Seila Law and Collins.
Such a failure is made more confusing by the fact that there has never been a court more likely to support a president’s agenda than the one that decided these cases (McMahon Reference McMahon2024). Not only were both cases decided by a court with a clear conservative majority, but two justices in the Seila Law majority and three in the Collins decision were appointed by Donald Trump, a president who has leaned upon the legal justifications of the unitary theory more than perhaps any other in history. Furthermore, of the new appointees, two have a history of hostility to either Humphrey’s specifically or IRCs in general. For example, Justice Neil Gorsuch penned both a controlling opinion and a concurrence for the Tenth Circuit Court, in which he criticizes the Supreme Court’s decision in Chevron v. NRDC and suggests that IRCs implicitly violate the separation of powers by concentrating legislative and judicial power in essentially executive agencies.Footnote 34 Similarly, Justice Brett Kavanaugh is known to have a history of explicit hostility toward both Humphrey’s and IRCs. While serving on the DC Circuit Court, Kavanaugh penned two opinions of note in this regard. The first was a concurrence in In re: Aiken County, in which he argues that the court’s decision in Humphrey’s could be “best explained” by the fact that the court was at the time “seemingly bent on resisting President Roosevelt and his New Deal policies,” subsequently refers to the decision as “egregious,” and heavily implies that it should be overturned like other cases decided around that time “as relics of an overly activist, anti-New Deal Supreme Court.”Footnote 35 Kavanaugh goes on, in that concurrence, to criticize a variety of regulatory agencies for their lack of accountability to the American public, making an argument essentially in line with those made by proponents of the unitary theory.Footnote 36 Most importantly, Kavanaugh draws attention to the fact that the court’s “rhetoric and reasoning” in Free Enterprise Fund are “notably in tension with Humphrey’s,” implying that the court has already set the stage for overturning that case.Footnote 37 The second was a dissent in PHH Corp. v. Consumer Financial Protection Bureau. In that opinion, he referred to independent agencies as “a headless fourth branch of government” and suggested that the exemptions carved out by Humphrey’s could only be applied to agencies headed by a nonpartisan body of experts rather than a single individual.Footnote 38 Even Roberts, the author of the court’s decision in Seila Law, had previously endorsed the unitary theory (Savage Reference Savage2007, 256). All in all, the court was in no better position to overturn Humphrey’s, and indeed, appeared to be poised to do so (Walker Reference Walker2020). Yet despite claims that Seila Law represented a “ringing endorsement of the unitary executive theory” by the court, it failed to explicitly overturn Humphrey’s (Skowronek, Dearborn, and King Reference Skowronek, Dearborn and King2022, 152).
Chief Justice Roberts, writing the controlling opinion for Seila Law, construes the decision as a refusal to extend the exceptions to the removal power granted by Humphrey’s to “an agency that wields significant executive power and is run by a single individual who cannot be removed by the president unless certain statutory criteria are met.”Footnote 39 The court thus held that that the “structure of the CFPB [Consumer Financial Protection Bureau] violates the separation of powers,” but neither suggested that the CFPB ought not exist nor that IRCs in general violate the Constitution.Footnote 40 In his opinion, Roberts seems comfortable with maintaining the exemptions established by Humphrey’s, so long as they were understood to “depend on the characteristics of the agency before the Court.”Footnote 41 Notably, however, Roberts adopts the framework present in Kavanaugh’s PHH Corp. dissent, arguing that one of the reasons the CFPB’s structure is unconstitutional is that it is not composed of a nonpartisan body of experts, and that it “is hardly merely a legislative or judicial aid” but rather an agency exercising significant executive powers.Footnote 42
As such, Roberts’s opinion may be justly understood as an attempt to confine and limit the effects of Humphrey’s—at least insofar as it further solidifies the limited nature of exemptions from presidential removal. There remains, however, a significant gulf between limiting the kinds of agencies and officials insulated from removal and doing away with the very possibility of protecting certain agencies and officers from removal. Seila Law can only be understood as an endorsement of the unitary executive theory insofar as it attempts to resolve certain ambiguities about the application of Humphrey’s that have arisen with the advent of a growing administrative state. Seila Law cannot, however, reasonably be construed as a decision that abolishes the exemptions outlined in Humphrey’s, or as one that abolishes agency independence as such.
Collins v. Yellen follows the reasoning of Seila Law directly, holding that recent legislation insulating the director of the Federal Housing Finance Agency from presidential removal violated the separation of powers.Footnote 43 The most important development provided by Collins is the way in which the majority opinion, written by Justice Samuel Alito, extends the effects of Seila Law by arguing that “the nature and breadth of an agency’s authority is not dispositive in determining whether Congress may limit the President’s power to remove its head.”Footnote 44 In essence, Alito suggests that, so long as an agency headed by a single individual is wielding executive power, no matter how insignificant or seemingly unimportant that power is, the agency must be answerable to the president. Significantly, this development still does not overturn Humphrey’s. IRCs with officials insulated from presidential removal remain essentially constitutional. At most, Collins augments the decision in Seila Law by enabling the president to remove agency directors who exercise any amount of executive power, rather than limiting removal to officials exercising “significant” executive powers.Footnote 45
Importantly, while the court has yet to fully hear more cases that build upon the logic of Seila Law and Collins, lower courts have. Indeed, recent lower-court jurisprudence suggests that the ramifications of these two cases are not nearly as damaging to agency independence as some feared. In Consumers’ Research et al. v. Consumer Product Safety Commission, the Fifth Circuit Court of Appeals decided that the existence of “for-cause removal” protections does not constitute a violation of the separation of powers, and that Seila Law did not overturn Humphrey’s. Footnote 46 Consequently, the Fifth Circuit decided that it was still obligated to enforce and apply Humphrey’s in all cases it may hear. Furthermore, despite the fact that the judge in Consumers’ Research practically begged the Supreme Court to overturn Humphrey’s, the court denied the plaintiffs in this case a writ of certiorari, implying a kind of finality to Seila Law and Collins. Footnote 47
This finality is, however, perhaps being reconsidered by the court. In a May 22, 2025, decision on Trump v. Wilcox et al., a recent emergency docket case regarding the president’s removal of multiple officials including members of the National Labor Relations Board (NLRB) and the Merit Systems Protection Board (MSPB), the court ruled in favor of the Trump administration’s request for a stay on the order of the DC District Court preventing the removal of two bureaucrats. This case is clearly another attempt by the Trump administration to force the court to overturn Humphrey’s Executor. Indeed, as the attorneys for Cathy A. Harris (one of the bureaucrats contesting their removal) have argued, allowing these removals would demand that the court overturn Humphrey’s. Footnote 48 While the court’s May 22, 2025, decision on this case is not a final decision on the legality of these removals, it does indicate the potential for the court to expand the removal power even further than it already has. Notably, however, the unsigned majority opinion is careful to frame its decision in such a way that the letter—if not the spirit—of Humphrey’s could be preserved. Thus, while Justice Elena Kagan rightly points out in her dissent that the majority neglects to mention Humphrey’s by name, the decision is carefully worded so as to emphasize the “considerable executive power” exercised by these bureaucrats, and thus indicates that their offices sit outside the removal protections Humphrey’s provides.Footnote 49
In summation, despite the fact that there are many conservative judges and legal scholars—Justices Neil Gorsuch and Clarence Thomas among them—who sincerely believe that “[t]he decision in Humphrey’s poses a direct threat to our constitutional structure,”Footnote 50 the overall results of recent removal-power jurisprudence are almost shockingly limited in nature (Grossman and Sandoloski Reference Grossman and Sandoloski2021, 223–25). This is, of course, not to say that Seila Law and Collins hold no legal significance. Together, these cases represent the most significant changes to the court’s approach to the removal power since Humphrey’s. By limiting the scope of the exemptions to removal provided by Humphrey’s to only those agencies made up of multimember, politically balanced boards, these cases take significant steps in the direction of restoring a version of the removal power conceived of in Myers. At the same time, these are only a few steps down a rather long road—a road that the current court, despite its obvious conservative affiliation, seems reluctant to continue traveling.Footnote 51 The version of the removal power necessary for a full-throated affirmation of the unitary theory ultimately requires that presidents be permitted to execute at-will removal of any official they are involved in appointing. While no committed unitarian would balk at the results of Seila Law and Collins, they would also acknowledge that these cases are not a true legal sanction of their theory. Indeed, the most shocking aspect of these decisions is their relatively limited significance in the grand scheme of removal-power jurisprudence given the composition of the court that decided them.
Explaining the Court’s Decisions
Simply because the court has not granted full legal sanction to the unitary theory does not mean, however, that it has repudiated it. As discussed above, it seems that Seila Law and Collins are examples of the court taking halting steps toward a more robust interpretation of the removal power, similar to those taken earlier in Bowsher, Morrison, and Free Enterprise Fund. It remains puzzling, however, that the court has chosen to continue limiting the exemptions provided in Humphrey’s rather than overturning the case in full. It seems that legal precedent in favor of overturning Humphrey’s had been building for years, with the court edging ever closer to doing so since its decision in Bowsher. Why then, has the court failed to do so? I offer two tentative explanations, both of which can be true without contradicting the other.
One explanation for the court’s behavior is that the conservative majority determined that they could effectively increase the president’s ability to use the removal power as a political tool without providing full legal sanction to the unitary theory. In doing so, justices would be able to make decisions on the basis of their preexisting political attitudes while also ensuring that future presidents who do not share their political proclivities can be strategically blocked from exercising the same powers (Segal and Spaeth Reference Segal and Spaeth2008). Similarly, the court’s decisions in Seila Law and Collins fit within the regime model of judicial behavior, in which the court is part of a dominant governing coalition (Bridge Reference Bridge2015, 422–25; Dahl Reference Dahl1957; Graber Reference Graber1993; Whittington Reference Whittington2005). According to this approach to judicial politics, the court exercises judicial review in ways that benefit the governing coalition to which the majority of its members belong. As such, justices do not necessarily decide on the basis of their own political attitudes, but rather on the basis of their coalition’s policy goals. Thus, Seila Law and Collins are simply instances of the court serving the needs of the governing coalition—in this case the Republican Party led by Donald Trump.
Even prior to the decisions in Seila Law and Collins, presidents were clearly capable of using the removal power for political purposes. A study carried out by the Brookings Institution indicates that Trump made effective use of the removal power to control both his cabinet officials and key officials in the executive office of the president during his first term, including the time before Seila Law and Collins were decided (Tenpas Reference Tenpas2021). That same study indicates that previous presidents were similarly capable of using the removal power to retain control over their administrations in this way. All these removals were justified even under the most expansive reading of Humphrey’s protections because they were removals of officials directly and unambiguously answerable to the president.
Approaching the court’s decisions from the perspective of these models also helps to explain why Humphrey’s has not yet been overturned. Under these models, the court, as a partisan actor, is concerned both with providing policy victories for a particular political coalition and obstructing the ends of the opposing coalition. As such, the court is incentivized to decide on narrow grounds so as to limit the possible ways in which the opposing coalition can benefit from its decisions. By maintaining Humphrey’s, the court sets the stage for a case-by-case evaluation of removals in which its members can decide in the moment whether to allow them, thus performing its role within the dominant coalition without abdicating its power in the future.Footnote 52
At the same time, however, granting these victories to its coalition comes with certain trade-offs. While the Seila Law and Collins decisions preserve the court’s power to review instances of presidential removal in the future, they do so on a more limited ground, and in such a way that grants an undiscerning expansion of presidential power. As such, members of the court have been unable to both provide victories for their coalition and fully prevent the exploitation of their decisions by opposing parties, even if the potential for future limitations remains. For example, President Joe Biden’s removal of the general counsel of the NLRB was upheld in a ruling by the Ninth Circuit Court, which cited Seila Law as the controlling precedent.Footnote 53 In 2021 Biden fired Administrative Conference of the United States council member Roger Severino, and had that removal upheld in court based on the decision in Seila Law. Footnote 54 The same year, Biden also fired Social Security Commissioner Andrew Saul.Footnote 55 The Justice Department argued, in a memo to the president, that the decisions in Seila Law and Collins rendered the commissioner’s statutory removal protections moot.Footnote 56 In each case, a previously questionable removal was justified because of the court’s limitation of Humphrey’s in Seila Law, Collins, or both.
It is also possible, however, to explain the court’s behavior as a result of the way that the Constitution structures power between the three branches of government. When discussing the separation of powers in Federalist no. 51, Publius explains that the Constitution arranges power in such a way that the people occupying a particular office of government will be compelled to defend their institution’s constitutional position (Hamilton, Jay, and Madison [1818] Reference Hamilton, Jay, Madison, Carey and McClellan2001, 268). For Publius, being selected for office compels the individual to adopt an “institutional perspective” on politics that requires them to view politics from the perspective of their own office, with a special eye toward the preservation and perpetuation of their office’s constitutional position.Footnote 57 When officeholders adopt a particular institutional perspective, their political positions can shift to accord with their new view of the political landscape (Cash Reference Cash2023, 47). Examining the court’s decisions in Seila Law and Collins with this in mind, the court’s behavior can be explained as a result of its institutional perspective. Had the court decided to overturn Humphrey’s, it would have hamstrung its own ability to further adjudicate removal questions. In doing so, the court would have significantly decreased its own power and increased that of the president—something that it is simply not constitutionally incentivized to do. Furthermore, it would have become more difficult for the court to weigh in on adjacent separation-of-powers cases that might arise in conjunction with the proliferation of the unitary theory. Thus, conservative justices who may have otherwise preferred to grant the unitary theory a legal stamp of approval were compelled by their institutional perspective to stop short. This allowed them to rule in a way that delivered positive outcomes for their coalition while also preserving their institutional power and position.
Justices Roberts and Kavanaugh’s flip-flops on overturning Humphrey’s presents compelling evidence in favor of this explanation. As discussed above, both Roberts and Kavanaugh had previous, well-documented records of hostility toward Humphrey’s during their prior government service. Yet as a Supreme Court justice, Roberts passed up the opportunity to finally reevaluate the “Constitutional anomaly” of independent agencies three separate times. Kavanaugh joined him in one of those decisions, despite his rather explicit skepticism of IRCs and Humphrey’s. If Roberts thought that “the time [was] ripe” to reevaluate the constitutionality of IRCs in 1983, why would he pass up not one, not two, but three opportunities to overturn Humphrey’s? Indeed, of the three most recent removal cases, Roberts penned the controlling opinion in both Free Enterprise Fund and Seila Law, and in doing so went out of his way both times to affirm Humphrey’s. Similarly, Kavanaugh’s decision to join Roberts’s pro-Humphrey’s opinion in Seila Law is made all the more significant given the fact that Justices Thomas and Gorsuch penned a concurrence far more hostile to Humphrey’s that Kavanaugh refused to join. It appears to be the case that Roberts’s and Kavanaugh’s new institutional perspectives caused a genuine shift in their positions, one that ultimately resulted in a more moderate ruling than would otherwise have been expected.
Importantly, both of these explanations can be true. It is entirely possible that the court made a prudential decision about what outcome would most effectively serve the interests of the coalition to which it belonged, and that the ruling of its justices was influenced by their institutional perspective. By refusing to grant legal sanction to the unitary theory, the majority of the court was able to serve both its partisan interests and preserve the court’s particular constitutional powers. Indeed, taking both explanations seriously better explains the court’s behavior than either one alone. If the institutional perspective of the court was controlling, then we would have likely not seen Justices Thomas and Gorsuch’s concurrence. At the same time, if political attitudes or coalitional affiliation were controlling, Roberts’s and Kavanaugh’s changes of position are difficult to explain. Thus, while both possible explanations for the court’s behavior individually provide adequate and complete explanations of their behavior, together they provide the most comprehensive explanation possible.
New Data and Anomalous Possibilities
Models for understanding the behavior of political institutions are always subject to change when new variables arise. Since the court has now signaled in Trump v. Wilcox et al. a willingness to further expand the removal power, it is clear that things have changed since the court’s refusal to hear Consumers’ Research in 2024. It is consequently necessary to explain the new variables at play, and determine the consequences that they may have, both in general and on the behavioral explanations provided above. Ultimately, I conclude that the court is in a unique and historically significant position regarding the removal power. With only four options—outlined in table 1—for deciding Gwynne Wilcox’s case when it comes to them in the future, the court is poised to make a decision that has the potential to either preserve a tenuous but acceptable status quo or place the future of judicial and executive power in uncharted—and arguably dangerous—waters, creating a situation that may even pose a threat to the separation of powers as such.
Possible Outcomes
To be sure, the court has considerable constitutional incentive not to overturn Humphrey’s. Indeed, it is manifestly unclear what the court would gain by doing so that would justify significantly narrowing its own authority. Furthermore, not enough time has passed between the court’s decision in Seila Law and now to reasonably explain why a court that had previously chosen to preserve Humphrey’s would change its mind, even given the new political situation. There are no new justices on the court and despite the Roberts court’s notably loose conception of stare decisis, such a radical shift in attitude would be difficult to explain. However, by staying the injunction against Trump’s recent removals from the NLRB and MSPB, the court has suggested it believes that “the Government is likely to show that both the NLRB and MSPB exercise considerable executive power.”Footnote 58 Thus, while the court has not definitely signaled that it is about to overturn Humphrey’s, it has at least indicated a willingness to narrow removal protections even further. As such, there are two ways the court could decide if it chooses to preserve Humphrey’s. In both of these scenarios, however, the court would preserve its authority to participate in interbranch constitutional conflict over the removal power. Consequently, it would be acting in line with its constitutional incentives, demonstrating the continued viability of the Constitution and the separation-of-powers system.
First, the court could rule in Wilcox’s favor, declaring her removal unconstitutional, and preserving the status quo established in Seila Law and affirmed in Collins and Consumers’ Research. Such an outcome would be optimal for preserving the court’s constitutional authority and demonstrating the strength of the separation-of-powers system. This possibility seems unlikely, however, given the court’s statements in the Wilcox decision. Second, the court could choose to rule in favor of the government—as its decision in Wilcox indicates is likely—but do so without overturning Humphrey’s. In this scenario, the court acts in a way that is consistent with the historical pattern established above, providing a short-term policy victory for the affiliated coalition of its conservative majority, but still preserves its overarching constitutional authority, affirming the continued efficacy of the separation of powers. Of course, the more the court limits the removal protections Humphrey’s provides, the more it limits its ability to check the president. Thus, while this outcome would be consistent with the status quo, and be within the scope of judicial action incentivized by the Constitution, it would still result in a more limited court.
Things have changed significantly enough since Seila Law was decided, however, that it may no longer be reasonable to suggest that the court will continue to abide by the status quo. When the court denied certiorari for Consumers’ Research in October of 2024, the presidential election was still a month away, and the outcome was far from certain. In the absence of an assured Republican victory, the court’s conservatives may have simply decided that it was too risky to hear the case and overturn Humphrey’s when doing so might give Democratic presidential nominee Kamala Harris unrestrained removal power. Trump’s subsequent victory may have changed things in the minds of the court’s conservatives, especially given the willingness Trump has shown to implement the vision laid out in “Project 2025.”Footnote 59 Empowering Trump to remove any and all bureaucrats at will would allow him to fundamentally reshape American government—a goal important enough to conservatives that it may outweigh any fears of future Democratic presidents wielding similar authority. As such, the court’s conservatives may be willing to overturn Humphrey’s when they finally hear Wilcox’s case in full, since they have a strong guarantee that doing so would fulfill their coalition’s immediate and long-term goals. This scenario leads to two potential outcomes, both of which would go against the court’s constitutional incentives, and demonstrate a degree of weakness in the Constitution’s separation-of-powers system.
In one scenario the court’s conservatives could finally overturn Humphrey’s Executor, destroying all removal protections for independent agencies and granting the president the legal authority to pursue the destruction of the federal bureaucracy with impunity. Such a decision would fully exclude the court from ruling on removal cases in the future. In doing so, the court would abdicate any constitutional authority it may have had to participate in interbranch debates surrounding the separation of powers in matters of bureaucratic removal. This outcome would be extremely problematic for the constitutional order, as it would severely limit the court’s power to check the president and exclude it from a significant role within the separation of powers. It would suggest a significant weakness in the Constitution, and call into question the continued efficacy of the separation of powers. Furthermore, it would be historically unusual, given the court’s steadfast refusal to overturn Humphrey’s over the past four decades. Still, it is a possible outcome, especially given the fact that the current conservatives on the Supreme Court are so thoroughly steeped in the anti-bureaucratic ideology of the Federalist Society (McMahon Reference McMahon2024, 186).
As lawyers in both Wilcox and Harris have argued, however, a ruling overturning Humphrey’s Executor in its entirety would potentially enable the president to remove members of any multimember independent agency, including the chairman of the Federal Reserve.Footnote 60 Such a situation is one that even the court’s conservatives appear uncomfortable with, leading them to write in the unsigned Wilcox decision that “[t]he 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.”Footnote 61 In a previous case, Justice Alito argued that exemptions to presidential removal could be carved out for the Federal Reserve even if Humphrey’s was overturned.Footnote 62 Importantly, the court’s decision in Wilcox suggests the independence of the Federal Reserve is entirely unrelated to the removal protections provided by Humphrey’s, which are instead rooted in the organization’s structure and similarity to the earlier national banks. Thus, the court’s conservatives appear divided, with some seeming to think that the Federal Reserve is exempt from presidential removal because of its similarity to historically constitutional entities, and others thinking that specific removal protections would need to be affirmed to preserve its independence. In any case, these comments suggest the possibility of another scenario in which the court overturns Humphrey’s but explicitly acknowledges the independence of the Federal Reserve. The constitutional consequences of this potential outcome would be largely identical to those of the first scenario, with the caveat that the court would preserve its authority to adjudicate removal disputes if an involved party is part of the Federal Reserve. Such a decision would still be historically anomalous and contrary to the court’s constitutional incentives but would be—at least slightly—less damaging to the separation of powers, as the court would not fully abandon its ability to adjudicate all removals.
The various possibilities for the court’s approach to the removal power are succinctly illustrated in table 1. While not all outcomes are equally likely, they each present unique and interesting possibilities for the court’s approach to the removal power. Some of these outcomes reaffirm the strength and salience of the Constitution and its incentives for real politics. Others indicate constitutional weakness and even suggest a breakdown of the separation of powers.
Conclusion
The result of this examination of the court’s removal-power jurisprudence leads to the following conclusions. First, the unitary theory demands an expansive removal power akin to the interpretation proposed in Myers. This interpretation of the removal power was explicitly denied in Humphrey’s, which placed significant limitations on the president’s use of the removal power. This decidedly nonunitarian version of the removal power remained uncontested in the court’s jurisprudence for nearly 50 years. While the court began to set the stage for overturning Humphrey’s in Bowsher and continued to do so in Morrison and Free Enterprise Fund, it passed up the opportunity to take advantage of this in Seila Law and Collins despite being in an ideal position to do so. As such, scholars who construe Seila Law and Collins as full-throated affirmations of the unitary theory overstate the effects of these cases. Similarly, any attempts by the court to affirm the unitary theory would be historically anomalous.
Second, the court possesses both political and constitutional incentives that affect the choices it makes. The court is both politically incentivized to expand the removal power to serve its affiliated coalition and constitutionally incentivized to preserve some degree of agency independence. Historically, it has navigated this tension by expanding the removal power while going out of its way to preserve the central holding of Humphrey’s Executor. This allowed the court to both deliver victories for its affiliated coalition and preserve its constitutional authority.
Finally, the court is poised to make an extremely significant decision regarding the nature of the president’s removal power. The decision that it makes will indicate the resiliency of the constitutional order as well as the functionality of the separation of powers. On one hand, if the court chooses to overturn Humphrey’s it risks abdicating significant aspects of its own authority, deeply changing the status quo, and undermining the separation of powers. Variations exist within this possibility, as the court has the option to overturn Humphrey’s while still preserving the independence of at least the Federal Reserve. Such an outcome, however, would be only slightly less damaging to the constitutional order. On the other hand, the court could choose to uphold Humphrey’s and in doing so affirm the strength of the Constitution and the efficacy of the separation of powers. Again, variations exist here as well, with the court being able to either stop the removal of Wilcox and other bureaucrats entirely or permit their removal without overturning the central holding of Humphrey’s. No matter what the court decides, however, this reflection provides both a comprehensive view of the development and trajectory of the court’s approach to the removal power and a detailed explanation of the potential impact of each of the four possible ways the court can decide in the future.
Open access
