Seila Law LLC v. Consumer Financial Protection Bureau, 591 U.S. ___ (2020)

Docket No. 19-7
Granted: October 18, 2019
Argued: March 3, 2020
Decided: June 29, 2020
Justia Summary

Following the 2008 financial crisis, the Consumer Financial Protection Bureau (CFPB), was established by the Dodd-Frank Act as an independent regulatory agency tasked with ensuring that consumer debt products are safe and transparent. The administration of 18 existing federal statutes was transferred to CFPB. A new prohibition on unfair and deceptive practices in the consumer-finance sector, 12 U.S.C. 5536(a)(1)(B), gave CFPB extensive rulemaking, enforcement, and adjudicatory powers, including the authority to conduct investigations, issue subpoenas and civil investigative demands, initiate administrative adjudications, prosecute civil actions in federal court, and issue binding decisions in administrative proceedings. CFPB is led by a single Director, appointed by the President with the advice and consent of the Senate, for a five-year term, during which the President may remove the Director only for “inefficiency, neglect of duty, or malfeasance,” 12 U.S.C. 5491(c)(1),(3).

Primary Holding

The Consumer Financial Protection Bureau’s leadership by a single Director removable only for inefficiency, neglect, or malfeasance violates the separation of powers.


NOTE: Where it is feasible, a syllabus (headnote) will be released, as is being done in connection with this case, at the time the opinion is issued. The syllabus constitutes no part of the opinion of the Court but has been prepared by the Reporter of Decisions for the convenience of the reader. See United States v. Detroit Timber & Lumber Co., 200 U.S. 321, 337.



Seila Law LLC v. Consumer Financial Protection Bureau

certiorari to the united states court of appeals for the ninth circuit

No. 19–7. Argued March 3, 2020—Decided June 29, 2020

NOTICE: This opinion is subject to formal revision before publication in the preliminary print of the United States Reports. Readers are requested to notify the Reporter of Decisions, Supreme Court of the United States, Washington, D. C. 20543, of any typographical or other formal errors, in order that corrections may be made before the preliminary print goes to press.



No. 19–7



on writ of certiorari to the united states court of appeals for the ninth circuit

[June 29, 2020]

Chief Justice Roberts delivered the opinion of the Court with respect to Parts I, II, and III.

In the wake of the 2008 financial crisis, Congress established the Consumer Financial Protection Bureau (CFPB), an independent regulatory agency tasked with ensuring that consumer debt products are safe and transparent. In organizing the CFPB, Congress deviated from the structure of nearly every other independent administrative agency in our history. Instead of placing the agency under the leadership of a board with multiple members, Congress provided that the CFPB would be led by a single Director, who serves for a longer term than the President and cannot be removed by the President except for inefficiency, neglect, or malfeasance. The CFPB Director has no boss, peers, or voters to report to. Yet the Director wields vast rulemaking, enforcement, and adjudicatory authority over a significant portion of the U. S. economy. The question before us is whether this arrangement violates the Constitution’s separation of powers.

Under our Constitution, the “executive Power”—all of it—is “vested in a President,” who must “take Care that the Laws be faithfully executed.” Art. II, §1, cl. 1; id., §3. Because no single person could fulfill that responsibility alone, the Framers expected that the President would rely on subordinate officers for assistance. Ten years ago, in Free Enterprise Fund v. Public Company Accounting Oversight Bd., 561 U.S. 477 (2010), we reiterated that, “as a general matter,” the Constitution gives the President “the authority to remove those who assist him in carrying out his duties,” id., at 513–514. “Without such power, the President could not be held fully accountable for discharging his own responsibilities; the buck would stop somewhere else.” Id., at 514.

The President’s power to remove—and thus supervise—those who wield executive power on his behalf follows from the text of Article II, was settled by the First Congress, and was confirmed in the landmark decision Myers v. United States, 272 U.S. 52 (1926). Our precedents have recognized only two exceptions to the President’s unrestricted removal power. In Humphrey’s Executor v. United States, 295 U.S. 602 (1935), we held that Congress could create expert agencies led by a group of principal officers removable by the President only for good cause. And in United States v. Perkins, 116 U.S. 483 (1886), and Morrison v. Olson, 487 U.S. 654 (1988), we held that Congress could provide tenure protections to certain inferior officers with narrowly defined duties.

We are now asked to extend these precedents to a new configuration: an independent 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. We decline to take that step. While we need not and do not revisit our prior decisions allowing certain limitations on the President’s removal power, there are compelling reasons not to extend those precedents to the novel context of an independent agency led by a single Director. Such an agency lacks a foundation in historical practice and clashes with constitutional structure by concentrating power in a unilateral actor insulated from Presidential control.

We therefore hold that the structure of the CFPB violates the separation of powers. We go on to hold that the CFPB Director’s removal protection is severable from the other statutory provisions bearing on the CFPB’s authority. The agency may therefore continue to operate, but its Director, in light of our decision, must be removable by the President at will.



In the summer of 2007, then-Professor Elizabeth Warren called for the creation of a new, independent federal agency focused on regulating consumer financial products. Warren, Unsafe at Any Rate, Democracy (Summer 2007). Professor Warren believed the financial products marketed to ordinary American households—credit cards, student loans, mortgages, and the like—had grown increasingly unsafe due to a “regulatory jumble” that paid too much attention to banks and too little to consumers. Ibid. To remedy the lack of “coherent, consumer-oriented” financial regulation, she proposed “concentrat[ing] the review of financial products in a single location”—an independent agency modeled after the multimember Consumer Product Safety Commission. Ibid.

That proposal soon met its moment. Within months of Professor Warren’s writing, the subprime mortgage market collapsed, precipitating a financial crisis that wiped out over $10 trillion in American household wealth and cost millions of Americans their jobs, their retirements, and their homes. In the aftermath, the Obama administration embraced Professor Warren’s recommendation. Through the Treasury Department, the administration encouraged Congress to establish an agency with a mandate to ensure that “consumer protection regulations” in the financial sector “are written fairly and enforced vigorously.” Dept. of Treasury, Financial Regulatory Reform: A New Foundation 55 (2009). Like Professor Warren, the administration envisioned a traditional independent agency, run by a multimember board with a “diverse set of viewpoints and experiences.” Id., at 58.

In 2010, Congress acted on these proposals and created the Consumer Financial Protection Bureau (CFPB) as an independent financial regulator within the Federal Reserve System. Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank), 124Stat. 1376. Congress tasked the CFPB with “implement[ing]” and “enforc[ing]” a large body of financial consumer protection laws to “ensur[e] that all consumers have access to markets for consumer financial products and services and that markets for consumer financial products and services are fair, transparent, and competitive.” 12 U. S. C. §5511(a). Congress transferred the administration of 18 existing federal statutes to the CFPB, including the Fair Credit Reporting Act, the Fair Debt Collection Practices Act, and the Truth in Lending Act. See §§5512(a), 5481(12), (14). In addition, Congress enacted a new prohibition on “any unfair, deceptive, or abusive act or practice” by certain participants in the consumer-finance sector. §5536(a)(1)(B). Congress authorized the CFPB to implement that broad standard (and the 18 pre-existing statutes placed under the agency’s purview) through binding regulations. §§5531(a)–(b), 5581(a)(1)(A), (b).

Congress also vested the CFPB with potent enforcement powers. The agency has the authority to conduct investigations, issue subpoenas and civil investigative demands, initiate administrative adjudications, and prosecute civil actions in federal court. §§5562, 5564(a), (f ). To remedy violations of federal consumer financial law, the CFPB may seek restitution, disgorgement, and injunctive relief, as well as civil penalties of up to $1,000,000 (inflation adjusted) for each day that a violation occurs. §§5565(a), (c)(2); 12 CFR §1083.1(a), Table (2019). Since its inception, the CFPB has obtained over $11 billion in relief for over 25 million consumers, including a $1 billion penalty against a single bank in 2018. See CFPB, Financial Report of the Consumer Financial Protection Bureau, Fiscal Year 2015, p. 3; CFPB, Bureau of Consumer Financial Protection Announces Settlement With Wells Fargo for Auto-Loan Administration and Mortgage Practices (Apr. 20, 2018).

The CFPB’s rulemaking and enforcement powers are coupled with extensive adjudicatory authority. The agency may conduct administrative proceedings to “ensure or enforce compliance with” the statutes and regulations it administers. 12 U. S. C. §5563(a). When the CFPB acts as an adjudicator, it has “jurisdiction to grant any appropriate legal or equitable relief.” §5565(a)(1). The “hearing officer” who presides over the proceedings may issue subpoenas, order depositions, and resolve any motions filed by the parties. 12 CFR §1081.104(b). At the close of the proceedings, the hearing officer issues a “recommended decision,” and the CFPB Director considers that recommendation and “issue[s] a final decision and order.” §§1081.400(d), 1081.402(b); see also §1081.405.

Congress’s design for the CFPB differed from the proposals of Professor Warren and the Obama administration in one critical respect. Rather than create a traditional independent agency headed by a multimember board or commission, Congress elected to place the CFPB under the leadership of a single Director. 12 U. S. C. §5491(b)(1). The CFPB Director is appointed by the President with the advice and consent of the Senate. §5491(b)(2). The Director serves for a term of five years, during which the President may remove the Director from office only for “inefficiency, neglect of duty, or malfeasance in office.” §§5491(c)(1), (3).

Unlike most other agencies, the CFPB does not rely on the annual appropriations process for funding. Instead, the CFPB receives funding directly from the Federal Reserve, which is itself funded outside the appropriations process through bank assessments. Each year, the CFPB requests an amount that the Director deems “reasonably necessary to carry out” the agency’s duties, and the Federal Reserve grants that request so long as it does not exceed 12% of the total operating expenses of the Federal Reserve (inflation adjusted). §§5497(a)(1), (2)(A)(iii), 2(B). In recent years, the CFPB’s annual budget has exceeded half a billion dollars. See CFPB, Fiscal Year 2019: Ann. Performance Plan and Rep., p. 7.


Seila Law LLC is a California-based law firm that provides debt-related legal services to clients. In 2017, the CFPB issued a civil investigative demand to Seila Law to determine whether the firm had “engag[ed] in unlawful acts or practices in the advertising, marketing, or sale of debt relief services.” 2017 WL 6536586, *1 (CD Cal., Aug. 25, 2017). See also 12 U. S. C. §5562(c)(1) (authorizing the agency to issue such demands to persons who “may have any information[ ] relevant to a violation” of one of the laws enforced by the CFPB). The demand (essentially a subpoena) directed Seila Law to produce information and documents related to its business practices.

Seila Law asked the CFPB to set aside the demand, objecting that the agency’s leadership by a single Director removable only for cause violated the separation of powers. The CFPB declined to address that claim and directed Seila Law to comply with the demand.

When Seila Law refused, the CFPB filed a petition to enforce the demand in the District Court. See §5562(e)(1) (creating cause of action for that purpose). In response, Seila Law renewed its defense that the demand was invalid and must be set aside because the CFPB’s structure violated the Constitution. The District Court disagreed and ordered Seila Law to comply with the demand (with one modification not relevant here).

The Court of Appeals affirmed. 923 F.3d 680 (CA9 2019). The Court observed that the “arguments for and against” the constitutionality of the CFPB’s structure had already been “thoroughly canvassed” in majority, concurring, and dissenting opinions by the en banc Court of Appeals for the District of Columbia Circuit in PHH Corp. v. CFPB, 881 F.3d 75 (2018), which had rejected a challenge similar to the one presented here. 923 F. 3d, at 682. The Court saw “no need to re-plow the same ground.” Ibid. Instead, it provided a brief explanation for why it agreed with the PHH Court’s core holding. The Court took as its starting point Humphrey’s Executor, which had approved for-cause removal protection for the Commissioners of the Federal Trade Commission (FTC). In applying that precedent, the Court recognized that the CFPB wields “substantially more executive power than the FTC did back in 1935” and that the CFPB’s leadership by a single Director (as opposed to a multimember commission) presented a “structural difference” that some jurists had found “dispositive.” 923 F. 3d, at 683–684. But the Court felt bound to disregard those differences in light of our decision in Morrison, which permitted a single individual (an independent counsel) to exercise a core executive power (prosecuting criminal offenses) despite being insulated from removal except for cause. Because the Court found Humphrey’s Executor and Morrison “controlling,” it affirmed the District Court’s order requiring compliance with the demand. 923 F. 3d, at 684.

We granted certiorari to address the constitutionality of the CFPB’s structure. 589 U. S. ___ (2019). We also requested argument on an additional question: whether, if the CFPB’s structure violates the separation of powers, the CFPB Director’s removal protection can be severed from the rest of the Dodd-Frank Act.

Because the Government agrees with petitioner on the merits of the constitutional question, we appointed Paul Clement to defend the judgment below as amicus curiae. He has ably discharged his responsibilities.


We first consider three threshold arguments raised by the appointed amicus for why we may not or should not reach the merits. Each is unavailing.

First, amicus argues that the demand issued to petitioner is not “traceable” to the alleged constitutional defect because two of the three Directors who have in turn played a role in enforcing the demand were (or now consider themselves to be) removable by the President at will. Brief for Court-Appointed Amicus Curiae 21–24. Amicus highlights the Government’s argument below that the demand, originally issued by former Director Richard Cordray, had been ratified by an acting CFPB Director who, according to the Office of Legal Counsel (OLC), was removable by the President at will. See Brief for Appellee in No. 17–56324 (CA9), pp. 1, 10, 13–19 (citing Designating an Acting Director of the Bureau of Consumer Financial Protection, 41 Op. OLC ___, ___ (Nov. 25, 2017)). Amicus further observes that current CFPB Director Kathleen Kraninger, now responsible for enforcing the demand, agrees with the Solicitor General’s position in this case that her for-cause removal protection is unconstitutional. See Brief for Respondent on Pet. for Cert. 20; Letter from K. Kraninger, CFPB Director, to M. McConnell, Majority Leader, U. S. Senate, p. 2 (Sept. 17, 2019); Letter from K. Kraninger, CFPB Director, to N. Pelosi, Speaker, U. S. House of Representatives, p. 2 (Sept. 17, 2019).[1] In amicus’ view, these developments reveal that the demand would have been issued—and would continue to be enforced—even in the absence of the CFPB Director’s removal protection, making the asserted separation of powers dispute “artificial.” Brief for Court-Appointed Amicus Curiae 22.

Even if that were true, it would not deprive us of jurisdiction. Amicus’ traceability argument appears to challenge petitioner’s Article III standing. See Lujan v. Defenders of Wildlife, 504 U.S. 555, 560 (1992) (explaining that the plaintiff ’s injury must be “fairly traceable to the challenged action of the defendant” (internal quotation marks and alterations omitted)). But amicus’ argument does not cast any doubt on the jurisdiction of the District Court because petitioner is the defendant and did not invoke the Court’s jurisdiction. See Bond v. United States, 564 U.S. 211, 217 (2011) (When the plaintiff has standing, “Article III does not restrict the opposing party’s ability to object to relief being sought at its expense.”).

It is true that “standing must be met by persons seeking appellate review, just as it must be met by persons appearing in courts of first instance.” Hollingsworth v. Perry, 570 U.S. 693, 705 (2013) (internal quotation marks omitted). But petitioner’s appellate standing is beyond dispute. Petitioner is compelled to comply with the civil investigative demand and to provide documents it would prefer to withhold, a concrete injury. That injury is traceable to the decision below and would be fully redressed if we were to reverse the judgment of the Court of Appeals and remand with instructions to deny the Government’s petition to enforce the demand.

Without engaging with these principles, amicus contends that a litigant wishing to challenge an executive act on the basis of the President’s removal power must show that the challenged act would not have been taken if the responsible official had been subject to the President’s control. See Brief for Court-Appointed Amicus Curiae 21–24. Our precedents say otherwise. We have held that a litigant challenging governmental action as void on the basis of the separation of powers is not required to prove that the Government’s course of conduct would have been different in a “counterfactual world” in which the Government had acted with constitutional authority. Free Enterprise Fund, 561 U. S., at 512, n. 12. In the specific context of the President’s removal power, we have found it sufficient that the challenger “sustain[s] injury” from an executive act that allegedly exceeds the official’s authority. Bowsher v. Synar, 478 U.S. 714, 721 (1986).

Second, amicus contends that the proper context for assessing the constitutionality of an officer’s removal restriction is a contested removal. See Brief for Court-Appointed Amicus Curiae 24–27. While that is certainly one way to review a removal restriction, it is not the only way. Our precedents have long permitted private parties aggrieved by an official’s exercise of executive power to challenge the official’s authority to wield that power while insulated from removal by the President. See Bowsher, 478 U. S., at 721 (lawsuit filed by aggrieved third party in the absence of contested removal); Free Enterprise Fund, 561 U. S., at 487 (same); Morrison, 487 U. S., at 668–669 (defense to subpoena asserted by third party in the absence of contested removal). Indeed, we have expressly “reject[ed]” the “argument that consideration of the effect of a removal provision is not ‘ripe’ until that provision is actually used,” because when such a provision violates the separation of powers it inflicts a “here-and-now” injury on affected third parties that can be remedied by a court. Bowsher, 478 U. S., at 727, n. 5 (internal quotation marks omitted). The Court of Appeals therefore correctly entertained petitioner’s constitutional defense on the merits.

Lastly, amicus contends that we should dismiss the case because the parties agree on the merits of the constitutional question and the case therefore lacks “adverseness.” Tr. of Oral Arg. 42–43, 45–46. That contention, however, is foreclosed by United States v. Windsor, 570 U.S. 744 (2013). There, we explained that a lower court order that presents real-world consequences for the Government and its adversary suffices to support Article III jurisdiction—even if “the Executive may welcome” an adverse order that “is accompanied by the constitutional ruling it wants.” Id., at 758. Here, petitioner and the Government disagree about whether petitioner must comply with the civil investigative demand. The lower courts sided with the Government, and the Government has not volunteered to relinquish that victory and withdraw the demand. To the contrary, while the Government agrees that the agency is unconstitutionally structured, it believes it may nevertheless enforce the demand on remand. See infra, at 30. Accordingly, our “decision will have real meaning” for the parties. INS v. Chadha, 462 U.S. 919, 939 (1983). And, as in Windsor, any prudential concerns with deciding an important legal question in this posture can be addressed by “the practice of entertaining arguments made by an amicus when the Solicitor General confesses error with respect to a judgment below,” which we have done. 570 U. S., at 760.

We therefore turn to the merits of petitioner’s constitutional challenge.


We hold that the CFPB’s leadership by a single individual removable only for inefficiency, neglect, or malfeasance violates the separation of powers.


Article II provides that “[t]he executive Power shall be vested in a President,” who must “take Care that the Laws be faithfully executed.” Art. II, §1, cl. 1; id., §3. The entire “executive Power” belongs to the President alone. But because it would be “impossib[le]” for “one man” to “perform all the great business of the State,” the Constitution assumes that lesser executive officers will “assist the supreme Magistrate in discharging the duties of his trust.” 30 Writings of George Washington 334 (J. Fitzpatrick ed. 1939).

These lesser officers must remain accountable to the President, whose authority they wield. As Madison explained, “[I]f any power whatsoever is in its nature Executive, it is the power of appointing, overseeing, and controlling those who execute the laws.” 1 Annals of Cong. 463 (1789). That power, in turn, generally includes the ability to remove executive officials, for it is “only the authority that can remove” such officials that they “must fear and, in the performance of [their] functions, obey.” Bowsher, 478 U. S., at 726 (internal quotation marks omitted).

The President’s removal power has long been confirmed by history and precedent. It “was discussed extensively in Congress when the first executive departments were created” in 1789. Free Enterprise Fund, 561 U. S., at 492. “The view that ‘prevailed, as most consonant to the text of the Constitution’ and ‘to the requisite responsibility and harmony in the Executive Department,’ was that the executive power included a power to oversee executive officers through removal.” Ibid. (quoting Letter from James Madison to Thomas Jefferson (June 30, 1789), 16 Documentary History of the First Federal Congress 893 (2004)). The First Congress’s recognition of the President’s removal power in 1789 “provides contemporaneous and weighty evidence of the Constitution’s meaning,” Bowsher, 478 U. S., at 723 (internal quotation marks omitted), and has long been the “settled and well understood construction of the Constitution,” Ex parte Hennen, 13 Pet. 230, 259 (1839).

The Court recognized the President’s prerogative to remove executive officials in Myers v. United States, 272 U.S. 52. Chief Justice Taft, writing for the Court, conducted an exhaustive examination of the First Congress’s determination in 1789, the views of the Framers and their contemporaries, historical practice, and our precedents up until that point. He concluded that Article II “grants to the President” the “general administrative control of those executing the laws, including the power of appointment and removal of executive officers.” Id., at 163–164 (emphasis added). Just as the President’s “selection of administrative officers is essential to the execution of the laws by him, so must be his power of removing those for whom he cannot continue to be responsible.” Id., at 117. “[T]o hold otherwise,” the Court reasoned, “would make it impossible for the President . . . to take care that the laws be faithfully executed.” Id., at 164.

We recently reiterated the President’s general removal power in Free Enterprise Fund. “Since 1789,” we recapped, “the Constitution has been understood to empower the President to keep these officers accountable—by removing them from office, if necessary.” 561 U. S., at 483. Although we had previously sustained congressional limits on that power in certain circumstances, we declined to extend those limits to “a new situation not yet encountered by the Court”—an official insulated by two layers of for-cause removal protection. Id., at 483, 514. In the face of that novel impediment to the President’s oversight of the Executive Branch, we adhered to the general rule that the President possesses “the authority to remove those who assist him in carrying out his duties.” Id., at 513–514.

Free Enterprise Fund left in place two exceptions to the President’s unrestricted removal power. First, in Humphrey’s Executor, decided less than a decade after Myers, the Court upheld a statute that protected the Commissioners of the FTC from removal except for “inefficiency, neglect of duty, or malfeasance in office.” 295 U. S., at 620 (quoting 15 U. S. C. §41). In reaching that conclusion, the Court stressed that Congress’s ability to impose such removal restrictions “will depend upon the character of the office.” 295 U. S., at 631.

Because the Court limited its holding “to officers of the kind here under consideration,” id., at 632, the contours of the Humphrey’s Executor exception depend upon the characteristics of the agency before the Court. Rightly or wrongly, the Court viewed the FTC (as it existed in 1935) as exercising “no part of the executive power.” Id., at 628. Instead, it was “an administrative body” that performed “specified duties as a legislative or as a judicial aid.” Ibid. It acted “as a legislative agency” in “making investigations and reports” to Congress and “as an agency of the judiciary” in making recommendations to courts as a master in chancery. Ibid. “To the extent that [the FTC] exercise[d] any executive function[,] as distinguished from executive power in the constitutional sense,” it did so only in the discharge of its “quasi-legislative or quasi-judicial powers.” Ibid. (emphasis added).[2]

The Court identified several organizational features that helped explain its characterization of the FTC as non-executive. Composed of five members—no more than three from the same political party—the Board was designed to be “non-partisan” and to “act with entire impartiality.” Id., at 624; see id., at 619–620. The FTC’s duties were “neither political nor executive,” but instead called for “the trained judgment of a body of experts” “informed by experience.” Id., at 624 (internal quotation marks omitted). And the Commissioners’ staggered, seven-year terms enabled the agency to accumulate technical expertise and avoid a “complete change” in leadership “at any one time.” Ibid.

In short, Humphrey’s Executor permitted Congress to give for-cause removal protections to a multimember body of experts, balanced along partisan lines, that performed legislative and judicial functions and was said not to exercise any executive power. Consistent with that understanding, the Court later applied “[t]he philosophy of Humphrey’s Executor” to uphold for-cause removal protections for the members of the War Claims Commission—a three-member “adjudicatory body” tasked with resolving claims for compensation arising from World War II. Wiener v. United States, 357 U.S. 349, 356 (1958).

While recognizing an exception for multimember bodies with “quasi-judicial” or “quasi-legislative” functions, Humphrey’s Executor reaffirmed the core holding of Myers that the President has “unrestrictable power . . . to remove purely executive officers.” 295 U. S., at 632. The Court acknowledged that between purely executive officers on the one hand, and officers that closely resembled the FTC Commissioners on the other, there existed “a field of doubt” that the Court left “for future consideration.” Ibid.

We have recognized a second exception for inferior officers in two cases, United States v. Perkins and Morrison v. Olson.[3] In Perkins, we upheld tenure protections for a naval cadet-engineer. 116 U. S., at 485. And, in Morrison, we upheld a provision granting good-cause tenure protection to an independent counsel appointed to investigate and prosecute particular alleged crimes by high-ranking Government officials. 487 U. S., at 662–663, 696–697. Backing away from the reliance in Humphrey’s Executor on the concepts of “quasi-legislative” and “quasi-judicial” power, we viewed the ultimate question as whether a removal restriction is of “such a nature that [it] impede[s] the President’s ability to perform his constitutional duty.” 487 U. S., at 691. Although the independent counsel was a single person and performed “law enforcement functions that typically have been undertaken by officials within the Executive Branch,” we concluded that the removal protections did not unduly interfere with the functioning of the Executive Branch because “the independent counsel [was] an inferior officer under the Appointments Clause, with limited jurisdiction and tenure and lacking policymaking or significant administrative authority.” Ibid.

These two exceptions—one for multimember expert agencies that do not wield substantial executive power, and one for inferior officers with limited duties and no policymaking or administrative authority—“represent what up to now have been the outermost constitutional limits of permissible congressional restrictions on the President’s removal power.” PHH, 881 F. 3d, at 196 (Kavanaugh, J., dissenting) (internal quotation marks omitted).


Neither Humphrey’s Executor nor Morrison resolves whether the CFPB Director’s insulation from removal is constitutional. Start with Humphrey’s Executor. Unlike the New Deal-era FTC upheld there, the CFPB is led by a single Director who cannot be described as a “body of experts” and cannot be considered “non-partisan” in the same sense as a group of officials drawn from both sides of the aisle. 295 U. S., at 624. Moreover, while the staggered terms of the FTC Commissioners prevented complete turnovers in agency leadership and guaranteed that there would always be some Commissioners who had accrued significant expertise, the CFPB’s single-Director structure and five-year term guarantee abrupt shifts in agency leadership and with it the loss of accumulated expertise.

In addition, the CFPB Director is hardly a mere legislative or judicial aid. Instead of making reports and recommendations to Congress, as the 1935 FTC did, the Director possesses the authority to promulgate binding rules fleshing out 19 federal statutes, including a broad prohibition on unfair and deceptive practices in a major segment of the U. S. economy. And instead of submitting recommended dispositions to an Article III court, the Director may unilaterally issue final decisions awarding legal and equitable relief in administrative adjudications. Finally, the Director’s enforcement authority includes the power to seek daunting monetary penalties against private parties on behalf of the United States in federal court—a quintessentially executive power not considered in Humphrey’s Executor.[4]

The logic of Morrison also does not apply. Everyone agrees the CFPB Director is not an inferior officer, and her duties are far from limited. Unlike the independent counsel, who lacked policymaking or administrative authority, the Director has the sole responsibility to administer 19 separate consumer-protection statutes that cover everything from credit cards and car payments to mortgages and student loans. It is true that the independent counsel in Morrison was empowered to initiate criminal investigations and prosecutions, and in that respect wielded core executive power. But that power, while significant, was trained inward to high-ranking Governmental actors identified by others, and was confined to a specified matter in which the Department of Justice had a potential conflict of interest. By contrast, the CFPB Director has the authority to bring the coercive power of the state to bear on millions of private citizens and businesses, imposing even billion-dollar penalties through administrative adjudications and civil actions.

In light of these differences, the constitutionality of the CFPB Director’s insulation from removal cannot be settled by Humphrey’s Executor or Morrison alone.


The question instead is whether to extend those precedents to the “new situation” before us, namely an independent agency led by a single Director and vested with significant executive power. Free Enterprise Fund, 561 U. S., at 483. We decline to do so. Such an agency has no basis in history and no place in our constitutional structure.


“Perhaps the most telling indication of [a] severe constitutional problem” with an executive entity “is [a] lack of historical precedent” to support it. Id., at 505 (internal quotation marks omitted). An agency with a structure like that of the CFPB is almost wholly unprecedented.

After years of litigating the agency’s constitutionality, the Courts of Appeals, parties, and amici have identified “only a handful of isolated” incidents in which Congress has provided good-cause tenure to principal officers who wield power alone rather than as members of a board or commission. Ibid. “[T]hese few scattered examples”—four to be exact—shed little light. NLRB v. Noel Canning, 573 U.S. 513, 538 (2014).

First, the CFPB’s defenders point to the Comptroller of the Currency, who enjoyed removal protection for one year during the Civil War. That example has rightly been dismissed as an aberration. It was “adopted without discussion” during the heat of the Civil War and abandoned before it could be “tested by executive or judicial inquiry.” Myers, 272 U. S., at 165. (At the time, the Comptroller may also have been an inferior officer, given that he labored “under the general direction of the Secretary of the Treasury.” Ch. 58, 12Stat. 665.)[5]

Second, the supporters of the CFPB point to the Office of the Special Counsel (OSC), which has been headed by a single officer since 1978.[6] But this first enduring single-leader office, created nearly 200 years after the Constitution was ratified, drew a contemporaneous constitutional objection from the Office of Legal Counsel under President Carter and a subsequent veto on constitutional grounds by President Reagan. See Memorandum Opinion for the General Counsel, Civil Service Commission, 2 Op. OLC 120, 122 (1978); Public Papers of the Presidents, Ronald Reagan, Vol. II, Oct. 26, 1988, pp. 1391–1392 (1991).[7] In any event, the OSC exercises only limited jurisdiction to enforce certain rules governing Federal Government employers and employees. See 5 U. S. C. §1212. It does not bind private parties at all or wield regulatory authority comparable to the CFPB.

Third, the CFPB’s defenders note that the Social Security Administration (SSA) has been run by a single Administrator since 1994. That example, too, is comparatively recent and controversial. President Clinton questioned the constitutionality of the SSA’s new single-Director structure upon signing it into law. See Public Papers of the Presidents, William J. Clinton, Vol. II, Aug. 15, 1994, pp. 1471–1472 (1995) (inviting a “corrective amendment” from Congress). In addition, unlike the CFPB, the SSA lacks the authority to bring enforcement actions against private parties. Its role is largely limited to adjudicating claims for Social Security benefits.

The only remaining example is the Federal Housing Finance Agency (FHFA), created in 2008 to assume responsibility for Fannie Mae and Freddie Mac. That agency is essentially a companion of the CFPB, established in response to the same financial crisis. See Housing and Economic Recovery Act of 2008, 122Stat. 2654. It regulates primarily Government-sponsored enterprises, not purely private actors. And its single-Director structure is a source of ongoing controversy. Indeed, it was recently held unconstitutional by the Fifth Circuit, sitting en banc. See Collins v. Mnuchin, 938 F.3d 553, 587–588 (2019).

With the exception of the one-year blip for the Comptroller of the Currency, these isolated examples are modern and contested. And they do not involve regulatory or enforcement authority remotely comparable to that exercised by the CFPB. The CFPB’s single-Director structure is an innovation with no foothold in history or tradition.[8]


In addition to being a historical anomaly, the CFPB’s single-Director configuration is incompatible with our constitutional structure. Aside from the sole exception of the Presidency, that structure scrupulously avoids concentrating power in the hands of any single individual.

“The Framers recognized that, in the long term, structural protections against abuse of power were critical to preserving liberty.” Bowsher, 478 U. S., at 730. Their solution to governmental power and its perils was simple: divide it. To prevent the “gradual concentration” of power in the same hands, they enabled “[a]mbition . . . to counteract ambition” at every turn. The Federalist No. 51, p. 349 (J. Cooke ed. 1961) (J. Madison). At the highest level, they “split the atom of sovereignty” itself into one Federal Government and the States. Gamble v. United States, 587 U. S. ___, ___ (2019) (slip op., at 9) (internal quotation marks omitted). They then divided the “powers of the new Federal Government into three defined categories, Legislative, Executive, and Judicial.” Chadha, 462 U. S., at 951.

They did not stop there. Most prominently, the Framers bifurcated the federal legislative power into two Chambers: the House of Representatives and the Senate, each composed of multiple Members and Senators. Art. I, §§2, 3.

The Executive Branch is a stark departure from all this division. The Framers viewed the legislative power as a special threat to individual liberty, so they divided that power to ensure that “differences of opinion” and the “jarrings of parties” would “promote deliberation and circumspection” and “check excesses in the majority.” See The Federalist No. 70, at 475 (A. Hamilton); see also id., No. 51, at 350. By contrast, the Framers thought it necessary to secure the authority of the Executive so that he could carry out his unique responsibilities. See id., No. 70, at 475–478. As Madison put it, while “the weight of the legislative authority requires that it should be . . . divided, the weakness of the executive may require, on the other hand, that it should be fortified.” Id., No. 51, at 350.

The Framers deemed an energetic executive essential to “the protection of the community against foreign attacks,” “the steady administration of the laws,” “the protection of property,” and “the security of liberty.” Id., No. 70, at 471. Accordingly, they chose not to bog the Executive down with the “habitual feebleness and dilatoriness” that comes with a “diversity of views and opinions.” Id., at 476. Instead, they gave the Executive the “[d]ecision, activity, secrecy, and dispatch” that “characterise the proceedings of one man.” Id., at 472.

To justify and check that authority—unique in our constitutional structure—the Framers made the President the most democratic and politically accountable official in Government. Only the President (along with the Vice President) is elected by the entire Nation. And the President’s political accountability is enhanced by the solitary nature of the Executive Branch, which provides “a single object for the jealousy and watchfulness of the people.” Id., at 479. The President “cannot delegate ultimate responsibility or the active obligation to supervise that goes with it,” because Article II “makes a single President responsible for the actions of the Executive Branch.” Free Enterprise Fund, 561 U. S., at 496–497 (quoting Clinton v. Jones, 520 U.S. 681, 712–713 (1997) (Breyer, J., concurring in judgment)).

The resulting constitutional strategy is straightforward: divide power everywhere except for the Presidency, and render the President directly accountable to the people through regular elections. In that scheme, individual executive officials will still wield significant authority, but that authority remains subject to the ongoing supervision and control of the elected President. Through the President’s oversight, “the chain of dependence [is] preserved,” so that “the lowest officers, the middle grade, and the highest” all “depend, as they ought, on the President, and the President on the community.” 1 Annals of Cong. 499 (J. Madison).

The CFPB’s single-Director structure contravenes this carefully calibrated system by vesting significant governmental power in the hands of a single individual accountable to no one. The Director is neither elected by the people nor meaningfully controlled (through the threat of removal) by someone who is. The Director does not even depend on Congress for annual appropriations. See The Federalist No. 58, at 394 (J. Madison) (describing the “power over the purse” as the “most compleat and effectual weapon” in representing the interests of the people). Yet the Director may unilaterally, without meaningful supervision, issue final regulations, oversee adjudications, set enforcement priorities, initiate prosecutions, and determine what penalties to impose on private parties. With no colleagues to persuade, and no boss or electorate looking over her shoulder, the Director may dictate and enforce policy for a vital segment of the economy affecting millions of Americans.

The CFPB Director’s insulation from removal by an accountable President is enough to render the agency’s structure unconstitutional. But several other features of the CFPB combine to make the Director’s removal protection even more problematic. In addition to lacking the most direct method of presidential control—removal at will—the agency’s unique structure also forecloses certain indirect methods of Presidential control.

Because the CFPB is headed by a single Director with a five-year term, some Presidents may not have any opportunity to shape its leadership and thereby influence its activities. A President elected in 2020 would likely not appoint a CFPB Director until 2023, and a President elected in 2028 may never appoint one. That means an unlucky President might get elected on a consumer-protection platform and enter office only to find herself saddled with a holdover Director from a competing political party who is dead set against that agenda. To make matters worse, the agency’s single-Director structure means the President will not have the opportunity to appoint any other leaders—such as a chair or fellow members of a Commission or Board—who can serve as a check on the Director’s authority and help bring the agency in line with the President’s preferred policies.

The CFPB’s receipt of funds outside the appropriations process further aggravates the agency’s threat to Presidential control. The President normally has the opportunity to recommend or veto spending bills that affect the operation of administrative agencies. See Art. I, §7, cl. 2; Art. II, §3. And, for the past century, the President has annually submitted a proposed budget to Congress for approval. See Budget and Accounting Act, 1921, ch. 18, §201, 42Stat. 20. Presidents frequently use these budgetary tools “to influence the policies of independent agencies.” PHH, 881 F. 3d, at 147 (Henderson, J., dissenting) (citing Pasachoff, The President’s Budget as a Source of Agency Policy Control, 125 Yale L. J. 2182, 2191, 2203–2204 (2016)). But no similar opportunity exists for the President to influence the CFPB Director. Instead, the Director receives over $500 million per year to fund the agency’s chosen priorities. And the Director receives that money from the Federal Reserve, which is itself funded outside of the annual appropriations process. This financial freedom makes it even more likely that the agency will “slip from the Executive’s control, and thus from that of the people.” Free Enterprise Fund, 561 U. S., at 499.[9]


Amicus raises three principal arguments in the agency’s defense. At the outset, amicus questions the textual basis for the removal power and highlights statements from Madison, Hamilton, and Chief Justice Marshall expressing “heterodox” views on the subject. Brief for Court-Appointed Amicus Curiae 4–5, 28–29. But those concerns are misplaced. It is true that “there is no ‘removal clause’ in the Constitution,” id., at 1, but neither is there a “separation of powers clause” or a “federalism clause.” These foundational doctrines are instead evident from the Constitution’s vesting of certain powers in certain bodies. As we have explained many times before, the President’s removal power stems from Article II’s vesting of the “executive Power” in the President. Free Enterprise Fund, 561 U. S., at 483 (quoting Art. II, §1, cl. 1). As for the opinions of Madison, Hamilton, and Chief Justice Marshall, we have already considered the statements cited by amicus and discounted them in light of their context (Madison), the fact they reflect initial impressions later abandoned by the speaker (Hamilton), or their subsequent rejection as ill-considered dicta (Chief Justice Marshall). See Free Enterprise Fund, 561 U. S., at 500, n. 6 (Madison); Myers, 272 U. S., at 136–139, 142–144 (Hamilton and Chief Justice Marshall).[10]

Next, amicus offers a grand theory of our removal precedents that, if accepted, could leave room for an agency like the CFPB—and many other innovative intrusions on Article II. According to amicus, Humphrey’s Executor and Morrison establish a general rule that Congress may impose “modest” restrictions on the President’s removal power, with only two limited exceptions. Brief for Court-Appointed Amicus Curiae 33–37. Congress may not reserve a role for itself in individual removal decisions (as it attempted to do in Myers and Bowsher). And it may not eliminate the President’s removal power altogether (as it effectively did in Free Enterprise Fund). Outside those two situations, amicus argues, Congress is generally free to constrain the President’s removal power. See also post, at 16–22 (Kagan, J., concurring in judgment with respect to severability and dissenting in part) (hereinafter dissent) (expressing similar view).

But text, first principles, the First Congress’s decision in 1789, Myers, and Free Enterprise Fund all establish that the President’s removal power is the rule, not the exception. While we do not revisit Humphrey’s Executor or any other precedent today, we decline to elevate it into a freestanding invitation for Congress to impose additional restrictions on the President’s removal authority.[11]

Finally, amicus contends that if we identify a constitutional problem with the CFPB’s structure, we should avoid it by broadly construing the statutory grounds for removing the CFPB Director from office. See Brief for Court-Appointed Amicus Curiae 50–53; Tr. of Oral Arg. 57–62. The Dodd-Frank Act provides that the Director may be removed for “inefficiency, neglect of duty, or malfeasance in office.” 12 U. S. C. §5491(c)(3). In amicus’ view, that language could be interpreted to reserve substantial discretion to the President. Brief for Court-Appointed Amicus Curiae 51.

We are not persuaded. For one, Humphrey’s Executor implicitly rejected an interpretation that would leave the President free to remove an officer based on disagreements about agency policy. See 295 U. S., at 619, 625–626. In addition, while both amicus and the House of Representatives invite us to adopt whatever construction would cure the constitutional problem, they have not advanced any workable standard derived from the statutory language. Amicus suggests that the proper standard might permit removals based on general policy disagreements, but not specific ones; the House suggests that the permissible bases for removal might vary depending on the context and the Presidential power involved. See Tr. of Oral Arg. 58–60, 76–77. They do not attempt to root either of those standards in the statutory text. Further, although nearly identical language governs the removal of some two-dozen multimember independent agencies, amicus suggests that the standard should vary from agency to agency, morphing as necessary to avoid constitutional doubt. Tr. of Oral Arg. 55–56. We decline to embrace such an uncertain and elastic approach to the text.

Amicus and the House also fail to engage with the Dodd-Frank Act as a whole, which makes plain that the CFPB is an “independent bureau.” 12 U. S. C. §5491(a); see also 44 U. S. C. §3502(5) (listing the CFPB as an “independent regulatory agency”). Neither amicus nor the House explains how the CFPB would be “independent” if its head were required to implement the President’s policies upon pain of removal. See Black’s Law Dictionary 838 (9th ed. 2009) (defining “independent” as “[n]ot subject to the control or influence of another”). The Constitution might of course compel the agency to be dependent on the President notwithstanding Congress’s contrary intent, but that result cannot fairly be inferred from the statute Congress enacted.

Constitutional avoidance is not a license to rewrite Congress’s work to say whatever the Constitution needs it to say in a given situation. Without a proffered interpretation that is rooted in the statutory text and structure, and would avoid the constitutional violation we have identified, we take Congress at its word that it meant to impose a meaningful restriction on the President’s removal authority.

The dissent, for its part, largely reprises points that the Court has already considered and rejected: It notes the lack of an express removal provision, invokes Congress’s general power to create and define executive offices, highlights isolated statements from individual Framers, downplays the decision of 1789, minimizes Myers, brainstorms methods of Presidential control short of removal, touts the need for creative congressional responses to technological and economic change, and celebrates a pragmatic, flexible approach to American governance. See post, at 1–25, 32–33, 38.

If these arguments sound familiar, it’s because they are. They were raised by the dissent in Free Enterprise Fund. Compare post, at 1–25, 32–33, 38, with Free Enterprise Fund, 561 U. S., at 515–524, 530 (Breyer, J., dissenting). The answers to these repeated concerns (beyond those we have already covered) are the same today as they were ten years ago. Today, as then, Congress’s “plenary control over the salary, duties, and even existence of executive offices” makes “Presidential oversight” more critical—not less—as the “[o]nly” tool to “counter [Congress’s] influence.” Id., at 500 (opinion of the Court). Today, as then, the various “bureaucratic minutiae” a President might use to corral agency personnel is no substitute for at will removal. Ibid. And today, as always, the urge to meet new technological and societal problems with novel governmental structures must be tempered by constitutional restraints that are not known—and were not chosen—for their efficiency or flexibility. Id., at 499.

As we explained in Free Enterprise Fund, “One can have a government that functions without being ruled by functionaries, and a government that benefits from expertise without being ruled by experts.” Ibid. While “[n]o one doubts Congress’s power to create a vast and varied federal bureaucracy,” the expansion of that bureaucracy into new territories the Framers could scarcely have imagined only sharpens our duty to ensure that the Executive Branch is overseen by a President accountable to the people. Ibid.


Having concluded that the CFPB’s leadership by a single independent Director violates the separation of powers, we now turn to the appropriate remedy. We directed the parties to brief and argue whether the Director’s removal protection was severable from the other provisions of the Dodd-Frank Act that establish the CFPB. If so, then the CFPB may continue to exist and operate notwithstanding Congress’s unconstitutional attempt to insulate the agency’s Director from removal by the President. There is a live controversy between the parties on that question, and resolving it is a necessary step in determining petitioner’s entitlement to its requested relief.

As the defendant in this action, petitioner seeks a straightforward remedy. It asks us to deny the Government’s petition to enforce the civil investigative demand and dismiss the case. The Government counters that the demand, though initially issued by a Director unconstitutionally insulated from removal, can still be enforced on remand because it has since been ratified by an Acting Director accountable to the President. The parties dispute whether this alleged ratification in fact occurred and whether, if so, it is legally sufficient to cure the constitutional defect in the original demand. That debate turns on case-specific factual and legal questions not addressed below and not briefed here. A remand for the lower Courts to consider those questions in the first instance is therefore the appropriate course—unless such a remand would be futile.

In petitioner’s view, it would be. Before the Court of Appeals, petitioner contended that, regardless of any ratification, the demand is unenforceable because the statutory provision insulating the CFPB Director from removal cannot be severed from the other statutory provisions that define the CFPB’s authority. See Brief for Appellant in No. 17–56324 (CA9), pp. 27–28, 30–32. If petitioner is correct, and the offending removal provision means the entire agency is unconstitutional and powerless to act, then a remand would be pointless. With no agency left with statutory authority to maintain this suit or otherwise enforce the demand, the appropriate disposition would be to reverse with instructions to deny the Government’s petition to enforce the agency’s demand for documents and dismiss the case, as petitioner requests.

Accordingly, there is a live controversy over the question of severability. And that controversy is essential to our ability to provide petitioner the relief it seeks: If the removal restriction is not severable, then we must grant the relief requested, promptly rejecting the demand outright. If, on the other hand, the removal restriction is severable, we must instead remand for the Government to press its ratification arguments in further proceedings. Unlike the lingering ratification issue, severability presents a pure question of law that has been fully briefed and argued by the parties. We therefore proceed to address it.[12]

It has long been settled that “one section of a statute may be repugnant to the Constitution without rendering the whole act void.” Loeb v. Columbia Township Trustees, 179 U.S. 472, 490 (1900) (quoting Treasurer of Fayette Cty. v. People’s & Drovers’ Bank, 47 Ohio St. 503, 523, 25 N.E. 697, 702 (1890)). Because a “statute bad in part is not necessarily void in its entirety,” “[p]rovisions within the legislative power may stand if separable from the bad.” Dorchy v. Kansas, 264 U.S. 286, 289–290 (1924).

“Generally speaking, when confronting a constitutional flaw in a statute, we try to limit the solution to the problem, severing any problematic portions while leaving the remainder intact.” Free Enterprise Fund, 561 U. S., at 508 (internal quotation marks omitted). Even in the absence of a severability clause, the “traditional” rule is that “the unconstitutional provision must be severed unless the statute created in its absence is legislation that Congress would not have enacted.” Alaska Airlines, Inc. v. Brock, 480 U.S. 678, 685 (1987). When Congress has expressly provided a severability clause, our task is simplified. We will presume “that Congress did not intend the validity of the statute in question to depend on the validity of the constitutionally offensive provision . . . unless there is strong evidence that Congress intended otherwise.” Id., at 686.

The only constitutional defect we have identified in the CFPB’s structure is the Director’s insulation from removal. If the Director were removable at will by the President, the constitutional violation would disappear. We must therefore decide whether the removal provision can be severed from the other statutory provisions relating to the CFPB’s powers and responsibilities.

In Free Enterprise Fund, we found a set of unconstitutional removal provisions severable even in the absence of an express severability clause because the surviving provisions were capable of “functioning independently” and “nothing in the statute’s text or historical context [made] it evident that Congress, faced with the limitations imposed by the Constitution, would have preferred no Board at all to a Board whose members are removable at will.” 561 U. S., at 509 (internal quotation marks omitted).

So too here. The provisions of the Dodd-Frank Act bearing on the CFPB’s structure and duties remain fully operative without the offending tenure restriction. Those provisions are capable of functioning independently, and there is nothing in the text or history of the Dodd-Frank Act that demonstrates Congress would have preferred no CFPB to a CFPB supervised by the President. Quite the opposite. Unlike the Sarbanes-Oxley Act at issue in Free Enterprise Fund, the Dodd-Frank Act contains an express severability clause. There is no need to wonder what Congress would have wanted if “any provision of this Act” is “held to be unconstitutional” because it has told us: “the remainder of this Act” should “not be affected.” 12 U. S. C. §5302.

Petitioner urges us to disregard this plain language for three reasons. None is persuasive. First, petitioner dismisses the clause as non-probative “boilerplate” because it applies “to the entire, 848-page Dodd-Frank Act” and “appears almost 600 pages before the removal provision at issue.” Brief for Petitioner 45. In petitioner’s view, that means we cannot be certain that Congress really meant to apply the clause to each of the Act’s provisions. But boilerplate is boilerplate for a reason—because it offers tried-and-true language to ensure a precise and predictable result. That is the case here. The language unmistakably references “any provision of this Act.” 12 U. S. C. §5302 (emphasis added). And it appears in a logical and prominent place, immediately following the Act’s title and definitions sections, reinforcing the conclusion that it applies to the entirety of the Act. Congress was not required to laboriously insert duplicative severability clauses, provision by provision, to accomplish its stated objective.

Second, petitioner points to an additional severability clause in the Act that applies only to one of the Act’s subtitles. See 15 U. S. C. §8232. In petitioner’s view, that clause would be superfluous if Congress meant the general severability clause to apply across the Act. But “our preference for avoiding surplusage constructions is not absolute.” Lamie v. United States Trustee, 540 U.S. 526, 536 (2004). In this instance, the redundant language appears to reflect the fact that the subtitle to which it refers originated as a standalone bill that was later incorporated into Dodd-Frank. Compare 15 U. S. C. §8232 with H. R. 2571, 111th Cong., 1st Sess., §302 (2009). And petitioner does not offer any construction that would give effect to both provisions, making the redundancy both inescapable and unilluminating. See Microsoft Corp. v. i4i L. P., 564 U.S. 91, 106 (2011) (“The canon against superfluity assists only where a competing interpretation gives effect to every clause and word of a statute.” (internal quotation marks omitted)).

Finally, petitioner argues more broadly that Congress would not have wanted to give the President unbridled control over the CFPB’s vast authority. Petitioner highlights the references to the CFPB’s independence in the statutory text and legislative history, as well as in Professor Warren’s and the Obama administration’s original proposals. See Brief for Petitioner 43–44 (collecting examples). And petitioner submits that Congress might not have exempted the CFPB from congressional oversight via the appropriations process if it had known that the CFPB would come under executive control.

These observations certainly confirm that Congress preferred an independent CFPB to a dependent one; but they shed little light on the critical question whether Congress would have preferred a dependent CFPB to no agency at all. That is the only question we have the authority to decide, and the answer seems clear. Petitioner assumes that, if we eliminate the CFPB, regulatory and enforcement authority over the statutes it administers would simply revert back to the handful of independent agencies previously responsible for them. See id., at 46. But, as the Solicitor General and House of Representatives explain, that shift would trigger a major regulatory disruption and would leave appreciable damage to Congress’s work in the consumer-finance arena. See Reply Brief for Respondent 21–22; Tr. of Oral Arg. 67–68. One of the agencies whose regulatory authority was transferred to the CFPB no longer exists. See 12 U. S. C. §§5412–5413 (Office of Thrift Supervision). The others do not have the staff or appropriations to absorb the CFPB’s 1,500-employee, 500-million-dollar operations. And none has the authority to administer the Dodd-Frank Act’s new prohibition on unfair and deceptive practices in the consumer-finance sector. Given these consequences, it is far from evident that Congress would have preferred no CFPB to a CFPB led by a Director removable at will by the President.

Justice Thomas would have us junk our settled severability doctrine and start afresh, even though no party has asked us to do so. See post, at 15–16, 21–24 (opinion concurring in part and dissenting in part). Among other things, he objects that it is sheer “speculation” that Congress would prefer that its consumer protection laws be enforced by a Director accountable to the President rather than not at all. Post, at 23–24. We think it clear that Congress would prefer that we use a scalpel rather than a bulldozer in curing the constitutional defect we identify today. And such an approach by this Court can come as no surprise to Congress, which was on notice of constitutional objections to single-Director agencies by multiple past Presidents from both political parties, supra, at 19–20, and enacted Dodd-Frank against the background of our established severability doctrine.

As in every severability case, there may be means of remedying the defect in the CFPB’s structure that the Court lacks the authority to provide. Our severability analysis does not foreclose Congress from pursuing alternative responses to the problem—for example, converting the CFPB into a multimember agency. The Court’s only instrument, however, is a blunt one. We have “the negative power to disregard an unconstitutional enactment,” Massachusetts v. Mellon, 262 U.S. 447, 488 (1923); see Marbury v. Madison, 1 Cranch 137, 178 (1803), but we cannot re-write Congress’s work by creating offices, terms, and the like. “[S]uch editorial freedom . . . belongs to the Legislature, not the Judiciary.” Free Enterprise Fund, 561 U. S., at 510.

Because we find the Director’s removal protection severable from the other provisions of Dodd-Frank that establish the CFPB, we remand for the Court of Appeals to consider whether the civil investigative demand was validly ratified.

*  *  *

A decade ago, we declined to extend Congress’s authority to limit the President’s removal power to a new situation, never before confronted by the Court. We do the same today. In our constitutional system, the executive power belongs to the President, and that power generally includes the ability to supervise and remove the agents who wield executive power in his stead. While we have previously upheld limits on the President’s removal authority in certain contexts, we decline to do so when it comes to principal officers who, acting alone, wield significant executive power. The Constitution requires that such officials remain dependent on the President, who in turn is accountable to the people.

The judgment of the United States Court of Appeals for the Ninth Circuit is vacated, and the case is remanded for further proceedings consistent with this opinion.

It is so ordered.

1  Director Kraninger did not indicate whether she would disregard her statutory removal protection if the President attempted to remove her without cause.
2  The Court’s conclusion that the FTC did not exercise executive power has not withstood the test of time. As we observed in Morrison v. Olson, 487 U.S. 654 (1988), “[I]t is hard to dispute that the powers of the FTC at the time of Humphrey’s Executor would at the present time be considered ‘executive,’ at least to some degree.” Id., at 690, n. 28. See also Arlington v. FCC, 569 U.S. 290, 305, n. 4 (2013) (even though the activities of administrative agencies “take ‘legislative’ and ‘judicial’ forms,” “they are exercises of—indeed, under our constitutional structure they must be exercises of—the ‘executive Power’ ” (quoting Art. II, §1, cl. 1)).
3  Article II distinguishes between two kinds of officers—principal officers (who must be appointed by the President with the advice and consent of the Senate) and inferior officers (whose appointment Congress may vest in the President, courts, or heads of Departments). §2, cl. 2. While “[o]ur cases have not set forth an exclusive criterion for distinguishing between principal and inferior officers,” we have in the past examined factors such as the nature, scope, and duration of an officer’s duties. Edmond v. United States, 520 U.S. 651, 661 (1997). More recently, we have focused on whether the officer’s work is “directed and supervised” by a principal officer. Id., at 663.
4  The dissent would have us ignore the reasoning of Humphrey’s Executor and instead apply the decision only as part of a reimagined Humphrey’s-through-Morrison framework. See post, at 18, n. 7, 19–22 (Kagan, J., concurring in judgment with respect to severability and dissenting in part) (hereinafter dissent). But we take the decision on its own terms, not through gloss added by a later Court in dicta. The dissent also criticizes us for suggesting that the 1935 FTC may have had lesser responsibilities than the present FTC. See post, at 27, n. 10. Perhaps the FTC possessed broader rulemaking, enforcement, and adjudicatory powers than the Humphrey’s Court appreciated. Perhaps not. Either way, what matters is the set of powers the Court considered as the basis for its decision, not any latent powers that the agency may have had not alluded to by the Court.
5  The dissent suggests that the Comptroller still enjoyed some degree of insulation after his removal protection was repealed because the President faced a new requirement to “communicate[ ]” his “reasons” for terminating the Comptroller to the Senate. Post, at 15 (quoting Act of June 3, 1864, ch. 106, §1, 13Stat. 100). But the President could still remove the Comptroller for any reason so long as the President was, in the dissent’s phrase, “in a firing mood.” Post, at 15.
6  The OSC should not be confused with the independent counsel in Morrison or the special counsel recently appointed to investigate allegations related to the 2016 Presidential election. Despite sharing similar titles, those individuals have no relationship to the OSC.
7  An Act similar to the one vetoed by President Reagan was eventually signed by President George H. W. Bush after extensive negotiations and compromises with Congress. See Public Papers of the Presidents, George H. W. Bush, Vol. I, Apr. 10, 1989, p. 391 (1990).
8  The dissent categorizes the CFPB as one of many “financial regulators” that have historically enjoyed some insulation from the President. See post, at 11–16. But even assuming financial institutions like the Second Bank and the Federal Reserve can claim a special historical status, the CFPB is in an entirely different league. It acts as a mini legislature, prosecutor, and court, responsible for creating substantive rules for a wide swath of industries, prosecuting violations, and levying knee-buckling penalties against private citizens. See supra, at 4–5. And, of course, it is the only agency of its kind run by a single Director.
9  Amicus and the dissent try to diminish the CFPB’s insulation from Presidential control by observing that the CFPB’s final rules can be set aside by a super majority of the Financial Stability and Oversight Council (FSOC). See Brief for Court-Appointed Amicus Curiae 40; post, at 33, n. 13, 36. But the FSOC’s veto power is statutorily reserved for extreme situations, when two-thirds of the Council concludes that a CFPB regulation would “put the safety and soundness of the United States banking system or the stability of the financial system of the United States at risk.” 12 U. S. C. §§5513(a), (c)(3). That narrow escape hatch has no impact on the CFPB’s enforcement or adjudicatory authority and has never been used in the ten years since the agency’s creation. It certainlydoes not render the CFPB’s independent, single-Director structureconstitutional.
10  The dissent likewise points to Madison’s statement in The Federalist No. 39 that the “tenure” of “ministerial offices generally will be a subject of legal regulation.” Post, at 10 (quoting The Federalist No. 39, p. 253 (J. Cooke ed. 1961)). But whatever Madison may have meant by that statement, he later led the charge in contending, on the floor of the First Congress, that “inasmuch as the power of removal is of an Executive nature . . . it is beyond the reach of the Legislative body.” 1 Annals of Cong. 464 (1789); see also id., at 462–464, 495–496. Like the dissent in Free Enterprise Fund, the dissent goes on to “attribute[ ] to Madison a belief that . . . the Comptroller[ ] could be made independent of the President. But Madison’s actual proposal, consistent with his view of the Constitution, was that the Comptroller hold office for a term of ‘years, unless sooner removed by the President’; he would thus be ‘dependent upon the President, because he can be removed by him,’ and also ‘dependent upon the Senate, because they must consent to his [reappointment] for every term of years.’ ” Free Enterprise Fund v. Public Company Accounting Oversight Bd., 561 U.S. 477, 499, 500 n. 6 (2010) (citation omitted) (quoting 1 Annals of Cong. 612). See post, at 10, n. 4. The dissent further notes that, at the time of the founding, some States placed limitations on their Governors’ removal power. See post, at 7. But the Framers hardly viewed State Governors as a reliable guide in fashioning the Federal Executive. Indeed, they expressly rejected the “executive council” structure favored by most States, fearing that subjecting the President to oversight, as the States had, would “distract and . . . enervate the whole system of administration” and inject it with “habitual feebleness and dilatoriness.” The Federalist No. 70, at 473, 476 (A. Hamilton).
11  Building on amicus’ proposal, the dissent would endorse whatever “the times demand, so long as the President retains the ability to carry out his constitutional functions.” Post, at 4. But that amorphous test provides no real limiting principle. The “clearest” (and only) “example” the dissent can muster for what may be prohibited is a for-cause removal restriction placed on the President’s “close military or diplomatic advisers.” Post, at 17. But that carveout makes no logical or constitutional sense. In the dissent’s view, for-cause removal restrictions are permissible because they guarantee the President “meaningful control” over his subordinates. Post, at 28 (internal quotation marks and alterations omitted); see also post, at 8, 20, 26, 36. If that is the theory, then what is the harm in giving the President the same “meaningful control” over his close advisers? The dissent claims to see a constitutional distinction between the President’s “own constitutional duties in foreign relations and war” and his duty to execute laws passed by Congress. Post, at 13. But the same Article that establishes the President’s foreign relations and war duties expressly entrusts him to take care that the laws be faithfully executed. And, from the perspective of the governed, it is far from clear that the President’s core and traditional powers present greater cause for concern than peripheral and modern ones. If anything, “[t]he growth of the Executive Branch, which now wields vast power and touches almost every aspect of daily life, heightens the concern that it may slip from the Executive’s control, and thus from that of the people.” Free Enterprise Fund, 561 U. S., at 499 (emphasis added).
12  Justice Thomas believes that any ratification is irrelevant. In his view, even if the issuance of the demand and initiation of this suit have been validly ratified, Director Kraninger’s activities in litigating the case—after inheriting it from an Acting Director, but before becoming removable at will herself in light of our decision—present a distinct constitutional injury requiring immediate dismissal. See post, at 17–19 (opinion concurring in part and dissenting in part). But whether and when the temporary involvement of an unconstitutionally insulated officer in an otherwise valid prosecution requires dismissal falls outside the questions presented, has not been fully briefed, and is best resolved by the lower courts in the first instance.


No. 19–7



on writ of certiorari to the united states court of appeals for the ninth circuit

[June 29, 2020]

Justice Thomas, with whom Justice Gorsuch joins, concurring in part and dissenting in part.

The Court’s decision today takes a restrained approach on the merits by limiting Humphrey’s Executor v. United States, 295 U.S. 602 (1935), rather than overruling it. At the same time, the Court takes an aggressive approach on severability by severing a provision when it is not necessary to do so. I would do the opposite.

Because the Court takes a step in the right direction by limiting Humphrey’s Executor to “multimember expert agencies that do not wield substantial executive power,” ante, at 16 (emphasis added), I join Parts I, II, and III of its opinion. I respectfully dissent from the Court’s severability analysis, however, because I do not believe that we should address severability in this case.


The decision in Humphrey’s Executor poses a direct threat to our constitutional structure and, as a result, the liberty of the American people. The Court concludes that it is not strictly necessary for us to overrule that decision. See ante, at 2, 13–17. But with today’s decision, the Court has repudiated almost every aspect of Humphrey’s Executor. In a future case, I would repudiate what is left of this erroneous precedent.


“The Constitution does not vest the Federal Government with an undifferentiated ‘governmental power.’ ” Department of Transportation v. Association of American Railroads, 575 U.S. 43, 67 (2015) (Thomas, J., concurring in judgment). It sets out three branches and vests a different form of power in each—legislative, executive, and judicial. See Art. I, §1; Art. II, §1, cl. 1; Art. III, §1.

Article II of the Constitution vests “[t]he executive Power” in the “President of the United States of America,” §1, cl. 1, and directs that he shall “take Care that the Laws be faithfully executed,” §3. Of course, the President cannot fulfill his role of executing the laws without assistance. See Myers v. United States, 272 U.S. 52, 117 (1926). He therefore must “select those who [are] to act for him under his direction in the execution of the laws.” Ibid. While these officers assist the President in carrying out his constitutionally assigned duties, “[t]he buck stops with the President.” Free Enterprise Fund v. Public Company Accounting Oversight Bd., 561 U.S. 477, 493 (2010). “Since 1789, the Constitution has been understood to empower the President to keep [his] officers accountable—by removing them from office, if necessary.” Id., at 483. The Framers “insist[ed]” upon “unity in the Federal Executive” to “ensure both vigor and accountability” to the people. Printz v. United States, 521 U.S. 898, 922 (1997); see also ante, at 22.

Despite the defined structural limitations of the Constitution and the clear vesting of executive power in the President, Congress has increasingly shifted executive power to a de facto fourth branch of Government—independent agencies. These agencies wield considerable executive power without Presidential oversight. They are led by officers who are insulated from the President by removal restrictions, “reduc[ing] the Chief Magistrate to [the role of] cajoler-in-chief.” Free Enterprise Fund, 561 U. S., at 502. But “[t]he people do not vote for the Officers of the United States. They instead look to the President to guide the assistants or deputies subject to his superintendence.” Id., at 497–498 (alterations, internal quotation marks and citation omitted). Because independent agencies wield substantial power with no accountability to either the President or the people, they “pose a significant threat to individual liberty and to the constitutional system of separation of powers and checks and balances.” PHH Corp. v. CFPB, 881 F.3d 75, 165 (CADC 2018) (Kavanaugh, J., dissenting).

Unfortunately, this Court “ha[s] not always been vigilant about protecting the structure of our Constitution,” at times endorsing a “more pragmatic, flexible approach” to our Government’s design. Perez v. Mortgage Bankers Assn., 575 U.S. 92, 115–116 (2015) (Thomas, J., concurring in judgment) (internal quotation marks omitted). Our tolerance of independent agencies in Humphrey’s Executor is an unfortunate example of the Court’s failure to apply the Constitution as written. That decision has paved the way for an ever-expanding encroachment on the power of the Executive, contrary to our constitutional design.



The lead up to Humphrey’s Executor begins with this Court’s decision in Myers, 272 U.S. 52. Myers involved a federal statute that prohibited the President from removing certain postmasters except “by and with the advice and consent of the Senate.” Id., at 107 (internal quotation marks omitted). The question presented was “whether under the Constitution the President has the exclusive power of removing executive officers of the United States whom he has appointed by and with the advice and consent of the Senate.” Id., at 106. In a 70-page opinion by Chief Justice Taft, the Court held that the Constitution did vest such power in the President.

The Court anchored its analysis in evidence from the founding era. It acknowledged that the “subject [of removal] was not discussed in the Constitutional Convention,” id., at 109–110, but it reviewed in detail the First Congress’ vigorous debate about the removal of executive officers in what is known as the Decision of 1789, id., at 111–135.[1] In the course of analyzing the Decision of 1789, the Court explained that Article II vests “the executive power of the Government . . . in one person”—the President—and that the executive power includes the authority to “select those who [are] to act for him under his direction in the execution of the laws.” Id., at 116–117. Reiterating the position of James Madison and other Members of the First Congress, the Court noted that allowing limits on the President’s removal authority would grant Congress “the means of thwarting the Executive in the exercise of his great powers and in the bearing of his great responsibility, by fastening upon him, as subordinate executive officers, men who by their inefficient service under him, by their lack of loyalty to the service, or by their different views of policy might make his taking care that the laws be faithfully executed most difficult or impossible.” Id., at 131. After “devot[ing] much space to [the] discussion and decision of the question of the Presidential power of removal in the First Congress” as well as its understanding of the executive power, id., at 136, the Court concluded that “the power to remove officers appointed by the President and the Senate vested in the President alone,” id., at 114. It repeatedly described this removal power as “unrestricted.” Id., at 115, 134, 150, 172, 176.

The Court noted that the First Congress’ understanding of the removal question was quickly “accepted as a final decision of the question by all branches of the Government.” Id., at 136. The decision was “affirmed by this Court in unmistakable terms.” Id., at 148, 152–153 (discussing Ex parte Hennen, 13 Pet. 230, 259 (1839); Parsons v. United States, 167 U.S. 324, 330 (1897)). Presidents had “uniform[ly]” adopted the First Congress’ view “whenever an issue ha[d] clearly been raised.” Myers, 272 U. S., at 169. And “Congress, in a number of acts, followed and enforced the legislative decision of 1789 for seventy-four years.” Id., at 145. While disputes with President Andrew Johnson over Reconstruction led Congress to “enact legislation to curtail the then acknowledged powers of the President,” id., at 165, the Myers Court declined to give these politically charged acts any weight, id., at 175–176.

After exhaustively analyzing the historical evidence, the Court had “no hesitation in holding that [the First Congress’] conclusion [was] correct.” Id., at 176. Accordingly, the Court held that “the provision of the law [at issue], by which the unrestricted power of removal of first class postmasters is denied to the President, [was] in violation of the Constitution, and invalid.” Ibid.


Nine years after Myers, the Court decided Humphrey’s Executor. That case arose from the attempted removal of Commissioner William Humphrey from the Federal Trade Commission (FTC). In 1931, President Herbert Hoover appointed Humphrey to serve a 7-year term as one of the FTC’s five Commissioners. By all accounts, Humphrey proved to be a controversial figure. See Crane, Debunking Humphrey’s Executor, 83 Geo. Wash. L. Rev. 1836, 1841 (2015); Winerman, The FTC at Ninety: History Through Headlines, 72 Antitrust L. J. 871, 878–879 (2005); Yoo, Calabresi, & Nee, The Unitary Executive During the Third Half-Century, 1889–1945, 80 Notre Dame L. Rev. 1, 64 (2004). He reportedly “vowed not to approve any Commission action that did not have as its goal to help business help itself,” “threaten[ed] criminal prosecution against other commissioners who publicly dissented,” and “called his fellow commissioners men drunk with their own greatness” when they voted to initiate an investigation. Crane, supra, at 1841 (internal quotation marks omitted).

Less than two years into Humphrey’s term, newly inaugurated President Franklin D. Roosevelt wrote Humphrey a letter, asking for his resignation. The President explained that, in his view, “the aims and purposes of the Administration with respect to the work of the Commission [could] be carried out most effectively with personnel of [his] own selection.” Humphrey’s Executor, 295 U. S., at 618 (internal quotation marks omitted). A little over a month after his first letter, President Roosevelt wrote Humphrey again to ask for his resignation. The letter stated: “You will, I know, realize that I do not feel that your mind and my mind go along together on either the policies or the administering of the [FTC], and, frankly, I think it is best for the people of this country that I should have a full confidence.” Id., at 619 (internal quotation marks omitted). Humphrey declined to resign. In October 1933, President Roosevelt informed Humphrey that he was removed from his position. Humphrey did not comply, continuing “to insist that he was still a member of the commission, entitled to perform its duties and receive the compensation provided by law.” Ibid.

Four months later, Humphrey died. The executor of his estate brought suit in the Court of Claims, seeking to recover Humphrey’s salary from the date of his removal until the date of his death. The Court of Claims certified two questions to this Court: (1) whether §1 of the Federal Trade Commission Act of 1914, ch. 311, 38Stat. 717, prohibited the President from removing FTC Commissioners except for “inefficiency, neglect of duty, or malfeasance in office,” and (2) if so, whether that restriction was constitutional. 295 U. S., at 619 (internal quotation marks omitted).

The Court answered both of these questions in favor of Humphrey’s estate. It first held that the FTC Act “limit[ed] the executive power of removal to the causes enumerated” therein—inefficiency, neglect of duty, or malfeasance in office. Id., at 626. In the Court’s view, this construction of the Act was clear from “the face of the statute” and “the character of the commission,” id., at 624, which the Court described as a “body of experts” that operates “independent of executive authority . . . and free to exercise its judgment without the leave or hindrance of any other official,” id., at 625–626.

Then, notwithstanding the text of Article II of the Constitution and the decision in Myers, the Court held that the Act’s restriction on the President’s authority to remove Commissioners was constitutional. The Court acknowledged that the “recently decided” Myers decision had “fully review[ed] the general subject of the power of executive removal” and “examine[d] at length the historical, legislative and judicial data bearing upon the question.” Humphrey’s Executor, 295 U. S., at 626. And it conceded that executive officers are “subject to the exclusive and illimitable power of removal by the Chief Executive.” Id., at 627; see also id., at 631 (recognizing “the President’s illimitable power of removal” over executive officers).[2] The Court, however, claimed that “[t]he office of a postmaster is so essentially unlike the office [of an FTC Commissioner] that the decision in the Myers case [could not] be accepted as controlling.” Id., at 627. In the Court’s view, unlike the postmaster in Myers, FTC commissioners did not qualify as “purely executive officers.” 295 U. S., at 632.

The Court grounded its analysis in its assertion that the FTC “occupies no place in the executive department and . . . exercises no part of the executive power vested by the Constitution in the President.” Id., at 628. Rather, in the Court’s view, by “filling in and administering the details embodied by [the FTC Act’s] general standard[,] the commission act[ed] in part quasi-legislatively and in part quasi-judicially.” Ibid. The Court stated that the FTC acted “as a legislative agency” by “making investigations and reports thereon for the information of Congress” and acted “as an agency of the judiciary” when performing its role “as a master in chancery under rules prescribed by the court.” Ibid. “Such a body,” the Court explained, “cannot in any proper sense be characterized as an arm or an eye of the executive.” Ibid.

After distinguishing “purely executive officers” from officers exercising “quasi-legislative or quasi-judicial powers,” ibid., the Court held that “[w]hether the power of the President to remove an officer shall prevail over the authority of Congress to condition the power by . . . precluding a removal except for cause, will depend upon the character of the office,” id., at 631. “[P]urely executive officers” are subject to the President’s “unrestrictable power . . . to remove.” Id., at 632. But with regard to “quasi-legislative” and “quasi-judicial” officers, the Court concluded that “no removal [could] be made . . . except for one or more of the causes named.” Ibid.


Humphrey’s Executor laid the foundation for a fundamental departure from our constitutional structure with nothing more than handwaving and obfuscating phrases such as “quasi-legislative” and “quasi-judicial.” Unlike the thorough analysis in Myers, the Court’s thinly reasoned decision is completely “devoid of textual or historical precedent for the novel principle it set forth.” Morrison v. Olson, 487 U.S. 654, 726 (1988) (Scalia, J., dissenting). The exceptional weakness of the reasoning could be a product of the circumstances under which the case was decided—in the midst of a bitter standoff between the Court and President Roosevelt[3]—or it could be just another example of this Court departing from the strictures of the Constitution for a “more pragmatic, flexible approach” to our government’s design. Perez, 575 U. S., at 116 (opinion of Thomas, J.) (internal quotation marks omitted). But whatever the motivation, Humphrey’s Executor does not comport with the Constitution.

Humphrey’s Executor relies on one key premise: the notion that there is a category of “quasi-legislative” and “quasi-judicial” power that is not exercised by Congress or the Judiciary, but that is also not part of “the executive power vested by the Constitution in the President.” Humphrey’s Executor, supra, at 628. Working from that premise, the Court distinguished the “illimitable” power of removal recognized in Myers, Humphrey’s Executor, 295 U. S., at 627–628, and upheld the FTC Act’s removal restriction, while simultaneously acknowledging that the Constitution vests the President with the entirety of the executive power, id., at 628.

The problem is that the Court’s premise was entirely wrong. The Constitution does not permit the creation of officers exercising “quasi-legislative” and “quasi-judicial powers” in “quasi-legislative” and “quasi-judicial agencies.” Id., at 628–629. No such powers or agencies exist. Congress lacks the authority to delegate its legislative power, Whitman v. American Trucking Assns., Inc., 531 U.S. 457, 472 (2001), and it cannot authorize the use of judicial power by officers acting outside of the bounds of Article III, Stern v. Marshall, 564 U.S. 462, 484 (2011). Nor can Congress create agencies that straddle multiple branches of Government. The Constitution sets out three branches of Government and provides each with a different form of power—legislative, executive, and judicial. See Art. I, §1; Art. II, §1, cl. 1; Art. III, §1. Free-floating agencies simply do not comport with this constitutional structure. “[A]gencies have been called quasi-legislative, quasi-executive or quasi-judicial, as the occasion required, in order to validate their functions within the separation-of-powers scheme of the Constitution.” FTC v. Ruberoid Co., 343 U.S. 470, 487 (1952) (Jackson, J., dissenting). But “[t]he mere retreat to the qualifying ‘quasi’ is implicit with confession that all recognized classifications have broken down, and ‘quasi’ is a smooth cover which we draw over our confusion as we might use a counterpane to conceal a disordered bed.” Id., at 487–488.

That is exactly what happened in Humphrey’s Executor. The Court upheld the FTC Act’s removal restriction by using the “quasi” label to support its claim that the FTC “exercise[d] no part of the executive power vested by the Constitution in the President.” Humphrey’s Executor, supra, at 628. But “it is hard to dispute that the powers of the FTC at the time of Humphrey’s Executor would at the present time be considered ‘executive,’ at least to some degree.” Morrison, supra, at 690, n. 28; see ante, at 14, n. 2; see post, at 18, n. 7 (Kagan, J., concurring in judgment with respect to severability and dissenting in part).


Today’s decision constitutes the latest in a series of cases that have significantly undermined Humphrey’s Executor. First, in Morrison, the Court repudiated the reasoning of the decision. 487 U. S., at 689. Then, in Free Enterprise Fund, we returned to the principles set out in the “landmark case of Myers.” 561 U. S., at 492. And today, the Court rightfully limits Humphrey’s Executor to “multimember expert agencies that do not wield substantial executive power.” Ante, at 16. After these decisions, the foundation for Humphrey’s Executor is not just shaky. It is nonexistent.

This Court’s repudiation of Humphrey’s Executor began with its decision in Morrison. There, the Court upheld a statute insulating an independent counsel from removal by the Attorney General absent a showing of “good cause.” Morrison, supra, at 659–660. In doing so, the Court set aside the reasoning of Humphrey’s Executor. It recognized that Humphrey’s Executor “rel[ied] on the terms ‘quasi- legislative’ and ‘quasi-judicial’ to distinguish the officials involved in Humphrey’s Executor . . . from those in Myers.” 487 U. S., at 689. But it then immediately stated that its “present considered view is that the determination of whether the Constitution allows Congress to impose a ‘good cause’-type restriction on the President’s power to remove an official cannot be made to turn on whether or not that official is classified as ‘purely executive.’ ” Ibid. The Court also rejected Humphrey’s Executor’s conclusion that the FTC did not exercise executive power, stating that “the powers of the FTC at the time of Humphrey’s Executor would at the present time be considered ‘executive.’ ” Morrison, supra, at 690, n. 28. The lone dissenter, Justice Scalia, disagreed with much of the Court’s analysis but noted that the Court had rightfully “swept” Humphrey’s Executor “into the dustbin of repudiated constitutional principles.” 487 U. S., at 725. Thus, all nine Members of the Court in Morrison rejected the core rationale of Humphrey’s Executor.

The reasoning of the Court’s decision in Free Enterprise Fund created further tension (if not outright conflict) with Humphrey’s Executor. In Free Enterprise Fund, the Court concluded that a dual layer of for-cause removal restrictions for members of the Public Company Accounting Oversight Board violated the Constitution. In its analysis, the Court recognized that allowing officers to “execute the laws” beyond the President’s control “is contrary to Article II’s vesting of the executive power in the President.” 561 U. S., at 496 (emphasis added). The Court acknowledged that “the executive power include[s] a power to oversee executive officers through removal.” Id., at 492. And it explained that, without the power of removal, the President cannot “be held fully accountable” for the exercise of the executive power, “ ‘greatly diminish[ing] the intended and necessary responsibility of the chief magistrate himself.’ ” Id., at 514 (quoting The Federalist No. 70, p. 478 (J. Cooke ed. 1961) (A. Hamilton)). Accountability, the Court repeatedly emphasized, plays a central role in our constitutional structure. See, e.g., Free Enterprise Fund, 561 U. S., at 498 (“[E]xecutive power without the Executive’s oversight . . . 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”); id., at 513 (“The Constitution that makes the President accountable to the people for executing the laws also gives him the power to do so”). Humphrey’s Executor is at odds with every single one of these principles: It ignores Article II’s Vesting Clause, sidesteps the President’s removal power, and encourages the exercise of executive power by unaccountable officers. The reasoning of the two decisions simply cannot be reconciled.

Finally, today’s decision builds upon Morrison and Free Enterprise Fund, further eroding the foundation of Humphrey’s Executor. The Court correctly notes that “[t]he entire ‘executive Power’ belongs to the President alone.” Ante, at 11. The President therefore must have “power to remove—and thus supervise—those who wield executive power on his behalf.” Ante, at 2. As a result, the Court concludes that Humphrey’s Executor must be limited to “multimember expert agencies that do not wield substantial executive power.” Ante, at 16 (emphasis added). And, at the same time, it recognizes (as the Court did in Morrison) that “[t]he Court’s conclusion that the FTC did not exercise executive power has not withstood the test of time.” Ante, at 14, n. 2. In other words, Humphrey’s Executor does not even satisfy its own exception.

In light of these decisions, it is not clear what is left of Humphrey’s Executor’s rationale.[4] But if any remnant of that decision is still standing, it certainly is not enough to justify the numerous, unaccountable independent agencies that currently exercise vast executive power outside the bounds of our constitutional structure.

*  *  *

Continued reliance on Humphrey’s Executor to justify the existence of independent agencies creates a serious, ongoing threat to our Government’s design. Leaving these unconstitutional agencies in place does not enhance this Court’s legitimacy; it subverts political accountability and threatens individual liberty. We have a “responsibility to ‘examin[e] without fear, and revis[e] without reluctance,’ any ‘hasty and crude decisions’ rather than leaving ‘the character of [the] law impaired, and the beauty and harmony of the [American constitutional] system destroyed by the perpetuity of error.’ ” Gamble v. United States, 587 U. S. ___, ___ (2019) (Thomas, J., concurring) (slip op., at 7) (quoting 1 J. Kent, Commentaries on American Law 444 (1826); some alterations in original). We simply cannot compromise when it comes to our Government’s structure. Today, the Court does enough to resolve this case, but in the future, we should reconsider Humphrey’s Executor in toto. And I hope that we will have the will to do so.


While I think that the Court correctly resolves the merits of the constitutional question, I do not agree with its decision to sever the removal restriction in 12 U. S. C. §5491(c)(3). See ante, at 30–36; post, at 37. To resolve this case, I would simply deny the Consumer Financial Protection Bureau (CFPB) petition to enforce the civil investigative demand.


Article III of the Constitution vests “[t]he judicial Power of the United States” in the “supreme Court” and the lower federal courts established by Congress. §1. “[T]he judicial power is, fundamentally, the power to render judgments in individual cases” or controversies that are properly before the court. Murphy v. National Collegiate Athletic Assn., 584 U. S. ___, ___–___ (2018) (Thomas, J., concurring) (slip op., at 2–3); see also Plaut v. Spendthrift Farm, Inc., 514 U.S. 211, 219 (1995) (“ ‘[A] “judicial Power” is one to render dispositive judgments’ ”); Baude, The Judgment Power, 96 Geo. L. J. 1807, 1815–1816 (2008). “[T]he power exercised is that of ascertaining and declaring the law applicable to the controversy.” Massachusetts v. Mellon, 262 U.S. 447, 488 (1923). In the context of a constitutional challenge, “[i]t amounts to little more than the negative power to disregard an unconstitutional enactment.” Ibid.; see also Mitchell, The Writ-of-Erasure Fallacy, 104 Va. L. Rev. 933, 936 (2018). Thus, if a party argues that a statute and the Constitution conflict, “then courts must resolve that dispute and, . . . follow the higher law of the Constitution.” Murphy, 584 U. S., at ___ (Thomas, J., concurring) (slip op., at 3).

Consistent with this understanding, “[e]arly American courts did not have a severability doctrine.” Id., at ___ (slip op., at 2) (citing Walsh, Partial Unconstitutionality, 85 N. Y. U. L. Rev. 738, 769 (2010)). If a statute was unconstitutional, the court would just decline to enforce the statute in the case before it. 584 U. S., at ___ (Thomas, J., concurring) (slip op., at 3). That was the end of the matter. “[T]here was no ‘next step’ in which [a] cour[t]” severed portions of a statute. Walsh, supra, at 777.

Our modern severability precedents create tension with this historic practice. Instead of declining to enforce an unconstitutional statute in an individual case, this Court has stated that courts must “seve[r] and excis[e]” portions of a statute to “remedy” the constitutional problem. United States v. Booker, 543 U.S. 220, 245 (2005); Alaska Airlines, Inc. v. Brock, 480 U.S. 678, 686 (1987). The Court’s rhetoric when discussing severance implies that a court’s decision to sever a provision “formally suspend[s] or erase[s it], when [the provision] actually remains on the books as a law.” Mitchell, supra, at 1017. The Federal Judiciary does not have the power to excise, erase, alter, or otherwise strike down a statute. Murphy, supra, at ___ (Thomas, J., concurring) (slip op., at 4); Mitchell, supra, at 936. And the Court’s reference to severability as a “remedy” is inaccurate. Traditional remedies—like injunctions, declarations, or damages—“ ‘operate with respect to specific parties,’ not ‘on legal rules in the abstract.’ ” Murphy, supra, at ___ (Thomas, J., concurring) (slip op., at 3) (quoting Harrison, Severability, Remedies, and Constitutional Adjudication, 83 Geo. Wash. L. Rev. 56, 85 (2014)).

Because the power of judicial review does not allow courts to revise statutes, Mitchell, supra, at 983, the Court’s severability doctrine must be rooted in statutory interpretation. But, even viewing severability as an interpretive question, I remain skeptical of our doctrine. As I have previously explained, “the severability doctrine often requires courts to weigh in on statutory provisions that no party has standing to challenge, bringing courts dangerously close to issuing advisory opinions.” Murphy, 584 U. S., at ___ (concurring opinion) (slip op., at 5). And the application of the doctrine “does not follow basic principles of statutory interpretation.” Id., at ___ (slip op., at 4). Instead of determining the meaning of a statute’s text, severability involves “nebulous inquir[ies] into hypothetical congressional intent.” Booker, supra, at 320, n. 7 (Thomas, J., dissenting in part).


Consistent with the traditional understanding of the judicial power, I would deny CFPB’s petition to enforce the civil investigative demand that it issued to Seila. See §5562(e)(1). Seila “challenge[d] the validity of both the civil investigative demand and the ensuing enforcement action.” Reply Brief for Petitioner 5. Seila has not countersued or sought affirmative relief preventing the CFPB from acting in the future; it simply asks us to “reverse the court of appeals’ judgment.” Brief for Petitioner 35. I would do just that. As the Court recognizes, the enforcement of a civil investigative demand by an official with unconstitutional removal protection injures Seila. See ante, at 9–10. Presented with an enforcement request from an unconstitutionally insulated Director, I would simply deny the CFPB’s petition for an order of enforcement. This approach would resolve the dispute before us without addressing the issue of severability.

The Court, however, does more. In the plurality’s view,[5] because the CFPB raised a ratification argument before the Court of Appeals, we can (and should) reach the question of severability. See ante, at 30–31. But as explained more fully below, resolving this question is wholly unnecessary. Regardless of whether the CFPB’s ratification theory is valid, the Court of Appeals on remand must reach the same outcome: The CFPB’s civil investigative demand cannot be enforced against Seila.

The ratification argument presented by the CFPB is quite simple. Since its creation in 2010, the CFPB has had three Directors—first Director Richard Cordray, then Acting Director Mick Mulvaney, and now Director Kathleen Kraninger. The CFPB’s first Director, Director Cordray, issued a civil investigative demand to Seila and initiated the enforcement action. The CFPB has conceded that these actions were unconstitutional. But, in the Ninth Circuit, the CFPB argued that the investigative demand was ratified by Acting Director Mulvaney, who it claimed was not insulated by the removal provision. Brief for Appellee in No. 17–56324, pp. 13–19. In the CFPB’s view, the President could remove Acting Director Mulvaney at will because the “removal provision by its terms applies only to ‘the Director,’ not to an Acting Director,” and the Federal Vacancy Reform Act “does not limit the President’s ability to designate a different person as Acting Director.” Id., at 14. Based on this ratification theory, the CFPB asked the Ninth Circuit to affirm the District Court’s order granting the CFPB’s petition to enforce its investigative demand.

The CFPB does not ask this Court to address ratification on the merits, but it does rely on its unresolved ratification theory to assert that the Court should reach severability. In doing so, the CFPB relies on the same theory that it presented to the Ninth Circuit. Thus, the only live ratification claim is the theory that Acting Director Mulvaney ratified the civil investigative demand. See ante, at 30–31.[6]

The resolution of the CFPB’s Acting-Director ratification theory, however, has no bearing on the outcome of the dispute before us and therefore provides no basis for addressing severability. If the Acting Director did not ratify the investigative demand, then there is obviously no need to address severability. And even if he did, the Court still does not need to address severability because the alleged ratification does not cure the constitutional injury—enforcement of an investigative demand by an unconstitutionally insulated Director. Seila “challenge[d] the validity of both the civil investigative demand and the ensuing enforcement action.” Reply Brief for Petitioner 5 (emphasis added). Acting Director Mulvaney may (or may not) have properly ratified the issuance of the investigative demand and the initiation of the enforcement proceedings. But he certainly could not ratify the continuance of the enforcement action by his successor, Director Kraninger. Id., at 7. Thus, even if the CFPB’s ratification theory is valid, Seila still has an injury: It has been (and continues to be) subjected to enforcement of an investigative demand by Director Kraninger, who “remains statutorily insulated from removal.” Reply Brief for Respondent 7; see also Free Enterprise Fund, 561 U. S., at 513; ante, at 10. Thus, we should decline to enforce the civil investigative demand against Seila. See supra, at 14–15.

Ultimately, I cannot see how the resolution of the severability question affects the dispute before us. And even if severability could affect this case in some hypothetical scenario, I would not reach out to resolve the issue given my growing discomfort with our current severability precedents.


Confident that it can address the question of severability, the plurality moves on to conduct its analysis. It starts by pointing to the severability clause in the Dodd-Frank Act. See ante, at 33. That clause states: “If any provision of this Act, an amendment made by this Act, or the application of such provision or amendment to any person or circumstance is held to be unconstitutional, the remainder of this Act, the amendments made by this Act, and the application of the provisions of such to any person or circumstance shall not be affected thereby.” §5302. The plurality states that “[i]f the Director were removable at will by the President, the constitutional violation would disappear.” Ante, at 32–33. Then, relying on language in the severability clause, it concludes that the removal provision, §5491(c)(3), should be severed.

The plurality suggests that its analysis is a matter of simply enforcing the “plain language” of the severability clause. See ante, at 33. But I am not sure it is that simple. For one, the plurality does not actually analyze the statutory language.[7] Second, the analysis the plurality does provide looks nothing like traditional statutory interpretation. Generally, when we interpret a statute, we do not hold that the text sets out a “presum[ption]” that can be rebutted by looking to atextual evidence of legislative intent. Ante, at 32. A text-based interpretation does not allow a free-ranging inquiry into what “ ‘Congress, faced with the limitations imposed by the Constitution, would have preferred’ ” had it known of a constitutional issue. Ante, at 33 (quoting Free Enterprise Fund, supra, at 509). Nor does it consider whether Congress would have wanted to avoid “a major regulatory disruption.” Ante, at 35. Statutory interpretation focuses on the text.

Even treating the question as a matter of pure statutory interpretation and assuming that the plurality points to the correct language, the text of the severability clause cannot, in isolation, justify severance of the removal provision. In some instances, a constitutional injury arises as a result of two or more statutory provisions operating together. See, e.g., Free Enterprise Fund, supra, at 509 (stating that the convergence of “a number of statutory provisions” produce a constitutional violation); Booker, 543 U. S., at 316–317 (opinion of Thomas, J.) (explaining that “the concerted action of [18 U. S. C.] §3553(b)(1) and the operative Guidelines and the relevant Rule of Criminal Procedure resulted in unconstitutional judicial factfinding”); Lea, Situation Severability, 103 Va. L. Rev. 735, 778–780 (2017) (discussing statutory convergences). That is precisely the situation we have in this case. As in Free Enterprise Fund, the provision requiring “good-cause removal is only one of [the] statutory provisions that, working together, produce a constitutional violation.” 561 U. S., at 509. The constitutional violation results from, at a minimum, the combination of the removal provision, 12 U. S. C. §5491(c)(3), and the provision allowing the CFPB to seek enforcement of a civil investigative demand, §5562(e)(1). When confronted with two provisions that operate together to violate the Constitution, the text of the severability clause provides no guidance as to which provision should be severed. Thus, we must choose, based on something other than the severability clause, which provision to sever.

Without text to guide us, the severability inquiry moves away from statutory interpretation and falls back on this Court’s questionable precedents. See Murphy, 584 U. S., at ___–___ (Thomas, J., concurring) (slip op., at 4–6). An analysis of the Court’s decisions in Booker and Free Enterprise Fund illustrates the Court’s approach to determining which provision to sever when confronting an injury caused by an unconstitutional convergence of multiple statutory provisions.

In Booker, a Rule of Criminal Procedure, a subset of provisions in the Sentencing Guidelines, and a statutory provision operated together to require unconstitutional judicial factfinding. To determine which aspect of the sentencing scheme to sever, the Court sought to divine “what Congress would have intended in light of the Court’s constitutional holding.” Booker, 543 U. S., at 246 (internal quotation marks omitted). The Court “recognize[d] that sometimes severability questions . . . can arise [in the context of] a legislatively unforeseen constitutional problem.” Id., at 247. But it nonetheless felt qualified to craft a remedy that would “move sentencing in Congress’ preferred direction.” Id., at 264. Surprisingly, that “move” did not involve enforcing the constitutional aspects of Congress’ sentencing scheme. The Court stated that “we cannot assume that Congress, if faced with the statute’s invalidity in key applications, would have preferred to apply the statute in as many other instances as possible.” Id., at 248.[8] Despite the fact that there were a plethora of cases in which mandatory Sentencing Guidelines would have posed no constitutional problem, the Court decided to “sever and excise . . . the provision that requires sentencing courts to impose a sentence within the applicable Guidelines range,” along with another provision which was not even at issue in the case. Id., at 259. In essence, the Court crafted a new sentencing scheme, transforming the Sentencing Guidelines into an entirely discretionary system based on its estimation that Congress would have wanted that result.

The Court in Free Enterprise Fund declined to explicitly engage in Booker’s free-wheeling inquiry into Congress’ hypothetical preferences, but it did not replace that inquiry with a clear standard. In that case, the Court held that a “number of statutory provisions . . . , working together, produce[d] a constitutional violation” similar to the violation at issue here. Free Enterprise Fund, 561 U. S., at 509. The Court decided to sever the Board’s removal restriction. It explicitly recognized that there were multiple ways to address the constitutional injury, stating that the Court could, for example, “blue-pencil a sufficient number of the Board’s responsibilities,” or “restrict the Board’s enforcement powers.” Ibid. But it described these alternative options as involving “editorial freedom—far more extensive than [the] holding today—[that] belongs to the Legislature, not the Judiciary.” Id., at 510. The Court did not explain, however, why the option that it chose was not also “editorial freedom” that belongs to the Legislature or why the alternatives involved “more extensive” “editorial freedom” than its preferred option. Ibid. The most that the Court provided was a suggestion that fewer provisions would have to be severed under its approach. Id., at 509–510.

Today’s plurality opinion provides no further guidance. In fact, the plurality does not even recognize that it has made a choice between the provisions that cause the constitutional injury. It merely states that “[i]f the Director were removable at will by the President, the constitutional violation would disappear.” Ante, at 32–33. Fair enough. But if the Director lacked executive authority under the statute to seek enforcement of a civil investigative demand, §5562(e)(1), the constitutional violation in this case would also disappear. The plurality thus chooses which of the provisions to sever.

In short, when multiple provisions of law combine to cause a constitutional injury, the Court’s current approach allows the Court to decide which provision to sever. The text of a severability clause does not guide that choice. Nor does the practice of early American courts. See supra, at 14–15. The Court is thus left to choose based on nothing more than speculation as to what the Legislature would have preferred. And the result of its choice can have a dramatic effect on the governing statutory scheme. See Booker, supra, at 259 (converting the entirety of the Sentencing Guidelines from a mandatory to a discretionary system). This is not a simple matter of following the “plain language” of a statute. Ante, at 33. It is incumbent on us to take a close look at our precedents to make sure that we are not exceeding the scope of the judicial power.

*  *  *

Given my concerns about our modern severability doctrine and the fact that severability makes no difference to the dispute before us, I would resolve this case by simply denying the CFPB’s petition to enforce the civil investigative demand.

1  For a comprehensive review of the Decision of 1789, see Prakash, New Light on the Decision of 1789, 91 Cornell L. Rev. 1021 (2006).
2  The explicit and repeated recognition of the President’s “illimitable power” in Humphrey’s Executor highlights the dissent’s error in claiming that Humphrey’s Executor “abandoned [the] view” set out in Myers v. United States, 272 U.S. 52 (1926). Post, at 17 (Kagan, J., concurring in judgment with respect to severability and dissenting in part) (hereinafter dissent). Humphrey’s Executor did not abandon Myers; it distinguished Myers based on the flawed premise that the FTC exercised “quasi-legislative” and “quasi-judicial” power that is not part of “the executive power vested by the Constitution in the President.” Humphrey’s Executor, 295 U. S., at 628; see also infra, at 9–11.
3  A number of historical sources indicate that President Roosevelt saw Humphrey’s Executor v. United States, 295 U.S. 602 (1935), as an attack on his administration. Given the Court’s recent decision in Myers, the Roosevelt administration was reportedly “stunned” by the Court’s decision in Humphrey’s Executor, and the President was particularly annoyed that the decision “ma[de] it appear that he had been willfully violating the Constitution.” See W. Leuchtenberg, The Supreme Court Reborn 78 (1995). Justice Jackson, who was serving in the Roosevelt administration at the time, stated in an interview that “ ‘the decision that made Roosevelt madder at the Court than any other decision was that . . . little case of Humphrey’s Executor v. United States. The President thought they went out of their way to spite him personally.’ ” E. Gerhart, America’s Advocate: Robert H. Jackson 99 (1958) (quoting 1949 interview with Justice Jackson).
4  The dissent, while vigorously defending the holding of Humphrey's Executor, can muster no defense for the reasoning of the decision. The dissent does not defend the notion of “quasi” powers or “quasi” agencies, recognizing that the power exercised by the FTC was executive power. See post, at 18, n. 7. And, in 39 pages, it cannot explain how any aspect of Humphrey’s Executor (other than its holding) survived Morrison v. Olson, 487 U.S. 654 (1988), and Free Enterprise Fund v. Public Company Accounting Oversight Bd., 561 U.S. 477 (2010). Instead, the dissent simply claims that Humphrey’s Executor was “extended” and “clarified” in Morrison, post, at 19, attempting to breathe validity into Humphrey’s Executor through the Court’s Morrison decision. But the dissent’s reading of Morrison as “extend[ing] Humphrey’s domain” is baffling. Post, at 19. Morrison expressly repudiated the substantive reasoning of Humphrey’s Executor. See supra, at 11–12.
5  The dissent provides no analysis of severability, simply stating “if the agency’s removal provision is unconstitutional, it should be severed.” Post, at 37.
6  The Court-appointed amicus suggests that the CFPB’s current Director, Director Kraninger, ratified the enforcement proceeding by maintaining the suit after she stated her belief that the removal provision is unconstitutional. But the CFPB expressly disclaimed the notion that Director Kraninger had the power to ratify the civil investigative demand, stating that she “remains statutorily insulated from removal, regardless whether she believes the law is invalid.” Reply Brief for Respondent 7.
7  The severability clause refers to three alternative scenarios: (1) a “provision of [the] Act . . . is held to be unconstitutional”; (2) “an amendment made by [the] Act . . . is held unconstitutional”; and (3) “the application of [a] provision or amendment [of the Act] to any person or circumstance is held to be unconstitutional.” 12 U. S. C. §5302. The plurality assumes, with no analysis, that this case falls in the first scenario, calling for a provision to be severed from the Dodd-Frank Act. See ante, at 33. But, as discussed below, there is no single “provision” of the Act that has led to the constitutional injury in this case. See infra, at 20–21. It is the attempted enforcement of a civil investigative demand under §5562(e)(1) by an unconstitutionally insulated Director that causes the constitutional injury in this case. There is at least a nonfrivolous argument that this case implicates the third scenario contemplated by the severability clause—i.e., “the application of [a] provision” in a certain “circumstance.” §5302. If that were so, the text of the severability clause would not require any “provision” to be severed; the unconstitutional application of §5562(e)(1) simply would not affect other provisions of the Dodd-Frank Act. Such a reading would be consistent with the traditional limits on the judicial power. See supra, at 14–15.
8  This statement in Booker is irreconcilable with the plurality’s assertion here that “Congress would prefer that we use a scalpel rather than a bulldozer in curing the constitutional defect.” Ante, at 35. Thus, it appears that the plurality either sub silentio “junk[s] our settled severability doctrine,” ibid., or invokes, without explanation, different assumptions for different cases.


No. 19–7



on writ of certiorari to the united states court of appeals for the ninth circuit

[June 29, 2020]

Justice Kagan, with whom Justice Ginsburg, Justice Breyer, and Justice Sotomayor join, concurring in the judgment with respect to severability and dissenting in part.

Throughout the Nation’s history, this Court has left most decisions about how to structure the Executive Branch to Congress and the President, acting through legislation they both agree to. In particular, the Court has commonly allowed those two branches to create zones of administrative independence by limiting the President’s power to remove agency heads. The Federal Reserve Board. The Federal Trade Commission (FTC). The National Labor Relations Board. Statute after statute establishing such entities instructs the President that he may not discharge their directors except for cause—most often phrased as inefficiency, neglect of duty, or malfeasance in office. Those statutes, whose language the Court has repeatedly approved, provide the model for the removal restriction before us today. If precedent were any guide, that provision would have survived its encounter with this Court—and so would the intended independence of the Consumer Financial Protection Bureau (CFPB).

Our Constitution and history demand that result. The text of the Constitution allows these common for-cause removal limits. Nothing in it speaks of removal. And it grants Congress authority to organize all the institutions of American governance, provided only that those arrangements allow the President to perform his own constitutionally assigned duties. Still more, the Framers’ choice to give the political branches wide discretion over administrative offices has played out through American history in ways that have settled the constitutional meaning. From the first, Congress debated and enacted measures to create spheres of administration—especially of financial affairs—detached from direct presidential control. As the years passed, and governance became ever more complicated, Congress continued to adopt and adapt such measures—confident it had latitude to do so under a Constitution meant to “endure for ages to come.” McCulloch v. Maryland, 4 Wheat. 316, 415 (1819) (approving the Second Bank of the United States). Not every innovation in governance—not every experiment in administrative independence—has proved successful. And debates about the prudence of limiting the President’s control over regulatory agencies, including through his removal power, have never abated.[1] But the Constitution—both as originally drafted and as practiced—mostly leaves disagreements about administrative structure to Congress and the President, who have the knowledge and experience needed to address them. Within broad bounds, it keeps the courts—who do not—out of the picture.

The Court today fails to respect its proper role. It recognizes that this Court has approved limits on the President’s removal power over heads of agencies much like the CFPB. Agencies possessing similar powers, agencies charged with similar missions, agencies created for similar reasons. The majority’s explanation is that the heads of those agencies fall within an “exception”—one for multimember bodies and another for inferior officers—to a “general rule” of unrestricted presidential removal power. Ante, at 13. And the majority says the CFPB Director does not. That account, though, is wrong in every respect. The majority’s general rule does not exist. Its exceptions, likewise, are made up for the occasion—gerrymandered so the CFPB falls outside them. And the distinction doing most of the majority’s work—between multimember bodies and single directors—does not respond to the constitutional values at stake. If a removal provision violates the separation of powers, it is because the measure so deprives the President of control over an official as to impede his own constitutional functions. But with or without a for-cause removal provision, the President has at least as much control over an individual as over a commission—and possibly more. That means the constitutional concern is, if anything, ameliorated when the agency has a single head. Unwittingly, the majority shows why courts should stay their hand in these matters. “Compared to Congress and the President, the Judiciary possesses an inferior understanding of the realities of administration” and the way “political power[ ] operates.” Free Enterprise Fund v. Public Company Accounting Oversight Bd., 561 U.S. 477, 523 (2010) (Breyer, J., dissenting).

In second-guessing the political branches, the majority second-guesses as well the wisdom of the Framers and the judgment of history. It writes in rules to the Constitution that the drafters knew well enough not to put there. It repudiates the lessons of American experience, from the 18th century to the present day. And it commits the Nation to a static version of governance, incapable of responding to new conditions and challenges. Congress and the President established the CFPB to address financial practices that had brought on a devastating recession, and could do so again. Today’s decision wipes out a feature of that agency its creators thought fundamental to its mission—a measure of independence from political pressure. I respectfully dissent.


The text of the Constitution, the history of the country, the precedents of this Court, and the need for sound and adaptable governance—all stand against the majority’s opinion. They point not to the majority’s “general rule” of “unrestricted removal power” with two grudgingly applied “exceptions.” Ante, at 13, 16. Rather, they bestow discretion on the legislature to structure administrative institutions as the times demand, so long as the President retains the ability to carry out his constitutional duties. And most relevant here, they give Congress wide leeway to limit the President’s removal power in the interest of enhancing independence from politics in regulatory bodies like the CFPB.


What does the Constitution say about the separation of powers—and particularly about the President’s removal authority? (Spoiler alert: about the latter, nothing at all.)

The majority offers the civics class version of separation of powers—call it the Schoolhouse Rock definition of the phrase. See Schoolhouse Rock! Three Ring Government (Mar. 13, 1979), pKSGyiT-o3o (“Ring one, Executive. Two is Legislative, that’s Congress. Ring three, Judiciary”). The Constitution’s first three articles, the majority recounts, “split the atom of sovereignty” among Congress, the President, and the courts. Ante, at 21 (internal quotation marks omitted). And by that mechanism, the Framers provided a “simple” fix “to governmental power and its perils.” Ibid.

There is nothing wrong with that as a beginning (except the adjective “simple”). It is of course true that the Framers lodged three different kinds of power in three different entities. And that they did so for a crucial purpose—because, as James Madison wrote, “there can be no liberty where the legislative and executive powers are united in the same person[ ] or body” or where “the power of judging [is] not separated from the legislative and executive powers.” The Federalist No. 47, p. 325 (J. Cooke ed. 1961) (quoting Baron de Montesquieu).

The problem lies in treating the beginning as an ending too—in failing to recognize that the separation of powers is, by design, neither rigid nor complete. Blackstone, whose work influenced the Framers on this subject as on others, observed that “every branch” of government “supports and is supported, regulates and is regulated, by the rest.” 1 W. Blackstone, Commentaries on the Laws of England 151 (1765). So as James Madison stated, the creation of distinct branches “did not mean that these departments ought to have no partial agency in, or no controul over the acts of each other.” The Federalist No. 47, at 325 (emphasis deleted).[2] To the contrary, Madison explained, the drafters of the Constitution—like those of then-existing state constitutions—opted against keeping the branches of government “absolutely separate and distinct.” Id., at 327. Or as Justice Story reiterated a half-century later: “[W]hen we speak of a separation of the three great departments of government,” it is “not meant to affirm, that they must be kept wholly and entirely separate.” 2 J. Story, Commentaries on the Constitution of the United States §524, p. 8 (1833). Instead, the branches have—as they must for the whole arrangement to work—“common link[s] of connexion [and] dependence.” Ibid.

One way the Constitution reflects that vision is by giving Congress broad authority to establish and organize the Executive Branch. Article II presumes the existence of “Officer[s]” in “executive Departments.” §2, cl. 1. But it does not, as you might think from reading the majority opinion, give the President authority to decide what kinds of officers—in what departments, with what responsibilities—the Executive Branch requires. See ante, at 11 (“The entire ‘executive Power’ belongs to the President alone”). Instead, Article I’s Necessary and Proper Clause puts those decisions in the legislature’s hands. Congress has the power “[t]o make all Laws which shall be necessary and proper for carrying into Execution” not just its own enumerated powers but also “all other Powers vested by this Constitution in the Government of the United States, or in any Department or Officer thereof.” §8, cl. 18. Similarly, the Appointments Clause reflects Congress’s central role in structuring the Executive Branch. Yes, the President can appoint principal officers, but only as the legislature “shall . . . establish[ ] by Law” (and of course subject to the Senate’s advice and consent). Art. II, §2, cl. 2. And Congress has plenary power to decide not only what inferior officers will exist but also who (the President or a head of department) will appoint them. So as Madison told the first Congress, the legislature gets to “create[ ] the office, define[ ] the powers, [and] limit[ ] its duration.” 1 Annals of Cong. 582 (1789). The President, as to the construction of his own branch of government, can only try to work his will through the legislative process.[3]

The majority relies for its contrary vision on Article II’s Vesting Clause, see ante, at 11–12, 25, but the provision can’t carry all that weight. Or as Chief Justice Rehnquist wrote of a similar claim in Morrison v. Olson, 487 U.S. 654 (1988), “extrapolat[ing]” an unrestricted removal power from such “general constitutional language”—which says only that “[t]he executive Power shall be vested in a President”—is “more than the text will bear.” Id., at 690, n. 29. Dean John Manning has well explained why, even were it not obvious from the Clause’s “open-ended language.” Separation of Powers as Ordinary Interpretation, 124 Harv. L. Rev. 1939, 1971 (2011). The Necessary and Proper Clause, he writes, makes it impossible to “establish a constitutional violation simply by showing that Congress has constrained the way ‘[t]he executive Power’ is implemented”; that is exactly what the Clause gives Congress the power to do. Id., at 1967. Only “a specific historical understanding” can bar Congress from enacting a given constraint. Id., at 2024. And nothing of that sort broadly prevents Congress from limiting the President’s removal power. I’ll turn soon to the Decision of 1789 and other evidence of Post-Convention thought. See infra, at 9–13. For now, note two points about practice before the Constitution’s drafting. First, in that era, Parliament often restricted the King’s power to remove royal officers—and the President, needless to say, wasn’t supposed to be a king. See Birk, Interrogating the Historical Basis for a Unitary Executive, 73 Stan. L. Rev. (forthcoming 2021). Second, many States at the time allowed limits on gubernatorial removal power even though their constitutions had similar vesting clauses. See Shane, The Originalist Myth of the Unitary Executive, 19 U. Pa. J. Const. L. 323, 334–344 (2016). Historical understandings thus belie the majority’s “general rule.”

Nor can the Take Care Clause come to the majority’s rescue. That Clause cannot properly serve as a “placeholder for broad judicial judgments” about presidential control. Goldsmith & Manning, The Protean Take Care Clause, 164 U. Pa. L. Rev. 1835, 1867 (2016); but see ante, at 11–12, 27–28, n. 11 (using it that way). To begin with, the provision—“he shall take Care that the Laws be faithfully executed”—speaks of duty, not power. Art. II, §3. New scholarship suggests the language came from English and colonial oaths taken by, and placing fiduciary obligations on, all manner and rank of executive officers. See Kent, Leib, & Shugerman, Faithful Execution and Article II, 132 Harv. L. Rev. 2111, 2121–2178 (2019). To be sure, the imposition of a duty may imply a grant of power sufficient to carry it out. But again, the majority’s view of that power ill comports with founding-era practice, in which removal limits were common. See, e.g., Corwin, Tenure of Office and the Removal Power Under the Constitution, 27 Colum. L. Rev. 353, 385 (1927) (noting that New York’s Constitution of 1777 had nearly the same clause, though the State’s executive had “very little voice” in removals). And yet more important, the text of the Take Care Clause requires only enough authority to make sure “the laws [are] faithfully executed”—meaning with fidelity to the law itself, not to every presidential policy preference. As this Court has held, a President can ensure “ ‘faithful execution’ of the laws”—thereby satisfying his “take care” obligation—with a removal provision like the one here. Morrison, 487 U. S., at 692. A for-cause standard gives him “ample authority to assure that [an official] is competently performing [his] statutory responsibilities in a manner that comports with the [relevant legislation’s] provisions.” Ibid.

Finally, recall the Constitution’s telltale silence: Nowhere does the text say anything about the President’s power to remove subordinate officials at will. The majority professes unconcern. After all, it says, “neither is there a ‘separation of powers clause’ or a ‘federalism clause.’ ” Ante, at 25. But those concepts are carved into the Constitution’s text—the former in its first three articles separating powers, the latter in its enumeration of federal powers and its reservation of all else to the States. And anyway, at-will removal is hardly such a “foundational doctrine[ ],” ibid.: You won’t find it on a civics class syllabus. That’s because removal is a tool—one means among many, even if sometimes an important one, for a President to control executive officials. See generally Free Enterprise Fund, 561 U. S., at 524 (Breyer, J., dissenting). To find that authority hidden in the Constitution as a “general rule” is to discover what is nowhere there.


History no better serves the majority’s cause. As Madison wrote, “a regular course of practice” can “liquidate & settle the meaning of ” disputed or indeterminate constitutional provisions. Letter to Spencer Roane (Sept. 2, 1819), in 8 Writings of James Madison 450 (G. Hunt ed. 1908); see NLRB v. Noel Canning, 573 U.S. 513, 525 (2014). The majority lays claim to that kind of record, asserting that its muscular view of “[t]he President’s removal power has long been confirmed by history.” Ante, at 12. But that is not so. The early history—including the fabled Decision of 1789—shows mostly debate and division about removal authority. And when a “settle[ment of] meaning” at last occurred, it was not on the majority’s terms. Instead, it supports wide latitude for Congress to create spheres of administrative independence.


Begin with evidence from the Constitution’s ratification. And note that this moment is indeed the beginning: Delegates to the Constitutional Convention never discussed whether or to what extent the President would have power to remove executive officials. As a result, the Framers advocating ratification had no single view of the matter. In Federalist No. 77, Hamilton presumed that under the new Constitution “[t]he consent of [the Senate] would be necessary to displace as well as to appoint” officers of the United States. Id., at 515. He thought that scheme would promote “steady administration”: “Where a man in any station had given satisfactory evidence of his fitness for it, a new president would be restrained” from substituting “a person more agreeable to him.” Ibid. By contrast, Madison thought the Constitution allowed Congress to decide how any executive official could be removed. He explained in Federalist No. 39: “The tenure of the ministerial offices generally will be a subject of legal regulation, conformably to the reason of the case, and the example of the State Constitutions.” Id., at 253. Neither view, of course, at all supports the majority’s story.[4]

The second chapter is the Decision of 1789, when Congress addressed the removal power while considering the bill creating the Department of Foreign Affairs. Speaking through Chief Justice Taft—a judicial presidentialist if ever there was one—this Court in Myers v. United States, 272 U.S. 52 (1926), read that debate as expressing Congress’s judgment that the Constitution gave the President illimitable power to remove executive officials. The majority rests its own historical claim on that analysis (though somehow also finding room for its two exceptions). See ante, at 12–13. But Taft’s historical research has held up even worse than Myers’ holding (which was mostly reversed, see infra, at 17–18). As Dean Manning has concluded after reviewing decades’ worth of scholarship on the issue, “the implications of the debate, properly understood, [are] highly ambiguous and prone to overreading.” Manning, 124 Harv. L. Rev., at 1965, n. 135; see id., at 2030–2031.

The best view is that the First Congress was “deeply divided” on the President’s removal power, and “never squarely addressed” the central issue here. Id., at 1965, n. 135; Prakash, New Light on the Decision of 1789, 91 Cornell L. Rev. 1021, 1072 (2006). The congressional debates revealed three main positions. See Corwin, 27 Colum. L. Rev., at 361. Some shared Hamilton’s Federalist No. 77 view: The Constitution required Senate consent for removal. At the opposite extreme, others claimed that the Constitution gave absolute removal power to the President. And a third faction maintained that the Constitution placed Congress in the driver’s seat: The legislature could regulate, if it so chose, the President’s authority to remove. In the end, Congress passed a bill saying nothing about removal, leaving the President free to fire the Secretary of Foreign Affairs at will. But the only one of the three views definitively rejected was Hamilton’s theory of necessary Senate consent. As even strong proponents of executive power have shown, Congress never “endorse[d] the view that [it] lacked authority to modify” the President’s removal authority when it wished to. Prakash, supra, at 1073; see Manning, supra, at 1965, n. 135, 2030–2031. The summer of 1789 thus ended without resolution of the critical question: Was the removal power “beyond the reach of congressional regulation?” Prakash, supra, at 1072.

At the same time, the First Congress gave officials han- dling financial affairs—as compared to diplomatic and military ones—some independence from the President. The title and first section of the statutes creating the Departments of Foreign Affairs and War designated them “executive departments.” Act of July 27, 1789, ch. 4, 1Stat. 28; Act of Aug. 7, 1789, ch. 7, 1Stat. 49. The law creating the Treasury Department conspicuously avoided doing so. See Act of Sept. 2, 1789, ch. 12, 1Stat. 65. That difference in nomenclature signaled others of substance. Congress left the organization of the Departments of Foreign Affairs and War skeletal, enabling the President to decide how he wanted to staff them. See Casper, An Essay in Separation of Powers, 30 Wm. & Mary L. Rev. 211, 239–241 (1989). By contrast, Congress listed each of the offices within the Treasury Department, along with their functions. See ibid. Of the three initial Secretaries, only the Treasury’s had an obligation to report to Congress when requested. See §2, 1Stat. 65–66. And perhaps most notable, Congress soon deemed the Comptroller of the Treasury’s settlements of public accounts “final and conclusive.” Act of Mar. 3, 1795, ch. 48, §4, 1Stat. 441–442. That decision, preventing presidential overrides, marked the Comptroller as exercising independent judgment.[5] True enough, no statute shielded the Comptroller from discharge. But even James Madison, who at this point opposed most removal limits, told Congress that “there may be strong reasons why an officer of this kind should not hold his office at the pleasure” of the Secretary or President. 1 Annals of Cong. 612. At the least, as Professor Prakash writes, “Madison maintained that Congress had the [constitutional] authority to modify [the Comptroller’s] tenure.” Prakash, supra, at 1071.

Contrary to the majority’s view, then, the founding era closed without any agreement that Congress lacked the power to curb the President’s removal authority. And as it kept that question open, Congress took the first steps—which would launch a tradition—of distinguishing financial regulators from diplomatic and military officers. The latter mainly helped the President carry out his own constitutional duties in foreign relations and war. The former chiefly carried out statutory duties, fulfilling functions Congress had assigned to their offices. In addressing the new Nation’s finances, Congress had begun to use its powers under the Necessary and Proper Clause to design effective administrative institutions. And that included taking steps to insulate certain officers from political influence.


As the decades and centuries passed, those efforts picked up steam. Confronting new economic, technological, and social conditions, Congress—and often the President—saw new needs for pockets of independence within the federal bureaucracy. And that was especially so, again, when it came to financial regulation. I mention just a few highlights here—times when Congress decided that effective governance depended on shielding technical or expertise-based functions relating to the financial system from political pressure (or the moneyed interests that might lie behind it). Enacted under the Necessary and Proper Clause, those measures—creating some of the Nation’s most enduring institutions—themselves helped settle the extent of Congress’s power. “[A] regular course of practice,” to use Madison’s phrase, has “liquidate[d]” constitutional meaning about the permissibility of independent agencies. See supra, at 9.

Take first Congress’s decision in 1816 to create the Second Bank of the United States—“the first truly independent agency in the republic’s history.” Lessig & Sunstein, The President and the Administration, 94 Colum. L. Rev. 1, 30 (1994). Of the twenty-five directors who led the Bank, the President could appoint and remove only five. See Act of Apr. 10, 1816, §8, 3Stat. 269. Yet the Bank had a greater impact on the Nation than any but a few institutions, regulating the Nation’s money supply in ways anticipating what the Federal Reserve does today. Of course, the Bank was controversial—in large part because of its freedom from presidential control. Andrew Jackson chafed at the Bank’s independence and eventually fired his Treasury Secretary for keeping public moneys there (a dismissal that itself provoked a political storm). No matter. Innovations in governance always have opponents; administrative independence predictably (though by no means invariably) provokes presidential ire. The point is that by the early 19th century, Congress established a body wielding enormous financial power mostly outside the President’s dominion.

The Civil War brought yet further encroachments on presidential control over financial regulators. In response to wartime economic pressures, President Lincoln (not known for his modest view of executive power) asked Congress to establish an office called the Comptroller of the Currency. The statute he signed made the Comptroller removable only with the Senate’s consent—a version of the old Hamiltonian idea, though this time required not by the Constitution itself but by Congress. See Act of Feb. 25, 1863, ch. 58, 12Stat. 665. A year later, Congress amended the statute to permit removal by the President alone, but only upon “reasons to be communicated by him to the Senate.” Act of June 3, 1864, §1, 13Stat. 100. The majority dismisses the original version of the statute as an “aberration.” Ante, at 19. But in the wake of the independence given first to the Comptroller of the Treasury and then to the national Bank, it’s hard to conceive of this newest Comptroller position as so great a departure. And even the second iteration of the statute preserved a constraint on the removal power, requiring a President in a firing mood to explain himself to Congress—a demand likely to make him sleep on the subject. In both versions of the law, Congress responded to new financial challenges with new regulatory institutions, alert to the perils in this area of political interference.[6]

And then, nearly a century and a half ago, the floodgates opened. In 1887, the growing power of the railroads over the American economy led Congress to create the Interstate

Commerce Commission. Under that legislation, the Presi- dent could remove the five Commissioners only “for inefficiency, neglect of duty, or malfeasance in office”—the same standard Congress applied to the CFPB Director. Act of Feb. 4, 1887, §11, 24Stat. 383. More—many more—for-cause removal provisions followed. In 1913, Congress gave the Governors of the Federal Reserve Board for-cause protection to ensure the agency would resist political pressure and promote economic stability. See Act of Dec. 23, 1913, ch. 6, 38Stat. 251. The next year, Congress provided similar protection to the FTC in the interest of ensuring “a continuous policy” “free from the effect” of “changing [White House] incumbency.” 51 Cong. Rec. 10376 (1914). The Federal Deposit Insurance Corporation (FDIC), the Securities and Exchange Commission (SEC), the Commodity Futures Trading Commission. In the financial realm, “independent agencies have remained the bedrock of the institutional framework governing U. S. markets.” Gadinis, From Independence to Politics in Financial Regulation, 101 Cal. L. Rev. 327, 331 (2013). By one count, across all subject matter areas, 48 agencies have heads (and below them hundreds more inferior officials) removable only for cause. See Free Enterprise Fund, 561 U. S., at 541 (Breyer, J., dissenting). So year by year by year, the broad sweep of history has spoken to the constitutional question before us: Independent agencies are everywhere.


What is more, the Court’s precedents before today have accepted the role of independent agencies in our governmental system. To be sure, the line of our decisions has not run altogether straight. But we have repeatedly upheld provisions that prevent the President from firing regulatory officials except for such matters as neglect or malfeasance. In those decisions, we sounded a caution, insisting that Congress could not impede through removal restrictions the President’s performance of his own constitutional duties. (So, to take the clearest example, Congress could not curb the President’s power to remove his close military or diplomatic advisers.) But within that broad limit, this Court held, Congress could protect from at-will removal the officials it deemed to need some independence from political pressures. Nowhere do those precedents suggest what the majority announces today: that the President has an “unrestricted removal power” subject to two bounded exceptions. Ante, at 2.

The majority grounds its new approach in Myers, ignoring the way this Court has cabined that decision. Myers, the majority tells us, found an unrestrained removal power “essential to the [President’s] execution of the laws.” Ante, at 13 (quoting Myers, 272 U. S., at 117). What the majority does not say is that within a decade the Court abandoned that view (much as later scholars rejected Taft’s one-sided history, see supra, at 10–11). In Humphrey’s Executor v. United States, 295 U.S. 602 (1935), the Court unceremoniously—and unanimously—confined Myers to its facts. “[T]he narrow point actually decided” there, Humphrey’s stated, was that the President could “remove a postmaster of the first class, without the advice and consent of the Senate.” 295 U. S., at 626. Nothing else in Chief Justice Taft’s prolix opinion “c[a]me within the rule of stare decisis.” Ibid. (Indeed, the Court went on, everything in Myers “out of harmony” with Humphrey’s was expressly “disapproved.” 295 U. S., at 626.) Half a century later, the Court was more generous. Two decisions read Myers as standing for the principle that Congress’s own “participation in the removal of executive officers is unconstitutional.” Bowsher v. Synar, 478 U.S. 714, 725 (1986); see Morrison, 487 U. S., at 686 (“As we observed in Bowsher, the essence” of “Myers was the judgment that the Constitution prevents Congress from draw[ing] to itself ” the power to remove (internal quotation marks omitted)). Bowsher made clear that Myers had nothing to say about Congress’s power to enact a provision merely “limit[ing] the President’s powers of removal” through a for-cause provision. 478 U. S., at 724. That issue, the Court stated, was “not presented” in “the Myers case.” Ibid. Instead, the relevant cite was Humphrey’s.

And Humphrey’s found constitutional a statute identical to the one here, providing that the President could remove FTC Commissioners for “inefficiency, neglect of duty, or malfeasance in office.” 295 U. S., at 619. The Humphrey’s Court, as the majority notes, relied in substantial part on what kind of work the Commissioners performed. See id., at 628, 631; ante, at 14. (By contrast, nothing in the decision turned—as the majority suggests, see ante, at 14–15—on any of the agency’s organizational features. See infra, at 30.) According to Humphrey’s, the Commissioners’ primary work was to “carry into effect legislative policies”—“filling in and administering the details embodied by [a statute’s] general standard.” 295 U. S., at 627–628. In addition, the Court noted, the Commissioners recommended dispositions in court cases, much as a special master does. Given those “quasi-legislative” and “quasi-judicial”—as opposed to “purely executive”—functions, Congress could limit the President’s removal authority. Id., at 628.[7] Or said another way, Congress could give the FTC some “independen[ce from] executive control.” Id., at 629.

About two decades later, an again-unanimous Court in Wiener v. United States, 357 U.S. 349 (1958), reaffirmed Humphrey’s. The question in Wiener was whether the President could dismiss without cause members of the War Claims Commission, an entity charged with compensating injuries arising from World War II. Disdaining Myers and relying on Humphrey’s, the Court said he could not. The Court described as “short-lived” Myers’ view that the President had “inherent constitutional power to remove officials, no matter what the relation of the executive to the discharge of their duties.” 357 U. S., at 352.[8] Here, the Commissioners were not close agents of the President, who needed to be responsive to his preferences. Rather, they exercised adjudicatory responsibilities over legal claims. Congress, the Court found, had wanted the Commissioners to do so “free from [political] control or coercive influence.” Id., at 355 (quoting Humphrey’s, 295 U. S., at 629). And that choice, as Humphrey’s had held, was within Congress’s power. The Constitution enabled Congress to take down “the Damocles’ sword of removal” hanging over the Commissioners’ heads. 357 U. S., at 356.

Another three decades on, Morrison both extended Humphrey’s domain and clarified the standard for addressing removal issues. The Morrison Court, over a one-Justice dissent, upheld for-cause protections afforded to an independent counsel with power to investigate and prosecute crimes committed by high-ranking officials. The Court well understood that those law enforcement functions differed from the rulemaking and adjudicatory duties highlighted in Humphrey’s and Wiener. But that difference did not resolve the issue. An official’s functions, Morrison held, were relevant to but not dispositive of a removal limit’s constitutionality. The key question in all the cases, Morrison saw, was whether such a restriction would “impede the President’s ability to perform his constitutional duty.” 487 U. S., at 691. Only if it did so would it fall outside Congress’s power. And the protection for the independent counsel, the Court found, did not. Even though the counsel’s functions were “purely executive,” the President’s “need to control the exercise of [her] discretion” was not “so central to the functioning of the Executive Branch as to require” unrestricted removal authority. Id., at 690–691. True enough, the Court acknowledged, that the for-cause standard prevented the President from firing the counsel for discretionary decisions or judgment calls. But it preserved “ample authority” in the President “to assure that the counsel is competently performing” her “responsibilities in a manner that comports with” all legal requirements. Id., at 692. That meant the President could meet his own constitutional obligation “to ensure ‘the faithful execution’ of the laws.” Ibid.; see supra, at 8.[9]

The majority’s description of Morrison, see ante, at 15–16, is not true to the decision. (Mostly, it seems, the majority just wishes the case would go away. See ante, at 17, n. 4.) First, Morrison is no “exception” to a broader rule from Myers. Morrison echoed all of Humphrey’s criticism of the by-then infamous Myers “dicta.” 487 U. S., at 687. It again rejected the notion of an “all-inclusive” removal power. Ibid. It yet further confined Myers’ reach, making clear that Congress could restrict the President’s removal of officials carrying out even the most traditional executive functions. And the decision, with care, set out the governing rule—again, that removal restrictions are permissible so long as they do not impede the President’s performance of his own constitutionally assigned duties. Second, as all that suggests, Morrison is not limited to inferior officers. In the eight pages addressing the removal issue, the Court constantly spoke of “officers” and “officials” in general. 487 U. S., at 685–693. By contrast, the Court there used the word “inferior” in just one sentence (which of course the majority quotes), when applying its general standard to the case’s facts. Id., at 691. Indeed, Justice Scalia’s dissent emphasized that the counsel’s inferior-office status played no role in the Court’s decision. See id., at 724 (“The Court could have resolved the removal power issue in this case by simply relying” on that status, but did not). As Justice Scalia noted, the Court in United States v. Perkins, 116 U.S. 483, 484–485 (1886), had a century earlier allowed Congress to restrict the President’s removal power over inferior officers. See Morrison, 487 U. S., at 723–724. Were that Morrison’s basis, a simple citation would have sufficed.

Even Free Enterprise Fund, in which the Court recently held a removal provision invalid, operated within the framework of this precedent—and in so doing, left in place a removal provision just like the one here. In that case, the Court considered a “highly unusual” scheme of double for-cause protection. 561 U. S., at 505. Members of an accounting board were protected from removal by SEC Commissioners, who in turn were protected from removal by the President. The Court found that the two-layer structure deprived the President of “adequate control” over the Board members. Id., at 508. The scheme “impaired” the President’s “ability to execute the laws,” the Court explained, because neither he nor any fully dependent agent could decide “whether[ ] good cause exists” for a discharge. Id., at 495–496. That holding cast no doubt on ordinary for-cause protections, of the kind in the Court’s prior cases (and here as well). Quite the opposite. The Court observed that it did not “take issue with for-cause limitations in general”—which do enable the President to determine whether good cause for discharge exists (because, say, an official has violated the law). Id., at 501. And the Court’s solution to the constitutional problem it saw was merely to strike one level of insulation, making the Board removable by the SEC at will. That remedy left the SEC’s own for-cause protection in place. The President could thus remove Commissioners for malfeasance or neglect, but not for policy disagreements. See ante, at 28.

So caselaw joins text and history in establishing the general permissibility of for-cause provisions giving some independence to agencies. Contrary to the majority’s view, those laws do not represent a suspicious departure from illimitable presidential control over administration. For almost a century, this Court has made clear that Congress has broad discretion to enact for-cause protections in pursuit of good governance.


The deferential approach this Court has taken gives Con-

gress the flexibility it needs to craft administrative agencies. Diverse problems of government demand diverse solutions. They call for varied measures and mixtures of democratic accountability and technical expertise, energy and efficiency. Sometimes, the arguments push toward tight presidential control of agencies. The President’s engagement, some people say, can disrupt bureaucratic stagnation, counter industry capture, and make agencies more responsive to public interests. See, well, Kagan, Presidential Administration, 114 Harv. L. Rev. 2245, 2331–2346 (2001). At other times, the arguments favor greater independence from presidential involvement. Insulation from political pressure helps ensure impartial adjudications. It places technical issues in the hands of those most capable of addressing them. It promotes continuity, and prevents short-term electoral interests from distorting policy. (Consider, for example, how the Federal Reserve’s independence stops a President trying to win a second term from manipulating interest rates.) Of course, the right balance between presidential control and independence is often uncertain, contested, and value-laden. No mathematical formula governs institutional design; trade-offs are endemic to the enterprise. But that is precisely why the issue is one for the political branches to debate—and then debate again as times change. And it’s why courts should stay (mostly) out of the way. Rather than impose rigid rules like the majority’s, they should let Congress and the President figure out what blend of independence and political control will best enable an agency to perform its intended functions.

Judicial intrusion into this field usually reveals only how little courts know about governance. Even everything I just said is an over-simplification. It suggests that agencies can easily be arranged on a spectrum, from the most to the least presidentially controlled. But that is not so. A given agency’s independence (or lack of it) depends on a wealth of features, relating not just to removal standards, but also to appointments practices, procedural rules, internal organization, oversight regimes, historical traditions, cultural norms, and (inevitably) personal relationships. It is hard to pinpoint how those factors work individually, much less in concert, to influence the distance between an agency and a President. In that light, even the judicial opinions’ perennial focus on removal standards is a bit of a puzzle. Removal is only the most obvious, not necessarily the most potent, means of control. See generally Free Enterprise Fund, 561 U. S., at 524 (Breyer, J., dissenting). That is because informal restraints can prevent Presidents from firing at-will officers—and because other devices can keep officers with for-cause protection under control. Of course no court, as Free Enterprise Fund noted, can accurately assess the “bureaucratic minutiae” affecting a President’s influence over an agency. Id., at 500 (majority opinion); ante, at 30 (reprising the point). But that is yet more reason for courts to defer to the branches charged with fashioning administrative structures, and to hesitate before ruling out agency design specs like for-cause removal standards.

Our Constitution, as shown earlier, entrusts such decisions to more accountable and knowledgeable actors. See supra, at 4–9. The document—with great good sense—sets out almost no rules about the administrative sphere. As Chief Justice Marshall wrote when he upheld the first independent financial agency: “To have prescribed the means by which government should, in all future time, execute its powers, would have been to change, entirely, the character of the instrument.” McCulloch, 4 Wheat., at 415. That would have been, he continued, “an unwise attempt to provide, by immutable rules, for exigencies which, if foreseen at all, must have been seen dimly.” Ibid. And if the Constitution, for those reasons, does not lay out immutable rules, then neither should judges. This Court has usually respected that injunction. It has declined to second-guess the work of the political branches in creating independent agencies like the CFPB. In reversing course today—in spurning a “pragmatic, flexible approach to American governance” in favor of a dogmatic, inflexible one, ante, at 29—the majority makes a serious error.


As the majority explains, the CFPB emerged out of disaster. The collapse of the subprime mortgage market “precipitat[ed] a financial crisis that wiped out over $10 trillion in American household wealth and cost millions of Americans their jobs, their retirements, and their homes.” Ante, at 3. In that moment of economic ruin, the President proposed and Congress enacted legislation to address the causes of the collapse and prevent a recurrence. An important part of that statute created an agency to protect consumers from exploitative financial practices. The agency would take over enforcement of almost 20 existing federal laws. See 12 U. S. C. §5581. And it would administer a new prohibition on “unfair, deceptive, or abusive act[s] or practice[s]” in the consumer-finance sector. §5536(a)(1)(B).

No one had a doubt that the new agency should be independent. As explained already, Congress has historically given—with this Court’s permission—a measure of independence to financial regulators like the Federal Reserve Board and the FTC. See supra, at 11–16. And agencies of that kind had administered most of the legislation whose enforcement the new statute transferred to the CFPB. The law thus included an ordinary for-cause provision—once again, that the President could fire the CFPB’s Director only for “inefficiency, neglect of duty, or malfeasance in office.” §5491(c)(3). That standard would allow the President to discharge the Director for a failure to “faithfully execute[ ]” the law, as well as for basic incompetence. U. S. Const., Art. II, §3; see supra, at 8, 20. But it would not permit removal for policy differences.

The question here, which by now you’re well equipped to answer, is whether including that for-cause standard in the statute creating the CFPB violates the Constitution.


Applying our longstanding precedent, the answer is clear: It does not. This Court, as the majority acknowledges, has sustained the constitutionality of the FTC and similar independent agencies. See ante, at 2, 13–16. The for-cause protections for the heads of those agencies, the Court has found, do not impede the President’s ability to perform his own constitutional duties, and so do not breach the separation of powers. See supra, at 18–22. There is nothing different here. The CFPB wields the same kind of power as the FTC and similar agencies. And all of their heads receive the same kind of removal protection. No less than those other entities—by now part of the fabric of government—the CFPB is thus a permissible exercise of Congress’s power under the Necessary and Proper Clause to structure administration.

First, the CFPB’s powers are nothing unusual in the universe of independent agencies. The CFPB, as the majority notes, can issue regulations, conduct its own adjudications, and bring civil enforcement actions in court—all backed by the threat of penalties. See ante, at 1; 12 U. S. C. §§5512, 5562–5565. But then again, so too can (among others) the FTC and SEC, two agencies whose regulatory missions parallel the CFPB’s. See 15 U. S. C. §§45, 53, 57a, 57b–3, 78u, 78v, 78w. Just for a comparison, the CFPB now has 19 enforcement actions pending, while the SEC brought 862 such actions last year alone. See Brief for Petitioner 7; SEC, Div. of Enforcement 2019 Ann. Rep. 14. And although the majority bemoans that the CFPB can “bring the coercive power of the state to bear on millions of private citizens,” ante, at 18, that scary-sounding description applies to most independent agencies. Forget that the more relevant factoid for those many citizens might be that the CFPB has recovered over $11 billion for banking consumers. See ante, at 5. The key point here is that the CFPB got the mass of its regulatory authority from other independent agencies that had brought the same “coercive power to bear.” See 12 U. S. C. §5581 (transferring power from, among others, the Federal Reserve, FTC, and FDIC). Congress, to be sure, gave the CFPB new authority over “unfair, deceptive, or abusive act[s] or practice[s]” in transactions involving a “consumer financial product or service.” §§5517(a)(1), 5536(a)(1). But again, the FTC has power to go after “unfair or deceptive acts or practices in or affecting commerce”—a portfolio spanning a far wider swath of the economy. 15 U. S. C. §45(a)(1).[10] And if influence on economic life is the measure, consider the Federal Reserve, whose every act has global consequence. The CFPB, gauged by that comparison, is a piker.

Second, the removal protection given the CFPB’s Director is standard fare. The removal power rests with the President alone; Congress has no role to play, as it did in the laws struck down in Myers and Bowsher. See supra, at 17–18. The statute provides only one layer of protection, unlike the law in Free Enterprise Fund. See supra, at 21–22. And the clincher, which you have heard before: The for-cause standard used for the CFPB is identical to the one the Court upheld in Humphrey’s. Both enable the President to fire an agency head for “inefficiency, neglect of duty, or malfeasance in office.” See 12 U. S. C. §5491(c)(3); 15 U. S. C. §41; supra, at 18. A removal provision of that kind applied to a financial agency head, this Court has held, does not “unduly trammel[ ] on executive authority,” even though it prevents the President from dismissing the official for a discretionary policy judgment. Morrison, 487 U. S., at 691. Once again: The removal power has not been “completely stripped from the President,” providing him with no means to “ensure the ‘faithful execution’ of the laws.” Id., at 692; see supra, at 20. Rather, this Court has explained, the for-cause standard gives the President “ample authority to assure that [the official] is competently performing his or her statutory responsibilities in a manner that comports with” all legal obligations. 487 U. S., at 692; see supra, at 20. In other words—and contra today’s majority—the President’s removal power, though not absolute, gives him the “meaningful[ ] control[ ]” of the Director that the Constitution requires. Ante, at 23.

The analysis is as simple as simple can be. The CFPB Director exercises the same powers, and receives the same removal protections, as the heads of other, constitutionally permissible independent agencies. How could it be that this opinion is a dissent?


The majority focuses on one (it says sufficient) reason: The CFPB Director is singular, not plural. “Instead of placing the agency under the leadership of a board with multiple members,” the majority protests, “Congress provided that the CFPB would be led by a single Director.” Ante, at 1.[11] And a solo CFPB Director does not fit within either of the majority’s supposed exceptions. He is not an inferior officer, so (the majority says) Morrison does not apply; and he is not a multimember board, so (the majority says) neither does Humphrey’s. Further, the majority argues, “[a]n agency with a [unitary] structure like that of the CFPB” is “novel”—or, if not quite that, “almost wholly unprecedented.” Ante, at 2, 18. Finally, the CFPB’s organizational form violates the “constitutional structure” because it vests power in a “single individual” who is “insulated from Presidential control.” Ante, at 2–3, 23.

I’m tempted at this point just to say: No. All I’ve explained about constitutional text, history, and precedent invalidates the majority’s thesis. But I’ll set out here some more targeted points, taking step by step the majority’s reasoning.

First, as I’m afraid you’ve heard before, the majority’s “exceptions” (like its general rule) are made up. See supra, at 16–22. To begin with, our precedents reject the very idea of such exceptions. “The analysis contained in our removal cases,” Morrison stated, shuns any attempt “to define rigid categories” of officials who may (or may not) have job protection. 487 U. S., at 689. Still more, the contours of the majority’s exceptions don’t connect to our decisions’ reasoning. The analysis in Morrison, as I’ve shown, extended far beyond inferior officers. See supra, at 20–21. And of course that analysis had to apply to individual officers: The independent counsel was very much a person, not a committee. So the idea that Morrison is in a separate box from this case doesn’t hold up.[12] Similarly, Humphrey’s and later precedents give no support to the majority’s view that the number of people at the apex of an agency matters to the constitutional issue. Those opinions mention the “groupness” of the agency head only in their background sections. The majority picks out that until-now-irrelevant fact to distinguish the CFPB, and constructs around it an until-now-unheard-of exception. So if the majority really wants to see something “novel,” ante, at 2, it need only look to its opinion.

By contrast, the CFPB’s single-director structure has a fair bit of precedent behind it. The Comptroller of the Currency. The Office of the Special Counsel (OSC). The Social Security Administration (SSA). The Federal Housing Finance Agency (FHFA). Maybe four prior agencies is in the eye of the beholder, but it’s hardly nothing. I’ve already explained why the earliest of those agencies—the Civil-War-era Comptroller—is not the blip the majority describes. See supra, at 14–15. The office is one in a long line, starting with the founding-era Comptroller of the Treasury (also one person), of financial regulators designed to do their jobs with some independence. As for the other three, the majority objects: too powerless and too contested. See ante, at 18–21. I think not. On power, the SSA runs the Nation’s largest government program—among other things, deciding all claims brought by its 64 million beneficiaries; the FHFA plays a crucial role in overseeing the mortgage market, on which millions of Americans annually rely; and the OSC prosecutes misconduct in the two-million-person federal workforce. All different from the CFPB, no doubt; but the majority can’t think those matters beneath a President’s notice. (Consider: Would the President lose more votes from a malfunctioning SSA or CFPB?) And controversial? Well, yes, they are. Almost all independent agencies are controversial, no matter how many directors they have. Or at least controversial among Presidents and their lawyers. That’s because whatever might be said in their favor, those agencies divest the President of some removal power. If signing statements and veto threats made independent agencies unconstitutional, quite a few wouldn’t pass muster. Maybe that’s what the majority really wants (I wouldn’t know)—but it can’t pretend the disputes surrounding these agencies had anything to do with whether their heads are singular or plural.

Still more important, novelty is not the test of constitutionality when it comes to structuring agencies. See Mistretta v. United States, 488 U.S. 361, 385 (1989) (“[M]ere anomaly or innovation” does not violate the separation of powers). Congress regulates in that sphere under the Necessary and Proper Clause, not (as the majority seems to think) a Rinse and Repeat Clause. See supra, at 6. The Framers understood that new times would often require new measures, and exigencies often demand innovation. See McCulloch, 4 Wheat., at 415; supra, at 24. In line with that belief, the history of the administrative sphere—its rules, its practices, its institutions—is replete with experiment and change. See supra, at 9–16. Indeed, each of the agencies the majority says now fits within its “exceptions” was once new; there is, as the saying goes, “a first time for everything.” National Federation of Independent Business v. Sebelius, 567 U.S. 519, 549 (2012). So even if the CFPB differs from its forebears in having a single director, that departure is not itself “telling” of a “constitutional problem.” Ante, at 18. In deciding what this moment demanded, Congress had no obligation to make a carbon copy of a design from a bygone era.

And Congress’s choice to put a single director, rather than a multimember commission, at the CFPB’s head violates no principle of separation of powers. The purported constitutional problem here is that an official has “slip[ped] from the Executive’s control” and “supervision”—that he has become unaccountable to the President. Ante, at 23, 25 (internal quotation marks omitted). So to make sense on the majority’s own terms, the distinction between singular and plural agency heads must rest on a theory about why the former more easily “slip” from the President’s grasp. But the majority has nothing to offer. In fact, the opposite is more likely to be true: To the extent that such matters are measurable, individuals are easier than groups to supervise.

To begin with, trying to generalize about these matters is something of a fool’s errand. Presidential control, as noted earlier, can operate through many means—removal to be sure, but also appointments, oversight devices (e.g., centralized review of rulemaking or litigating positions), budgetary processes, personal outreach, and more. See Free Enterprise Fund, 561 U. S., at 524 (Breyer, J., dissenting); supra, at 23–24.[13] The effectiveness of each of those control mechanisms, when present, can then depend on a multitude of agency-specific practices, norms, rules, and organizational features. In that complex stew, the difference between a singular and plural agency head will often make not a whit of difference. Or to make the point more concrete, a multimember commission may be harder to control than an individual director for a host of reasons unrelated to its plural character. That may be so when the two are subject to the same removal standard, or even when the individual director has greater formal job protection. Indeed, the very category of multimember commissions breaks apart under inspection, spoiling the majority’s essential dichotomy. See generally Brief for Rachel E. Barkow et al. as Amici Curiae. Some of those commissions have chairs appointed by the President; others do not. Some of those chairs are quite powerful; others are not. Partisan balance requirements, term length, voting rules, and more—all vary widely, in ways that make a significant difference to the ease of presidential control. Why, then, would anyone distinguish along a simple commission/single-director axis when deciding whether the Constitution requires at-will removal?

But if the demand is for generalization, then the majority’s distinction cuts the opposite way: More powerful control mechanisms are needed (if anything) for commissions. Holding everything else equal, those are the agencies more likely to “slip from the Executive’s control.” Ante, at 25. Just consider your everyday experience: It’s easier to get one person to do what you want than a gaggle. So too, you know exactly whom to blame when an individual—but not when a group—does a job badly. The same is true in bureaucracies. A multimember structure reduces accountability to the President because it’s harder for him to oversee, to influence—or to remove, if necessary—a group of five or more commissioners than a single director. Indeed, that is why Congress so often resorts to hydra-headed agencies. “[M]ultiple membership,” an influential Senate Report concluded, is “a buffer against Presidential control” (especially when combined, as it often is, with partisan-balance requirements). Senate Committee on Governmental Affairs, Study on Federal Regulation, S. Doc. No. 95–91, vol. 5, p. 75 (1977). So, for example, Congress constructed the Federal Reserve as it did because it is “easier to protect a board from political control than to protect a single appointed official.” R. Cushman, The Independent Regulatory Commissions 153 (1941).[14] It is hard to know why Congress did not take the same tack when creating the CFPB. But its choice brought the agency only closer to the President—more exposed to his view, more subject to his sway. In short, the majority gets the matter backward: Where presidential control is the object, better to have one than many.

Because it has no answer on that score, the majority slides to a different question: Assuming presidential control of any independent agency is vanishingly slim, is a single-head or a multi-head agency more capable of exercising power, and so of endangering liberty? See ante, at 21–23. The majority says a single head is the greater threat because he may wield power “unilaterally” and “[w]ith no colleagues to persuade.” Ante, at 23 (emphasis in original). So the CFPB falls victim to what the majority sees as a constitutional anti-power-concentration principle (with an exception for the President).

If you’ve never heard of a statute being struck down on that ground, you’re not alone. It is bad enough to “extrapolat[e]” from the “general constitutional language” of Article II’s Vesting Clause an unrestricted removal power constraining Congress’s ability to legislate under the Necessary and Proper Clause. Morrison, 487 U. S., at 690, n. 29; see supra, at 7. It is still worse to extrapolate from the Constitution’s general structure (division of powers) and implicit values (liberty) a limit on Congress’s express power to create administrative bodies. And more: to extrapolate from such sources a distinction as prosaic as that between the SEC and the CFPB—i.e., between a multi-headed and single-headed agency. That is, to adapt a phrase (or two) from our precedent, “more than” the emanations of “the text will bear.” Morrison, 487 U. S., at 690, n. 29. By using abstract separation-of-powers arguments for such purposes, the Court “appropriate[s]” the “power delegated to Congress by the Necessary and Proper Clause” to compose the government. Manning, Foreword: The Means of Constitutional Power, 128 Harv. L. Rev. 1, 78 (2014). In deciding for itself what is “proper,” the Court goes beyond its own proper bounds.

And in doing so, the majority again reveals its lack of interest in how agencies work. First, the premise of the majority’s argument—that the CFPB head is a mini-dictator, not subject to meaningful presidential control, see ante, at 23—is wrong. As this Court has seen in the past, independent agencies are not fully independent. A for-cause removal provision, as noted earlier, leaves “ample” control over agency heads in the hands of the President. Morrison, 487 U. S., at 692; see supra, at 20. He can discharge them for failing to perform their duties competently or in accordance with law, and so ensure that the laws are “faithfully executed.” U. S. Const., Art. II, §3; see supra, at 8, 20. And he can use the many other tools attached to the Office of the Presidency—including in the CFPB’s case, rulemaking review—to exert influence over discretionary policy calls. See supra, at 33, and n. 13. Second, the majority has nothing but intuition to back up its essentially functionalist claim that the CFPB would be less capable of exercising power if it had more than one Director (even supposing that were a suitable issue for a court to address). Ante, at 21, 23. Maybe the CFPB would be. Or maybe not. Although a multimember format tends to frustrate the President’s control over an agency, see supra, at 34–35, it may not lessen the agency’s own ability to act with decision and dispatch. (Consider, for a recent example, the Federal Reserve Board.) That effect presumably would depend on the agency’s internal organization, voting rules, and similar matters. At the least: If the Court is going to invalidate statutes based on empirical assertions like this one, it should offer some empirical support. It should not pretend that its assessment that the CFPB wields more power more dangerously than the SEC comes from someplace in the Constitution. But today the majority fails to accord even that minimal respect to Congress.


Recall again how this dispute got started. In the midst of the Great Recession, Congress and the President came together to create an agency with an important mission. It would protect consumers from the reckless financial practices that had caused the then-ongoing economic collapse. Not only Congress but also the President thought that the new agency, to fulfill its mandate, needed a measure of independence. So the two political branches, acting together, gave the CFPB Director the same job protection that innumerable other agency heads possess. All in all, those branches must have thought, they had done a good day’s work. Relying on their experience and knowledge of administration, they had built an agency in the way best suited to carry out its functions. They had protected the public from financial chicanery and crisis. They had governed.

And now consider how the dispute ends—with five unelected judges rejecting the result of that democratic process. The outcome today will not shut down the CFPB: A different majority of this Court, including all those who join this opinion, believes that if the agency’s removal provision is unconstitutional, it should be severed. But the majority on constitutionality jettisons a measure Congress and the President viewed as integral to the way the agency should operate. The majority does so even though the Constitution grants to Congress, acting with the President’s approval, the authority to create and shape administrative bodies. And even though those branches, as compared to courts, have far greater understanding of political control mechanisms and agency design.

Nothing in the Constitution requires that outcome; to the contrary. “While the Constitution diffuses power the better to secure liberty, it also contemplates that practice will integrate the dispersed powers into a workable government.” Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579, 635 (1952) (Jackson, J., concurring). The Framers took pains to craft a document that would allow the structures of governance to change, as times and needs change. The Constitution says only a few words about administration. As Chief Justice Marshall wrote: Rather than prescribing “immutable rules,” it enables Congress to choose “the means by which government should, in all future time, execute its powers.” McCulloch, 4 Wheat., at 415. It authorizes Congress to meet new exigencies with new devices. So Article II does not generally prohibit independent agencies. Nor do any supposed structural principles. Nor do any odors wafting from the document. Save for when those agencies impede the President’s performance of his own constitutional duties, the matter is left up to Congress.

Our history has stayed true to the Framers’ vision. Congress has accepted their invitation to experiment with administrative forms—nowhere more so than in the field of financial regulation. And this Court has mostly allowed it to do so. The result is a broad array of independent agencies, no two exactly alike but all with a measure of insulation from the President’s removal power. The Federal Reserve Board; the FTC; the SEC; maybe some you’ve never heard of. As to each, Congress thought that formal job protection for policymaking would produce regulatory outcomes in greater accord with the long-term public interest. Congress may have been right; or it may have been wrong; or maybe it was some of both. No matter—the branches accountable to the people have decided how the people should be governed.

The CFPB should have joined the ranks. Maybe it will still do so, even under today’s opinion: The majority tells Congress that it may “pursu[e] alternative responses” to the identified constitutional defect—“for example, converting the CFPB into a multimember agency.” Ante, at 36. But there was no need to send Congress back to the drawing board. The Constitution does not distinguish between single-director and multimember independent agencies. It instructs Congress, not this Court, to decide on agency design. Because this Court ignores that sensible—indeed, that obvious—division of tasks, I respectfully dissent.

1  In the academic literature, compare, e.g., Kagan, Presidential Administration, 114 Harv. L. Rev. 2245, 2331–2346 (2001) (generally favoring presidential control over agencies), with, e.g., Strauss, Overseer, or “The Decider”? The President in Administrative Law, 75 Geo. Wash. L. Rev. 696, 704, 713–715 (2007) (generally favoring administrative independence).
2  The principle of separation of powers, Madison continued, main-tained only that “where the whole power of one department is exercised by the same hands which possess the whole power of another department, the fundamental principles of a free constitution[ ] are subverted.” The Federalist No. 47, at 325–326.
3  Article II’s Opinions Clause also demonstrates the possibility of limits on the President’s control over the Executive Branch. Under that Clause, the President “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.” §2, cl. 1. For those in the majority’s camp, that Clause presents a puzzle: If the President must always have the direct supervisory control they posit, including by threat of removal, why would he ever need a constitutional warrant to demand agency heads’ opinions? The Clause becomes at least redundant—though really, inexplicable—under the majority’s idea of executive power.
4  The majority dismisses Federalist Nos. 77 and 39 as “reflect[ing] initial impressions later abandoned.” Ante, at 26, and n. 10. But even Hamilton’s and Madison’s later impressions are less helpful to the majority than it suggests. Assuming Hamilton gave up on the Senate’s direct participation in removal (the evidence is sketchy but plausible), there is no evidence to show he accepted the majority’s view. And while Madison opposed the first Congress’s enactment of removal limits (as the majority highlights), he also maintained that the legislature had constitutional power to protect the Comptroller of the Treasury from at-will firing. See infra, at 12–13. In any event, such changing minds and inconstant opinions don’t usually prove the existence of constitutional rules.
5  As President Jefferson explained: “[W]ith the settlement of the accounts at the Treasury I have no right to interfere in the least,” because the Comptroller of the Treasury “is the sole & supreme judge for all claims of money against the US. and would no more receive a direction from me” than would “one of the judges of the supreme court.” Letter from T. Jefferson to B. Latrobe (June 2, 1808), in Thomas Jefferson and the National Capital 429, 431 (S. Padover ed. 1946). A couple of decades later, Attorney General William Wirt reached the same conclusion, stating that “the President has no right to interpose in the settling of accounts” because Congress had “separated” the Comptroller from the President’s authority. 1 Op. Atty. Gen. 636, 637 (1824); 1 Op. Atty. Gen. 678, 680 (1824). And indeed, Wirt believed that Congress could restrict the President’s authority to remove such officials, at least so long as it “express[ed] that intention clearly.” 1 Op. Atty. Gen. 212, 213 (1818).
6  The Comptroller legislation of the Civil War provided a key precedent for what does appear a historical “aberration”—the Tenure of Office Act of 1867. See ch. 154, 14Stat. 430. Anxious to prevent President Andrew Johnson from interfering with reconstruction policies—including through his command of the military—Congress barred presidential removal of any Senate-confirmed officials without the Senate’s consent. The law thus severed the President’s removal authority over even officials like the Secretaries of War and State. The statute became the basis for the Nation’s first presidential impeachment, but was repealed in 1887. See Act of Mar. 3, 1887, ch. 353, 24Stat. 500. In one sense, the two-decade-long existence of the Tenure of Office Act reveals the 19th-century political system’s comfort with expansive restrictions on presidential removal. But the ultimate repudiation of the law, and the broad historical consensus that it went too far, just as strongly shows the limits that system later accepted on legislative power—that Congress may not impose removal restrictions preventing the President from carrying out his own constitutionally assigned functions in areas like war or foreign affairs. See Morrison v. Olson, 487 U.S. 654, 689–691 (1988) (recognizing that limit as the constitutional standard).
7  The majority is quite right that today we view all the activities of administrative agencies as exercises of “the ‘executive Power.’ ” Arlington v. FCC, 569 U.S. 290, 305, n. 4 (2013) (quoting Art. II, §1, cl.1); see ante, at 14, n. 2. But we well understand, just as the Humphrey’s Court did, that those activities may “take ‘legislative’ and ‘judicial’ forms.” Arlington, 569 U. S., at 305, n. 4. The classic examples are agency rulemakings and adjudications, endemic in agencies like the FTC and CFPB. In any event, the Court would soon make clear that Congress can also constrain the President’s removal authority over officials performing even the most “executive” of functions. See infra, at 19–20.
8  Expressing veiled contempt as only he could, Justice Frankfurter wrote for the Court that Chief Justice Taft’s opinion had “laboriously traversed” American history and that it had failed to “restrict itself to the immediate issue before it.” 357 U. S., at 351. No wonder Humphrey’s had “narrowly confined the scope of the Myers decision.” 357 U. S., at 352. Justice Frankfurter implied that the “Chief Justice who himself had been President” was lucky his handiwork had not been altogether reversed. Id., at 351.
9 Pretending this analysis is mine rather than Morrison’s, the majority registers its disagreement. See ante, at 27–28, n. 11. In its view, a test asking whether a for-cause provision impedes the President’s ability to carry out his constitutional functions has “no real limiting principle.” Ibid. If the provision leaves the President with constitutionally sufficient control over some subordinates (like the independent counsel), the majority asks, why not over even his close military or diplomatic advisers? See ibid. But the Constitution itself supplies the answer. If the only presidential duty at issue is the one to ensure faithful execution of the laws, a for-cause provision does not stand in the way: As Morrison recognized, it preserves authority in the President to ensure (just as the Take Care Clause requires) that an official is abiding by law. See 487 U. S., at 692. But now suppose an additional constitutional duty is implicated—relating, say, to the conduct of foreign affairs or war. To carry out those duties, the President needs advisers who will (beyond complying with law) help him devise and implement policy. And that means he needs the capacity to fire such advisers for disagreeing with his policy calls.
10  The majority suggests that the FTC was a different animal when this Court upheld its independent status in Humphrey’s. See ante, at 17. But then, as now, the FTC’s organic statute broadly “empowered and directed” the agency “to prevent persons” or businesses “from using unfair methods of competition in commerce.” Act of Sept. 26, 1914, §5, 38Stat. 719. To fulfill that mandate, the agency could and did run investigations, bring administrative charges, and conduct adjudications. See ibid.; §6(a), id., at 721; FTC Ann. Rep. (1935) (describing the FTC’s extensive enforcement activities in the year before Humphrey’s). And if any person refused to comply with an order, the agency could seek its enforcement in federal court under a highly deferential standard. See §5, 38Stat. 720; FTC v. Pacific States Paper Trade Assn., 273 U.S. 52, 63 (1927). Still more, the FTC has always had statutory rulemaking authority, even though (like several other agencies) it relied on adjudications until the 1960s. See §6(g), 38Stat. 722; National Petroleum Refiners Assn. v. FTC, 482 F.2d 672, 686 (CADC 1973). (The majority’s reply that a court including Charles Evans Hughes, Louis Brandeis, Benjamin Cardozo, and Harlan Stone somehow misunderstood these powers, see ante, at 17, n. 4, lacks all plausibility.) And in any case, the relevant point of comparison is the present-day FTC, which remains independent even if it now has some expanded powers—and which remains constitutional under not only Humphrey’s but also Morrison. See supra, at 18–20.
11  The majority briefly mentions, but understandably does not rely on, two other features of Congress’s scheme. First, the majority notes that the CFPB receives its funding outside the normal appropriations process. See ante, at 24–25. But so too do other financial regulators, including the Federal Reserve Board and the FDIC. See 12 U. S. C. §§243, 1815(d), 1820(e). And budgetary independence comes mostly at the expense of Congress’s control over the agency, not the President’s. (Because that is so, it actually works to the President’s advantage.) Second, the majority complains that the Director’s five-year term may prevent a President from “shap[ing the agency’s] leadership” through appointments. Ante, at 24. But again that is true, to one degree or another, of quite a few longstanding independent agencies, including the Federal Reserve, the FTC, the Merit Systems Protection Board, and the Postal Service Board of Governors. See, e.g., §§241, 242; 15 U. S. C. §41; 5 U. S. C. §§1201, 1202; 39 U. S. C. §202. (If you think the last is unimportant, just ask the current President whether he agrees.)
12  The majority, seeking some other way to distinguish Morrison, asserts that the independent counsel’s “duties” were more “limited” than the CFPB Director’s. Ante, at 17–18. That’s true in a sense: All (all?) the special counsel had to do was decide whether the President and his top advisers had broken the law. But I doubt (and I suspect Presidents would too) whether the need to control those duties was any less “central to the functioning of the Executive Branch” than the need to control the CFPB’s. Morrison, 487 U. S., at 691–692. And in any event, as I’ve shown, Morrison did much more than approve a specific removal provision; it created a standard to govern all removal cases that is at complete odds with the majority’s reasoning. See supra, at 19–21.
13  To use one important example, Congress provided for executive oversight of all the CFPB’s rulemaking. The Financial Stability Oversight Council can veto by a two-thirds vote any CFPB regulation it deems a threat to the “safety and soundness” of the financial system. 12 U. S. C. §5513(a). The FSOC is chaired by the Treasury Secretary, and most of its members are under the direct supervision of the President. See §5321. So the majority is wrong in saying that the CFPB’s Director can “unilaterally” issue final regulations. Ante, at 23 (emphasis in original). Indeed, the President has more control over rulemaking at the CFPB than at any similar independent agency. And the majority is similarly wrong to think that because the FSOC has not yet issued a formal veto, its review authority makes no practical difference. See ante, at 25, n. 9. Regulatory review, whether by the Office of Management and Budget or the FSOC, usually relies more on the threat of vetoes than on their execution. OMB casts a long shadow over rulemaking in the Executive Branch, but rarely uses its veto pen. See Sunstein, The Office of Information and Regulatory Affairs: Myths and Realities, 126 Harv. L. Rev. 1838, 1846–1847, n. 37 (2013).
14  I could go on. A recent study prepared for the Administrative Conference of the United States noted that “[g]overnance by multiple members limits the President’s influence.” J. Selin & D. Lewis, Sourcebook of United States Executive Agencies 89 (2d ed. 2018). And the General Accounting Office has recognized that the desire for “greater independence” is what “most likely explains why the Congress in the past has opted to head independent regulatory bodies with multimember commissions rather than single administrators.” Hearing before the Senate Subcommittee on the Consumer of the Committee on Commerce, Science, and Transportation, 100th Cong., 1st Sess., 135 (1987) (Statement of F. Frazier).