Jack M. Beermann1Jack M. Beermann is Professor of Law and Harry Elwood Warren Scholar at Boston University School of Law and a 1983 graduate of the University of Chicago Law School.
* * *
A part of the Seila Law and the Roberts Court series.
In Seila Law LLC v. Consumer Financial Protection Bureau, the Supreme Court, in an opinion by Chief Justice John Roberts, invalidated the provision of the Dodd-Frank Act restricting the president’s removal of the director of the Consumer Financial Protection Bureau (CFPB) to cases of “inefficiency, neglect of duty, or malfeasance in office.” The Court’s decision leaves the director subject to removal by the president for any reason or no reason at all.
This Essay on the Seila Law decision makes three points. First, because there is no legal or historical support for the Court’s distinction between independent agencies headed by multiple members and independent agencies headed by single officials, Seila Law creates a novel constitutional prohibition: Congress may not create an independent agency with significant regulatory power headed by a single director. Second, when rejecting the argument that it could interpret “inefficiency, neglect of duty or malfeasance in office” to include the director’s refusal to execute lawful presidential commands, the Court declined an opportunity to take a major step toward presidential control of independent agencies. Third, the dissent’s argument that multimember independent agencies may actually be more difficult for the president to control than independent agencies headed by single officials provides the Court with a possible basis to invalidate the entire independent agency form. Whether this will actually happen, only time will tell.
I. The Legal and Historical Background
There is no legal or historical basis for the Court’s distinction between single-headed independent agencies and multiheaded independent agencies. The Court’s primary argument for such a distinction in Seila Law is simple: Article II vests the executive power in the president and requires that they “take Care that the Laws be faithfully executed.” That power has long been construed to require that the president have some measure of control over all of the officials in the executive branch, but it has also been understood, since at least 1935, to allow Congress to impose restrictions on the president’s power to remove principal officers appointed to positions created by Congress. In the Seila Law Court’s view, however, Congress’s power to restrict removal does not extend to single-headed agencies because that would be too great of a restriction on the president’s ability to fulfill their constitutional role. This is an unprecedented conclusion.
No Supreme Court decision has ever before relied upon the multiheaded nature of an independent agency as a reason for upholding removal restrictions or opined that similar restrictions on the removal of the director of a single-headed agency would be unconstitutional. In cases like Humphrey’s Executor v. United States, in which the Court upheld an identical restriction on the president’s power to remove members of the Federal Trade Commission (FTC), the Court mentioned multiple agency heads only to describe the agency. In fact, the Court’s statement of its conclusion in Seila Law reveals the absurdity of its claim that it is merely following precedent under which the president’s removal authority is absolute except in cases involving multiheaded agencies: “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.” Given that the Constitution’s text is silent on the non-impeachment removal of executive branch officials, and, as the Court concedes, the Framers held conflicting views on routine removal, characterizing the Constitution’s removal rules as “a carefully calibrated system” is amusing but not particularly enlightening or persuasive.
The chief justice’s opinion distinguishes Humphrey’s Executor on two grounds: first, as noted, that the Federal Trade Commission is a multimember body, and second, that the Court in that case relied heavily on the characterization of the FTC’s duties as “quasi-judicial” and “quasi-legislative.” The Court’s characterization is accurate, but neither factor provides support for the distinction. On the first point, the Court in Humphrey’s Executor did not utter a single word that remotely supports the notion that the plural nature of the Commission’s membership was important to the decision. On the second point, the Court ignores the near-universal acceptance in recent decades, which it notes in the opinion itself, of the understanding that everything an agency does is law-execution even when agency action takes a legislative or judicial form. I do not mean to say that the Seila Law Court’s machinations in twisting its removal precedents are beyond the pale. If judges are experts at anything, it is at converting specious distinctions into legal principles while claiming, with a straight face, that they are simply following the law.
One answer to my claim that the Court’s reliance on the single-headed nature of the CFPB is unprecedented might be that the Seila Law Court engaged in traditional legal reasoning based on plenty of precedent, both legal and historical, and arrived at its decision by weighing competing legal principles including presidential control over execution of the law and Congress’s authority to create and structure federal agencies under the Necessary and Proper Clause in light of the entire corpus of its separation-of-powers jurisprudence. This may be an accurate description, but I assert that it is an inappropriate methodology in separation-of-powers cases, where the Court has traditionally deferred to Congress’s judgment except when Congress transgresses a clearly stated procedural or structural provision of the Constitution. There are good reasons for the traditional reluctance of the Supreme Court to question the wisdom of Congress’s structural decisions including that (1) the Constitution’s detailed rules on many issues such as appointment of Officers of the United States and the legislative process imply a lack of binding law on other matters; (2) the Necessary and Proper Clause expressly delegates to Congress authority over matters concerning the power of other branches not covered in the Constitution itself; (3) the judiciary lacks comparative advantage in evaluating the optimal structure of government; and (4) Congress is clearly more accountable than the Court, allowing for easier correction of mistakes concerning agency structure. Although I might agree with the Court that multiheaded agencies are far superior to single-headed ones, deciding the case on that basis is not consistent with the Court’s self-professed methodology in constitutional cases.
The only real historical support the Court offers for its distinction between single-headed and multiheaded independent agencies is the fact that Congress has rarely, and only recently, established independent agencies headed by a single director. This fact should be of no importance to determining whether Congress has the power to do so now. It is true that the Court often looks at history for guidance. Consistent long-running practices can provide evidence of general acceptance of the constitutionality of the practice, especially actions of early Congresses. But the lack of past practice is exceedingly weak evidence of unconstitutionality unless, perhaps, the practice had been proposed and rejected by Congress or by presidential veto on constitutional grounds. Otherwise, the argument amounts to a rejection of innovation with no constitutional grounding.
There are many non-constitutional reasons why Congress may not have previously created single-headed independent agencies. Perhaps they agreed with the Seila Law majority that multiheaded agencies would make better decisions or that decisions from multimember commissions would be more politically acceptable. Perhaps they thought that there would be too much work for a single director to handle. Perhaps they wanted more opportunities for placing their aides in powerful federal positions. Perhaps they just never thought of it because early independent agencies were multiheaded due to their adjudicatory responsibilities. None of these reasons for following the multimember form reflects on the constitutionality of the single-headed form. In effect, by finding the lack of prior congressional practice important to its decision, the Court endows earlier electorates with the power, by inaction, to limit the decisions of future electorates who have chosen a new, perhaps better, perhaps inferior, path. When Chief Justice John Marshall sounded his famous reminder that “it is a Constitution we are expounding,” he was warning his colleagues and future generations not to read the document to unduly constrain innovation and adaptation. The lack of single-headed independent agencies before recent decades should have been irrelevant to the Court’s decision in Seila Law.
The Court’s most powerful practical argument is that the lack of accountability of the director of the CFPB is more striking than that of other independent agencies because of its singularity, its sweeping powers, and because by law it receives its funding from the Federal Reserve Board without congressional input or control. Even if these doubtful propositions are true, the Court ignores the simple reality that all it would take to change any of these features of the CFPB would be an act of Congress. If Congress, the most democratically accountable branch of the federal government, determines that the CFPB has gone off the rails, it could restructure the agency or limit the agency’s funding, restrict its expenditures, or even alter the agency’s substantive authority. And the Court also ignores the availability of judicial review of agency action, which is broadly understood to impose a great deal of accountability across the administrative state. The CFPB’s single director cannot exercise powers or discretion beyond that granted by Congress. The federal courts will not allow the CFPB to stray beyond Congress’s delegation of authority or to make decisions not supported by rationally evaluated, permissible policy objectives.
II. The Meaning of “Inefficiency, Neglect of Duty, or Malfeasance in Office”
The Court could have made a major reform to the administrative state by reinterpreting “inefficiency, neglect of duty, or malfeasance in office” to allow the president to remove any official who refuses to follow a lawful presidential order. In effect, such a construction would put an end to the most important aspect of the independent agency form, a result long sought by advocates of the unitary executive theory of separation of powers.
As noted, the problem with the CFPB identified by Seila Law, LLC was that its single director could be removed by the president only for “inefficiency, neglect of duty, or malfeasance in office.” This standard, Seila Law argued, deprived the president of sufficient control over the director’s conduct to meet the requirements of Article II’s vesting of the executive power in the president. The Trump administration agreed, and thus it was left to a Court-appointed amicus curiae to argue in favor of the constitutionality of the statute. In that brief, Paul Clement argued that “the terms ‘inefficiency, neglect of duty, or malfeasance in office’ can be interpreted to impose only a permissible degree of restraint.” While Clement did not offer an alternative definition of the standard, he pointed out that in Bowsher v. Synar, the Court had construed identical language as making the comptroller general “subservient” to the removing authority, which in that case was Congress. Clement also conceded, in unfortunately sexist language, that the standard required the president “to have some reason, beyond simply wanting his ‘own man’ to remove an officer.”
The Court rejected Clement’s suggestion to construe the removal clause to give the president sufficient control for three fairly weak reasons: first, that the Court had declined a similar invitation in its 1935 decision in Humphrey’s Executor; second, that Clement’s reading “fails to engage with the Dodd-Frank Act as a whole, which makes plain that the CFPB is an ‘independent bureau’”; and third, because neither Clement nor any other amicus offered a plausible alternative construction of the clause.
The Court’s reliance on Humphrey’s Executor for its refusal to reinterpret the removal clause borders on laughable in light of the opinion’s overall treatment of Humphrey’s Executor as a wayward decision. Space does not allow a complete explication of all the weaknesses of each of these grounds for rejecting a limiting construction, but the second basis merits special attention. If independence of the CFPB is so important to the Dodd-Frank Act as a whole, it makes no sense for the Court to eliminate the most important aspect of that independence, namely the director’s protection from removal without good cause. If Chief Justice Roberts truly believed that independence was so central to Congress’s vision for the CFPB, he should have voted with Justices Clarence Thomas and Neil Gorsuch to find the removal provision inseverable from the remainder of the statute and halted the operations of the CFPB completely.
I was surprised that the Court did not engage with the possibility that refusal to follow lawful presidential policy directives constitutes “inefficiency” or “neglect of duty,” allowing the president to remove noncomplying agency heads. (“Lawful” because it has been long understood that the president does not have the constitutional power to order a subordinate to violate the law.) It would certainly be within standard English usage to construe “inefficiency, neglect of duty, or malfeasance in office” to include the director’s failure to follow the lawful orders of the president. Agents are expected to obey their principals’ instructions, and it would not be surprising if a principal dismissed an agent for failing to do so. A response to this suggestion might be that Congress intended for the CFPB to be independent of the president, either to satisfy the romantic notion that independent agencies are outside of politics or the more realistic understanding that Congress wants to maintain as much influence over the agency as possible. But we know from numerous cases, including the very recent decision that Title VII prohibits discrimination based on sexual orientation and gender identity, that the Supreme Court is more than willing to depart from Congress’s intent when a statute’s language is capacious enough to accommodate its desired result in even the most controversial of cases.
This reading would preserve at least a modicum of the director’s independence by disallowing discharge simply so the president can install his own person or a person of his own political party, both of which are now legally proper under the Court’s decision to allow the president to remove the director without cause. But more importantly, it recognizes that an agency official’s refusal to obey a president’s policy-based directive stands as a direct barrier to the fulfillment of the president’s primary constitutional duty to “take Care that the Laws be faithfully executed.” A subordinate’s refusal to obey a superior’s lawful instruction certainly falls within the ordinary meaning of the words “neglect of duty.” Given the Court’s attention to the president’s constitutional power over the executive branch, the Court should have at least offered a better explanation for rejecting that reading.
It seems highly likely that the chief justice and his conservative colleagues realized that construing the statutory removal restriction to allow discharge for failure to follow policy-based directives would significantly advance the cause of the unitary executive theory. Thus, I am constrained to conclude that even they are not willing to go that far to completely outlaw independent agencies. This conclusion is consistent with the chief justice’s apparent endorsement of the independence of multimember independent agencies in Seila Law and in Free Enterprise Fund v. Public Accounting Oversight Board, in which he endorsed reading a for-cause restriction into a statute that was silent on the matter, apparently accepting the independence of multimember independent agencies. There’s no legal basis for any of this, just a possible sense among some of the justices, at least Chief Justice Roberts, that regardless of their constitutional vision, they should not completely wrench control over such an important aspect of the structure of the administrative state away from Congress. Perhaps this is based on a principle of judicial restraint or perhaps it grows out of a fear, which in recent years has grown more salient, that governance by presidential fiat might not be best for the country.
In any case, it is notable that the Court did not even explore an alternative that held promise to inflict serious damage to Congress’s preferred structure of the administrative state. Construing removal restrictions such as those contained in the Dodd-Frank Act to allow for removal based on policy disagreement would go far toward placing all independent agencies under a previously unknown degree of presidential influence. It would effectively resolve the long-running dispute over the propriety of the independence of independent agencies in favor of the president. Given the Court’s professed attention to precedent, the decision against construing the removal provision at issue in Seila Law to allow more presidential control makes that alternative less likely, at least in the near future.
III. The Accountability Puzzle: Single versus Plural
The most questionable aspect of the Seila Law opinion is its conclusion that a single director is less accountable to the president than the multiple heads of most independent agencies. Although it makes sense when you say it fast, the Court’s distinction between single-headed and multiheaded independent agencies does not hold up to sober scrutiny. The dissent, picking up on arguments made rather gently in Paul Clement’s brief, argues persuasively—perhaps too persuasively—that multiheaded agencies are more difficult for the president to control than single-headed agencies. Because most agency members are appointed to terms that are longer than the president’s, a new president may have to wait years, perhaps more than a full first term, to gain control over a multiheaded agency. By contrast, if a single-headed directorship becomes open, the president can gain control more quickly with one appointment. Clement pointed out that the requirement of partisan balance that applies to virtually every multiheaded independent agency guarantees that the president will always face opposition to policy change from within the agency. Further, some agencies may operate by consensus, and the views of agency members of the president’s party may not be monolithic, meaning that the opposition party’s representatives will have influence even after they become a minority on the agency. Further, it may often be easier for the president to deploy the presidential bully pulpit against a single director than against an entrenched group. There is strength in numbers.
In truth, the majority’s most persuasive arguments favoring multimember agencies over single-headed agencies have nothing to do with separation of powers or presidential control. Rather, the majority’s most persuasive point boils down to a simple understanding that group decision-making is likely to be superior to decision-making by a single director because the necessity of persuading others reduces the possibility of harmful mistakes. This is certainly plausible. Bad ideas are less likely to get through a multimember decision-making process. In political decision-making, this concern also sounds in democratic and pluralistic values: a multimember agency is likely to represent more points of view and more constituencies than a single headed agency. Both of these are very similar to arguments traditionally made for bicameralism in legislative bodies and for presentment of legislation to a chief executive for signature or veto. But neither has any implication for presidential control except perhaps that the president who appoints the head of a single-headed agency is likely to have more influence over that body than over a multiheaded agency.
The Court’s analysis seems to be giving a nod toward a relatively new school of thought in administrative law that has been called “internal separation of powers.” Advocates of internal separation of powers analogize to separation of powers in the Constitution, arguing that concentration of executive, legislative, and judicial power in the same entity poses a danger to liberty. Avoiding that danger by creating three distinct branches of government is the genius of the Constitution and was the primary argument made to convince skeptics to accept increased central government power. Similar logic applies to agencies: fragmenting decision-making power within agencies that are statutorily authorized to employ executive, legislative, and judicial forms decreases the danger of arbitrary exercises of coercive governmental power. The advocates of internal separation of powers may be on to something in terms of optimal agency structure. But, putting due process in agency adjudication aside, the internal structure of an agency is not constitutionally relevant except under a purely pragmatic view of the Constitution. That is why, as the Seila Law Court recognizes, the law rightly regards all exercises of agency authority as executive in nature even if they take a legislative or judicial form. Whenever an agency acts, it is executing the law, regardless of the form its action takes.
In dissent, Justice Kagan argued, somewhat persuasively, that multiheaded agencies are more difficult for the president to control than single-headed agencies. She provided several reasons for her conclusion: (1) Experience shows it is easier to get one person do what you want than a gaggle; (2) it is easier to assign blame when an action is taken by a single individual than by a group; (3) multimember agencies were created by Congress to reduce presidential control; and (4) the president’s power to fire the CFPB’s director, even as currently constrained, is more likely to be used than the power to fire multiple agency heads. (This last argument is how I interpret Justice Kagan’s somewhat vague discussion of the president’s power to discharge agency heads for cause.)
This is the point at which the opinion in Seila Law packs the greatest potential to upend the administrative state as we know it, by eliminating the independence of independent agencies regardless of the number of directors or commissioners. The Court’s conservative majority could rely on Justice Kagan’s arguments to strike down removal protections for all principal officers, including the heads of multimember agencies. That is the clear implication of the majority’s expressed concern over presidential control of the entire executive branch. The Court makes no secret in Seila Law that it views Myers v. United States as authoritative on presidential control over agency officials even though, as the Court recognized in Humphrey’s Executor and Morrison v. Olson, virtually everything Myers said on the subject was dicta of the worst kind: a judge reaching out to address important unsettled legal matters not presented by the case at bar. The Myers dicta would allow removal constraints on inferior officers but not principal officers. If you believe that the Constitution establishes a unitary executive and that Congress cannot insulate the execution of the law from presidential control to any significant extent, the next natural step after Seila Law would be to abandon the distinction between single-headed and multiheaded agencies and strike down removal restrictions that apply to any agency head across the entire government.
If the Court ultimately applies the rule in Seila Law to all independent agency heads, it would not be the first time that the Court would have disavowed a distinction that it earlier claimed was an important limitation on a new rule of constitutional law. For example, after it struck down state and local minority-contractor set-aside programs under strict equal protection scrutiny, it stated that it would continue to apply intermediate scrutiny to federal programs of a similar nature. Later, it abandoned the distinction and held that all programs, state and federal, were subject to strict scrutiny under equal protection principles. Restrictions on presidential removal of independent agency heads could follow a similar trajectory.
This is where the potential for real disaster lies. A president who is willing to push an extreme, ill-considered, unwise agenda, fueled by political support from a narrow sector of society and egged on by an overly obsequious cadre of like-minded supporters in the White House and in Congress, could do real harm to the country. This nightmare scenario is multiplied if such a president were to achieve a second term, during which the concern over re-election is gone. For all of their faults, independent agencies are a bulwark against this possibility. Even if they are subject to congressional influence, the plurality of Congress and its constantly evolving membership and priorities present much less of a danger of extremism and harm to the country than a president off on a unilateral adventure.
The Court’s portrayal of Humphrey’s Executor as creating a narrow exception to a general rule that the president has constitutional power to remove all executive branch officials without cause, is the strongest indication that the Court might be willing to reconsider the whole mess. Judge-made exceptions to general constitutional rules would seem to be more vulnerable to judicial changes of heart than the rules themselves. The problem is that there is not and never has been a general rule granting the president the constitutional power to remove anyone charged with executing the law. As noted, the Constitution is silent on the matter, and the Constitution’s structure is deliciously ambiguous on the extent of Congress’s power to structure the executive branch. The only thing we know for sure is that Congress may not prohibit principal officers from providing written answers to the president’s questions. The importance of the vesting of executive power in a single president must be understood in light of both Congress’s power to make all laws that are “necessary and proper” for carrying into effect its constitutional powers and the all of the powers granted to the national government and the Constitution’s imposition of responsibility on the president to “take Care that the Laws are faithfully executed.” With regard to execution of the law, this language implies that the president is Congress’s agent.
Despite all of this, the most likely possibility is that the Court will not reject multimember independent agencies, at least as long as Chief Justice Roberts remains the swing vote on the Court. We have been at this precipice numerous times since the 1970s and 1980s when a series of Supreme Court decisions that seemed to portend major change fizzled into moderate reform to agency structure. Further, the Court’s removal decisions, including Seila Law, seem to enshrine the multimember independent agency into the constitutional firmament. Therefore, it may be that Seila Law, along with Free Enterprise Fund’s rule against double layered for-cause restrictions on the removal of members of multiheaded independent, will remain minor limitations on Congress’s power to create independent agencies––similar to the minor limits on Congress’s power to regulate interstate commerce that the Rehnquist Court created, namely limitation of legislation based on effects on interstate commerce to economic matters and the prohibition on federal “commandeering” of state and local officials to execute federal law. In fact, the prohibitions on two levels of protection from removal and single-headed independent agencies are analogous to the anti-commandeering doctrine. Just as Justice Sandra Day O’Connor created the anti-commandeering doctrine out of whole cloth in her dissenting opinion in FERC v. Mississippi, Justice Brett Kavanaugh created the prohibition on single-headed independent agencies in his panel opinion in the PHH Case. Commandeering of state and local officials is uncommon, just as single-headed independent agencies are exceedingly rare. If the Court remains true to form, Seila Law will be a footnote to a story of substantial congressional power to insulate agency heads from presidential control.
What would be gained if the Supreme Court were to effectively prohibit independent agencies? In terms of fidelity to the Constitution, nothing. Rather, the Court would be negating Congress’s power to “make all laws which shall be necessary and proper” to effectuate federal powers and would be substituting the Court’s judgment for Congress’s on matters of important policy and political salience. In terms of the efficient operation of the government, again, nothing. Further, nothing in the Court’s training or historical performance points toward it knowing more about what’s good for the country than Congress, even a Congress as widely reviled as the current one. Historically, the Court has made colossal policy mistakes on, for example, economic regulation, civil rights, and gun control. For all of Congress’s faults, and all of the ugly politicking that infects it, unless a statute violates a clearly stated structural provision of the Constitution, the Supreme Court should defer to Congress’s judgments about the appropriate structure of the federal government, including the degree to which agency heads should be shielded from direct presidential control. Anyone who thinks that the problem with the U.S. government is that the president lacks sufficient control over the Executive Branch simply hasn’t been paying attention.
* * *
Jack M. Beermann is Professor of Law and Harry Elwood Warren Scholar at Boston University School of Law and a 1983 graduate of the University of Chicago Law School.
* * *