Arthrex and Collins: The Roberts Court and Presidential Authority
Lerner Family Associate Dean for Public Interest and Public Service Law; Professorial Lecturer in Law, George Washington University Law School
Alan B. Morrison[*]
The U.S. Supreme Court decided two cases this term, United States v. Arthrex, Inc. and Collins v. Yellen, in which the Court followed the prior path set by Chief Justice John Roberts in which he enshrined into the Constitution his vision of the role of the president. In Arthrex, the Court concluded that administrative patent judges, who are inferior officers, cannot make final decisions in inter partes proceedings, but those rulings must come from an officer who is appointed by the president, confirmed by the Senate, and subject to removal by the president. In Collins, which followed last year’s ruling in Seila Law, LLC v. Consumer Financial Protection Bureau, the Court, in an opinion by Justice Samuel Alito, held that the single head of the Federal Housing Finance Administration (FHFA) must also be subject to removal by the president at will. These rulings, issued under the banner of assuring accountability of agency officials and the president, may promise more than they can deliver in the real world. In addition, their applications to the wide variety of agency arrangements and duties are very much up in the air. The only certainty is that there will be much more litigation and many calls for congressional action, most of which are likely to go unanswered.
I. The Removal Limitations Cases
The constitutional provision underlying these cases is the Appointments Clause in Article II, section 2, which was the portion of the Constitution on which William Marbury relied in seeking to obtain his commission as a justice of the peace in Marbury v. Madison. Although its words have not been the focus of these recent decisions, they are worth quoting in full, largely because of what is not said there:
[The president] shall nominate, and by and with the Advice and Consent of the Senate, shall appoint Ambassadors, other public Ministers and Consuls, Judges of the supreme Court, and all other Officers of the United States, whose Appointments are not herein otherwise provided for, and which shall be established by Law: but the Congress may by Law vest the Appointment of such inferior Officers, as they think proper, in the President alone, in the Courts of Law, or in the Heads of Departments.
Notice that there is no mention of removal in this clause, and its only appearance in the Constitution is in the Impeachment Clause in Article I, section 3, clause 7. Yet, both Seila Law and Collins held that the president had the power, under the Appointments Clause, to remove the officers at issue at will, despite the statutes which limited their removal except for cause.
These cases were not the first ones in which the Court ruled on the extent of the power of Congress to limit the ability of the president to remove his appointees at will. The common law rule for all employees was that they could be discharged at will; thus, unless Congress imposed restrictions on the president’s removal authority, the presumption was that he could dismiss his appointees whenever he chose. In Myers v. United States, Chief Justice William Howard Taft, himself a former president, ruled that the statute that prevented the president from removing First Class Postmaster Myers, unless he first obtained the consent of the Senate, was unconstitutional. The exhaustive opinion had much in it that would support the conclusion that any limitations on the president’s removal power would be held unconstitutional. But less than a decade later, the Court decided Humphrey’s Executor v United States, which upheld a statute that limited the ability of the president to remove members of the Federal Trade Commission except for certain specified causes.
The next time that the Court faced the question of whether the Constitution permitted restrictions on the power of the president to remove federal officers was in Bowsher v Synar, a challenge to the role that the comptroller general (CG), as head of the General Accounting Office (as it was then known), was assigned to implement the Gramm-Rudman-Hollings Act. The CG could only be removed if both Houses of Congress concurred, which meant that the president had essentially no ability to remove him. The law also limited the president’s choices to fill a vacancy in the office to a list of individuals proposed by Congress, and in addition, as the concurring opinion of Justice John Paul Stevens pointed out, the CG had been viewed by everyone as working for Congress rather than the president. Thus, it was not necessary to revisit Humphrey’s Executor in order to conclude that, as a matter of separation of powers, the CG could not serve as an executive-branch official who was executing the law.
Of course, Myers was not really (and certainly not only) about postmasters, because Congress had imposed similar Senate-consent requirements for removal of many high-ranking officials, and it would be intolerable if the Senate could block the president from deciding that he no longer wanted to keep his secretary of state or attorney general. But that was not the stated basis of the actual ruling in Myers, and so the Court in Humphrey’s Executor had to find another rationale to uphold the limitation, which it did by reaching the dubious conclusion that a Federal Trade Commission (FTC) commissioner is someone “who occupies no place in the executive department and who exercises no part of the executive power vested by the Constitution in the President.” The uneasy tension between at least the rationales of the two cases, if not their holdings, was at issue in Morrison v. Olson, where the Court upheld the Independent Counsel Act that limited the ability of the attorney general (and the president) to remove an independent counsel except for cause. In his opinion for the Court, Chief Justice William Rehnquist expressed his view that Myers was about Senate interference with executive functions and that Humphrey’s statements regarding whether the FTC was part of the Executive Branch may not have been entirely accurate. In any event, the Court’s bottom line was that excluding the president from choosing independent counsels (a special court did that) and from removing them (only the attorney general could do and even then only for cause) did not interfere with constitutional principles of accountability and separation of powers.
The next case in which limits on removal of a federal officer were challenged, and the first under Chief Justice Roberts, was Free Enterprise Fund v. Public Company Accounting Oversight Board. Enacted in the wake of the Enron and other accounting scandals, the law created a board, under the Securities and Exchange Commission (SEC), that had the authority to issue accounting rules and bring enforcement actions against those responsible for corporate financial statements that violated those rules. The Board was appointed by the SEC, and its members could only be removed for cause. The Board’s decisions were subject to review by the SEC and then in the courts of appeals. The members of the SEC were appointed by the president and confirmed by the Senate and, like the members of the FTC, could also only be removed for cause.
The Free Enterprise lawsuit was a broad challenge to the Board’s powers, with only one portion focused on the limits on removal, but that is where the chief justice found a constitutional flaw. Relying on the principle of presidential accountability, the Court concluded that double for-cause removal restrictions violated that principle—whose source in the words of the constitution was unstated—and so the Court struck the for-cause removal limitation as applied to the board.
The next case, Lucia v. SEC, was more important for what it did not decide—whether the limit to for-cause removals for administrative law judges (ALJs) at the SEC unconstitutionally interfered with the president’s ability to carry out the duties of his office. The principal issue in the case was whether the ALJs were inferior officers, who had to be appointed by the SEC itself, or were employees, who could be (and were) selected by the Commission staff. The Court held that the ALJs were inferior officers, and since they had not been properly appointed, it required the SEC to give Lucia a new hearing before a properly appointed ALJ. The solicitor general had also asked the Court to strike down the removal restrictions, but the Court declined to reach this second issue.
Then came Seila Law, where the Court held that, as applied to the Consumer Financial Protection Bureau, which is headed by a single director, the for-cause restrictions interfere with the ability of the president “to take care that the laws be faithfully executed” under Article II and undermined his accountability to the people who have elected him. The majority declined to “revisit” Humphrey’s Executor and limited its removal holdings to agencies headed by a single director, whereas Justices Clarence Thomas and Neil Gorsuch would have extended the ruling to multi-member agencies such as the FTC.
This past term, the Court in Collins was faced with similar restrictions on another single member agency—the Federal Housing Finance Administration (FHFA)—and it found Seila Law to be “all but dispositive” but provided no independent rationale for setting aside the removal limitation there. That was not the end of Collins. That lawsuit was brought by shareholders of Fannie Mae and Freddie Mac to set aside an administrative order by the FHFA that had vastly reduced the value of their holdings. Their first step, on which they succeeded, was to obtain a ruling, similar to that in Seila Law, that undermined the power of the head of the FHFA to issue the order in question because he could not be removed at will by the president. But there was a further barrier: Although the initial order was said to be invalid, it was subsequently affirmed by a successor who held the position on an acting basis, which meant that he could be removed at will by the president. According to the Court, that ratification by an acting director eliminated the unconstitutional cloud on the director’s status, and therefore the provisions to which the shareholders objected were valid. I return to this conclusion below.
II. What About Constitutional Avoidance in Difficult Constitutional Cases?
Collins and Seila Law decided a constitutional issue of great importance, but could and should the Court have avoided deciding the question in both cases? There are two reasons to support such a conclusion. First, on what basis did the plaintiffs have standing to claim that the inability of the president to fire the agency head injured them and gave them standing to raise the constitutional questions?
Standing was not an issue in Myers or Humphrey’s Executor because the president had fired both of the plaintiffs, and he claimed that the restrictions on those actions were unconstitutional. Standing to raise the removal issue was not challenged in Bowsher, but if it had been, the removal objection was only part of a broader challenge to the authority of the comptroller general who was a congressional officer assigned to carry out executive functions, a clear violation of the separation of powers that went to the validity of the order being challenged. Similarly, in Morrison, the removal objection was closely tied to the appointment problems, and the president had entered the case to oppose all aspects of the law, which explains, if it does not justify, why no one raised the standing issue regarding the limits on removal. And in Free Enterprise, the standing issue does not appear to have been raised.
However, standing was raised by several amici in Seila Law, and the Court treated the removal claim as an objection to the power of the agency to issue the order in question, much like a flaw in the director’s appointment. But there was no question that the director of the CFPB had been validly appointed—unlike Lucia and Morrison, where that was the basis of the legal challenge. There is also no doubt that had the president fired either official, the president would have had standing to raise the defense that the restrictions on removal were unconstitutional. But it is hard to understand why a private party has standing to object to the inability of the president to remove the agency head.
In response, this is what the Court said:
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.
The opinion then cited Free Enterprise and Bowsher to support standing, even though standing was not raised in either case. The Court then observed that there were “real world consequences to this case because although the parties agreed on the constitutional issue, petitioner and the Government disagree about whether petitioner must comply with the civil investigative demand.” None of that, in my view, explains how the private law firm was injured by the inability of the president to fire the director who issued the investigative demand to it, at least in the absence of a plausible claim that the president opposed the order. For a Court that has otherwise been very hard on standing in many other contexts, a cynic might suggest that standing is not a barrier when the Court wants to reach the merits.
In Seila Law, the government argued that the orders at issue had been ratified by an officer who was removable at will by the president, and thus the order was enforceable despite the removal flaw. The Court declined to decide that question, because it had not been passed on below (the court of appeals had rejected the constitutional objection), but it left the question open on remand from its constitutional ruling. On remand, the court of appeals held that a ratification of the purely prospective order after the Supreme Court’s decision eliminated any objection, and therefore it did not have to decide whether an earlier ratification by an acting director sufficed. As a result, in a case where the standing of the private party was far from clear, the Court could have avoided deciding a difficult and important constitutional question, as it was specifically asked to do by the court-appointed amicus and by the amicus House of Representatives. Instead, it chose to reach out and rule on the merits. It is especially hard to see why the Court was justified in doing that because the Obama-appointed head of the CFPB remained in office for ten months after Donald Trump became president and no effort was made to remove him during that period.
In Collins, the Supreme Court struck down the restriction on the removal of the head of the FHFA, but that same opinion upheld the order at issue on the ground that an acting official had made the actual decision being challenged. That was acceptable, according to the Court, because the acting official could be removed by the president at will. I return below to the question of whether an acting official, who is not a principal officer, can constitutionally make such a decision, but for now I observe only that the Court should have invoked the doctrine of constitutional avoidance and not reached the removal issue, but instead upheld the order on acting-director rationale which it subsequently employed in that opinion.
III. Arthrex and More Presidential Accountability
In 2011, Congress passed the America Invents Act which provides for an alternative means—inter partes review—by which a person accused of patent infringement could challenge the validity of the patent, instead of litigating that defense in federal court. That option allows any person to petition the Patent and Trademark Office (PTO) to hear the challenge to these often very valuable patents, with the administrative patent judges (APJs) who make up the Patent Trial and Appeal Board given the authority to decide whether to grant the request. If the request is granted, the case is heard by a panel of at least three (and usually no more) APJs, with one side arguing for the patent’s validity and the other taking the contrary position. The statute provides for no further internal agency review, but the losing party has a right to an appeal on the legal issues to the U.S Court of Appeals for the Federal Circuit, which is where infringement cases decided in district courts would be heard. The director of the PTO is a principal officer, having been appointed by the president and confirmed by the Senate, whereas the APJs, who now number more than two hundred, are inferior officers. The appointments of ALJs are vested in the secretary of commerce, who is the head of the department in which the PTO is lodged. APJs are not removable from their office by the secretary except for cause that would promote the efficiency of the service, and such actions are subject to review by the Merits System Protection Board (as are the appeals of many other federal employees).
The issue presented in Arthrex was whether these adversary adjudications performed by APJs could properly be done by inferior officers or whether a principal officer’s involvement was required. The United States defended the statute, arguing that the director, while having no authority to overturn a decision by an APJ panel, had other sufficient supervisory authority over their work so that their decisions were, in effect, his decisions. While not purporting to lay down a rule applicable to all administrative agency decisions, the Court held that decisions made by APJs could not be made by APJs alone, because no principal officer adequately supervised the actual rulings. Once again, in reaching the conclusion that the APJ decisions at issue were invalid, the chief justice gave great weight to the necessity of assuring presidential accountability, which he argued could only be achieved if the decisions were made by principal officers who are directly accountable to the president, which APJs are plainly not.
That left open the question of remedy. The Federal Circuit concluded that APJs could take on the duties of principal officers, despite their lack of presidential appointment and Senate confirmation, by eliminating the for-cause removal protection. The Trump administration embraced that approach, because it would have, in essence, given them the victory on the issue of restrictions on ALJ removal that the Court refused to reach in Lucia. But the majority on that issue—which included three of the justices who dissented on the merits—took a different approach. They applied the principle of severability and struck the provision that limited the power of the PTO director to review APJ decisions on the merits, thus adding a new level of potential review by a principal officer. The Court did not require the director to rehear every case in the future, but it did require him to consider rehearing the ruling in this case, and presumably others in which this type of challenge was made. There are open question as to how often the director will choose to grant rehearings, and whether the Court will be satisfied if he rehears only a handful of cases each year out of the five hundred APJ decisions issued annually. Moreover, the Court’s solution is at most temporary, as Congress can amend the statute by creating an appellate board comprised of principal officers, or it could make APJs principal officers and eliminate other internal agency review.
IV. The Next Cases
In the wake of these Supreme Court cases, there is the question of what comes next. But there is no doubt here that there will be similar litigation in the future, quite likely involving the PTO, as well as other federal agencies. Indeed, there are already a number of cases in which the issue is whether ALJs at the Social Security Administration (SSA) are acting unconstitutionally as inferior officers because they were not appointed in the manner required by the Appointments Clause. As a prelude to deciding that issue, this Term the Court also decided that the failure of SSA disability claimants to raise their constitutional claim at the administrative level did not result in a forfeiture of their right to raise it in court.
There are about 1,420 ALJs at SSA who rule on disability and other claims. By statute, they were chosen by a merit selection process conducted by the Office of Personnel Management, not by the agency head, which would be required if they are inferior officers. Shortly after Lucia was decided, the Trump administration issued an Executive Order proclaiming that all ALJs were inferior officers, and in response, the commissioner of SSA, who holds a Senate-confirmed position, reappointed all of the agency’s ALJs as inferior officers, which solves the Lucia problem going forward. But there are thousands of cases like Carr still in court in which the ALJ decision was made before the reappointment, and, even though their decisions are subject to further review within SSA, that does not, as it did not in Lucia, save them from challenge.
There are, however, two related grounds on which Lucia can be distinguished that may be a proper basis for concluding that the ALJs at SSA are employees rather than inferior officers. In contrast to the proceedings at issue at the SEC in Lucia, in which the agency sought to impose significant penalties and an injunction against the private party, SSA proceedings are non-adversarial, in which the payment of government benefits is at stake. Thus, if every action in which a federal agency denied anyone a federal benefit had to be done by an inferior officer, there would be little left that employees could do. Moreover, in contrast to Arthrex, all ALJ decisions at SSA are subject to further internal agency review below the department-head level, and on that basis the Court should sensibly conclude that the ALJs at SSA are employees, not inferior officers.
But now there is also an Arthrex issue at SSA because final adverse decisions there are not made by the commissioner or any other principal officer. Instead, a claimant who objects to an ALJ decision is required to ask the Appeals Council at SSA to review the decision before they can sue in district court. There are approximately seventy-one members of the Appeals Council who appear to be inferior officers. If they are, their rulings are subject to an Arthrex objection because the final agency determination is not made by, or even subject to review by, a principal officer at SSA—unless the non-adversarial nature of these proceedings excludes them from the scope of Arthrex.
However, there is no SSA statute standing in the way of solving that problem, as the operative provision broadly assigns all duties at SSA to the commissioner, who has delegated the responsibility to decide disability claims to the agency’s ALJs and Appeals Council.  In that sense, the fix is easier than in Arthrex: “All” the commissioner has to do is to make final Appeals Council decisions subject to review by the commissioner, as the Court did in Arthrex. That would still leave open the questions of whether the theoretical possibility of commissioner review will suffice for the approximately one-hundred thousand adverse Appeals Council rulings issued each year, and how many cases must the commissioner actually review to satisfy the Court. I am unaware of any SSA cases raising the Arthrex issue, but there are sure to be some, especially if the ALJs are held not to be inferior officers. But prudence dictates that the commissioner should change the agency’s regulations immediately to provide for the right to seek commissioner review, which is a necessary and perhaps sufficient remedy for any Arthrex problem.
The other major agency for which Lucia and Arthrex present problems is the Department of Justice, and in particular the Executive Office of Immigration Review (EOIR). That office adjudicates all cases in which the government is seeking to remove (deport) an alien or in which the alien is seeking asylum or other adjustments of status. There are many different ways in which the issues arise, but for simplicity, here I will discuss the most common proceeding in which the government has located an alien who it alleges is subject to deportation, and the alien contests the allegations. The case is assigned to one of over five-hundred or so immigration judges (IJs) who are appointed by the Department but are not ALJs. Although removal proceedings are not formal adjudications like those at the SEC, they are adversarial, in contrast to those at SSA. Based on the nature of these IJ adjudications, it is unclear whether the Court would conclude that IJs are inferior officers who must be appointed by the attorney general.
There is another basis on which the Court might conclude that IJs are employees, not inferior officers. In contrast to the SEC, where decisions are appealable only to the agency itself, adverse rulings by IJs are appealable to three-judge panels of the Board of Immigration Appeals (BIA), whose twenty-three members are inferior officers, appointed by the attorney general. There is also a further possibility of agency review by the attorney general, which can occur if he takes a case sua sponte, or if the BIA, or in some situations the Department of Homeland Security, refers a decision for attorney-general review. It is possible that the Court would conclude that review of IJ decisions by the inferior officers at the BIA is a further reason why IJs are not inferior officers.
But a holding that IJs are not inferior officers does not solve the problem of whether decisions of the BIA are invalid because they are made by inferior, not principal, officers. As was true for decisions of APJs prior to Arthrex, there is no formal means by which aliens, in contrast to the government, can seek attorney-general review of a decision adverse to them. And even then, attorney-general review is a rarity. According to a recent study, attorney-general review was almost never used until the Trump administration, where the attorney general reviewed twelve cases, all of which were undertaken to issue broad decisions designed to make it more difficult for aliens to prevail.
Following Arthrex, there would seem to be a strong argument that every alien who loses before the BIA must be entitled to seek review by a principal officer. For now, that would have to be the attorney general, but it might also be other Senate-confirmed officers such as the deputy or associate attorneys general. Even if that avenue were open, the issue would still be whether the possibility of that review, even if it were almost never realized, would suffice. One way to avoid that problem would be for Congress to provide that the BIA (or a subset of it), must be appointed as principal officers and then take the attorney general out of the review process entirely.
The current process for deciding immigration cases is also subject to a different challenge. Because IJs and members of the BIA have no for-cause removal protection, they lack the independence that basic fairness and due process require. In 2019, the Federal Bar Association drafted a bill to create an independent Article I immigration court, like the Tax Court. It could easily be structured to comply with Arthrex, and it would eliminate the problem that the attorneys who prosecute those cases for the government and the attorneys who judge them all work for the attorney general and serve at his pleasure.
V. Reassessing Presidential Accountability in the Context of Arthrex & Collins
With the impact of these decisions in mind, both for the cases themselves and for those that are likely to follow, it is time to look at the premise of presidential accountability which undergirds them. The Appointments Clause expressly requires Senate confirmation for principal officers and requires Congress to enact a law to create an inferior office, which may be filled only by one of three approved methods. By contrast, the Appointments Clause says nothing to support the Court’s conclusion that all (significant) decisions made by agencies must have been approved by (or had the opportunity to be reviewed by) a principal officer. Rather, the textual source for that conclusion seems to be the Take Care Clause, which vests in the president the power—or, as some read it, the duty—to “take care that the laws be faithfully executed,” but says nothing about a presidential removal power. The Appointments Clause makes clear that the president was never expected to act on his own in carrying out that responsibility, and there is nothing else in the text of the Constitution that compels the conclusion that Congress cannot vest in inferior officers the power to make decisions under laws created by Congress, in contrast to those decisions that are inherent in the role of the president.
Even accepting that the president must have, as Seila Law holds, the power to remove at least the heads of agencies with a single director, that alone does not preclude Congress from allowing inferior officers to have the final administrative say on a matter, especially when the ruling will be subject to full judicial review in a federal court. In other words, the result in Arthrex is based on a very muscular theory of presidential power, sometimes going by the name of the unitary executive. As the prior discussion shows, that theory will have significant implications for the administration of the laws assigned by Congress to federal agencies for implementation. Thus, Arthrex is likely either to require the appointment of many more principal officers who will have final agency decisional authority, or to require the heads of departments or other principal officers, who already have more than full plates, to take on the additional duty of reviewing administrative adjudications.
There are further objections to the Court’s theory of presidential accountability for agency adjudications generally. The theory assumes that the principal officer, who is responsible to the president, actually makes these decisions. We do not know what the director of the PTO will do with his new-found power to review APJ decisions, but we do know that there are hundreds of inter partes rulings each year, that the issues are complex and very significant to the parties and often to others, and that the director had a full-time job before these new duties were thrust on him by the Court—with no opportunity to debate alternatives or to explain what a burden this will be. Of course, he can hire staff to help him decide which cases to review in full and help him decide how to decide them. That will permit him to put his name on the final ruling and avoid any problem under the Morgan line of cases, but no one who pays the slightest attention will think that the decisions are his, except for the very few cases the director accepts for full review on the merits. To this observer, adding a layer of possible director review does not add much (if any) accountability simply because a very busy PTO director might reconsider a few rulings a year made by full-time, independent APJs.
If the Court is serious that the person in the executive branch who has the final decisional authority must be a principal officer subject to removal by the president at will, then the result in Edmond v. United States, will be called into question. At issue there was whether the civilian members of the Coast Guard Court of Military Review were inferior officers, with the Court agreeing that they were. Their decisions were subject to review only in the United States Court of Appeals for the Armed Forces, which is part of the Defense Department and whose members are appointed for fifteen-year terms and confirmed by the Senate. But, with limited exceptions, review in that court is discretionary, and in 2019, that court reviewed 425 petitions and granted only fifty-two, or 12.2%. In terms of accountability and fairness to the parties, it is by no means clear why, per Edmond, discretionary review by military judges in the executive branch, who cannot be removed except for cause, was constitutionally sufficient to supervise the courts of military review, but mandatory review by an Article III court for decisions of APJs in Arthrex was not.
A further flaw in the presidential-accountability theory is that no one—well, perhaps almost no one—will likely, or even conceivably, think that the president is responsible for an erroneous decision by the director who reviews an APJ panel decision. Yet that seems to be the Court’s animating theory. Assigning responsibility to the president is even less likely when, as is almost certain to be the result, the director declines to review the vast majority of panel decisions. In assessing the likelihood of the president being held accountable for an erroneous decision by a director, a knowledgeable observer would also factor in the right of a losing party to seek judicial review in the Federal Circuit of questions of law (which are what the panels are directed to decide) to assure adherence to the requirements of patentability that Congress has enacted. Indeed, it is not as though the president has the authority to override the APJs or the director in a particular case. At most, he can remove the director because of the way in which he reviews or does not review APJ rulings generally or even in specific cases. But even that seems highly unlikely to occur, both because the director has many other duties on which his performance is being judged and because the president will then have to find a replacement and obtain Senate approval of him or her.
There is another way to consider presidential accountability, but it also does not support the outcome in Arthrex. Presidents are held responsible when big things go wrong during their administrations, such as the failure to get Covid under control, the failure to stem the flood of aliens at the border, the bills that the president cannot persuade Congress to enact, or major foreign policy forays that he undertakes like the war in Iraq or the inability to evacuate all of Afghan citizen who helped the United States when we pulled out of that country. Even then, the president serves a four-year term and cannot serve more than two of them. Thus, other than the possibility of impeachment, elections—his or her own as well as that of any successor and perhaps those of his party in Congress—are the only formal ways by which the people can hold the president accountable. Does anyone think that more than a handful of voters will even be aware of bad decisions by the PTO director, let alone connect them to the president, or that such awareness will ever change a single vote in a presidential or other election? If the answer to that question is no, as I believe, then it is very hard to see how any notion of presidential accountability can support the result in Arthrex.
On top of these serious questions about the application of the theory of presidential accountability, one must add the Court’s willingness in Collins to allow those who are acting in the place of principal officers to perform their duties, even if the acting person has not been appointed by the president and confirmed by the Senate. That willingness raises two questions, one of which was raised by Justice Thomas in a recess appointment case that was resolved adversely to the agency on statutory grounds: “Whether directing Lafe Solomon to serve as acting general counsel of the National Labor Relations Board (NLRB or Board), without the advice and consent of the Senate, complied with the Constitution.” In other words, are the laws providing for the filling of vacancies at federal agencies constitutional if the office being filled is a principal one, and the acting person has not been appointed by the president and confirmed by the Senate?
That constitutional objection would be strengthened by the existence of the Recess Appointment Clause, from which it could be argued that the Framers were aware of the potential problems with vacancies and that recess appointments were their sole solution. To be sure, the Court has recognized that there may be an emergency exception to the mandates of the Appointments Clause, and not every action taken by a principal officer needs to be taken by someone who has been Senate-confirmed, although neither of those rationales would support allowing an acting head of FHFA to make the momentous decision at issue in Collins. Even then, as Justice Thomas also recognized in his concurrence, this constitutional limit on the filling of vacancies would raise many practical problems in many agencies. However, in departments such as Justice, State, and Defense, where there are many Senate-confirmed officers, they could satisfy the Constitution even if they were not confirmed for the position that is vacant. Despite the ruling in Collins approving the action of the acting director, this issue is likely to be raised in future cases, including in the cases to be remanded under Arthrex unless the vacant PTO director position is promptly and properly filled, because there is no other Senate-confirmed officer at the PTO. Indeed, it is possible that this argument remains open for the plaintiffs in Collins because the Court did not rule on that specific claim.
The second question is not a legal one, but asks whether, in effect, accountability is a one-way street. According to the Court, the doctrine, which is found nowhere in the Constitution, forbids Congress from allowing officers not subject to presidential at-will removal from taking significant agency action. Yet, the Court is perfectly comfortable with allowing acting officials, who have not been appointed under the express terms of the Appointments Clause, to perform the very acts that the duly confirmed officers, like the director of the FHFA, or properly appointed inferior officers, like APJs, are forbidden from doing. The requirement of Senate confirmation is to provide a check on the president’s appointment power, yet the Court has allowed him to make vacancy appointments for principal offices that, effectively, do an end run around that process. From the perspective of the public and the goal of holding the president and the Senate accountable, it seems far preferable to have these decisions made by duly appointed officers, regardless of their removability than it does to allow the president to fill a vacancy with a person who has never been approved by the Senate.
* * *
Stated in the abstract, the idea of greater presidential accountability may not rank up there with motherhood and apple pie, but it comes fairly close. Yet, when that notion is translated into constitutional mandates as applied to agencies created by Congress to implement the laws that it has enacted, and it is used to overturn the means that Congress has chosen to carry out those laws, its application is very dubious, and the result looks more like what five justices would like our government to look like and less like what the Constitution requires.
[*] Associate Dean, George Washington University Law School. The author teaches civil procedure and constitutional law and has filed amicus briefs in several of the cases discussed in this essay opposing rulings that would give the president increased authority over agency tribunals. Thanks to my GW colleagues Dmitry Karshtedt, John Whealan, Richard Pierce, and Robert Glicksman for their help in this endeavor.
 United States v. Arthrex, Inc., 141 S. Ct. 1970 (2021).
 Collins v. Yellen, 141 S. Ct. 1761 (2021).
 Arthrex, 141 S. Ct. at 1985–86.
 Seila Law, LLC v. Consumer Fin. Prot. Bureau, 140 S. Ct. 2183 (2020) (holding that leadership by a single director removable only for inefficiency, neglect, or malfeasance violates the separation of powers).
 Marbury v. Madison, 5 U.S. 137 (1803).
 Collins, 141 S. Ct. at 1784 (citing Seila Law, 140 S. Ct. at 2192).
 For a brief overview of this history, see Myers v. United States, 272 U.S. 52, 118 (1927) (“In the British system, the crown, which was the executive, had the power of appointment and removal of executive officers, and it was natural, therefore, for those who framed our Constitution to regard the words ‘executive power’ as including both.”).
Id. at 52.
 Humphrey’s Ex’r v. United States, 295 U.S. 602 (1935).
 Congress has not been consistent in its description of what constitutes cause for removal, and some litigants have argued for narrower constructions of some for-cause provisions. The Court has not accepted those arguments, and so this essay will treat all for-cause removal limitations as if they are identically worded. Seila Law, LLC v. Consumer Fin. Prot. Bureau, 140 S. Ct. 2183, 2206 (2020).
 Bowsher v. Synar, 478 U.S. 714 (1986).
 Id. at 737 (Stevens, J., concurring).
 Humphrey’s Ex’r, 295 U.S. at 628.
 Morrison v. Olson, 487 U.S. 654 (1988).
 Id. at 686.
 Id. at 689–90.
 Free Enter. Fund v. Pub. Co. Accounting Oversight Bd., 561 U.S. 477 (2010).
 Id. at 484–87.
 Id. at 497–98.
 Lucia v. SEC, 138 S. Ct. 2044 (2018).
 Id. at 2055.
 Id. at 2050 n.1.
 Seila Law, LLC v. Consumer Fin. Prot. Bureau, 140 S. Ct. 2183, 2203-04 (2020) (quoting U.S. Const. art. II., § 3, cl. 3). Although I find the dissent of Justice Elena Kagan to be more persuasive, this essay focuses more on the question of whether the chief justice’s removal rulings in fact lead to presidential accountability.
 Id. at 2206.
 Id. at 2211–19 (Thomas, J., dissenting).
 Collins v. Yellen, 141 S. Ct. 1761, 1783 (2021).
 Id. at 1775–78, 1787.
 Myers v. United States, 272 U.S. 52, 107–08 (1927); Humphrey’s Ex’r v. United States, 295 U.S. 602, 609 (1935).
 See generally Bowsher v. Synar, 478 U.S. 714 (1986); Morrison v. Olson, 487 U.S. 654 (1988); Free Enter. Fund v. Pub. Co. Accounting Oversight Bd., 561 U.S. 477 (2010).
 Brief of Amicus Curiae Alan B. Morrison in Support of Neither Party at 7–16, Seila Law, LLC v. Consumer Fin. Prot. Bureau, 140 S. Ct. 2183 (2020) (No. 19‒7); Brief For Court-Appointed Amicus Curiae in Support of Judgment Below at 21–24, Seila Law, 140 S. Ct. 2183 (No. 19‒7); Brief of Amicus Curiae the United States House of Representatives in Support of the Judgment Below at 3, Seila Law, 140 S. Ct. 2183 (No. 19‒7).
 Seila Law, 140 S. Ct. at 2196.
 See TransUnion LLC v. Ramirez, 141 S. Ct. 2190 (2021); Clapper v. Amnesty Int’l USA, 568 U.S. 398 (2013).
 In his concurring opinion in Collins, Justice Thomas questioned whether the shareholder plaintiffs there had shown any unlawful act that adversely affected them, which might also suggest that they lacked standing, but that was not the basis of his concurrence. Collins v. Yellen, 141 S. Ct. 1761, 1789–99 (2021) (Thomas, J., concurring).
 TransUnion LLC, 141 S. Ct. at 2211.
 Consumer Fin. Prot. Bureau v. Seila Law LLC, 984 F.3d 715, 717 (9th Cir. 2020). Rehearing en banc was denied, with four judges dissenting. Consumer Fin. Prot. Bureau v. Seila Law LLC, 997 F.3d 837 (9th Cir. 2021). As of mid-September 2021, a petition for certiorari had not been filed.
 Then-CFPB Director Richard Cordray resigned in November 2017, over ten months after Donald Trump became president. Adam Edelman, Consumer Financial Watchdog Cordray Resigns, NBC News (Nov. 15, 2017).
 Collins v. Yellen, 141 S. Ct. 1761, 1783–84 (2021).
 35 U.S.C. §§ 311 et seq.
 35 U.S.C. § 3(a)(1).
 35 U.S.C. § 3(c) (providing APJs with civil service protections outlined in 5 U.S.C. § 7513); 35 U.S.C. § 6(a) (concerning the appointment of APJs).
 United States v. Arthrex, Inc., 141 S. Ct. 1970, 1981–86 (2021).
 Collins 141 S. Ct. at 1779‒84. Alternative views on why the existing supervision of APJs satisfy the Appointments Clause are set forth in the various amicus and party briefs filed in Arthrex. In an amicus brief in which I was joined by four other law professors, we argued that Congress should be given substantial leeway in deciding whether an officer was “inferior,” in large part because the text of the Appointments Clause gives Congress considerable discretion through its phrase “as they think proper” in describing how inferior offices are to be created. Brief of Amicus Curiae Administrative, Constitutional, and Intellectual Property Law Professors Urging Reversal and Supporting Petitioners, Arthrex, 141 S. Ct. 1970 (Nos. 19‒1434 & 19‒1452) [hereinafter Law Professors’ Amicus Brief in Arthrex]. In the opening portion of Justice Stephen Breyer’s opinion dissenting on the merits and concurring on the remedy, which was joined by Justices Sonia Sotomayor and Elena Kagan, he supported that approach, and then pointed out other reasons why the majority was in error. Arthrex, 141 S. Ct. at 1994 (Breyer, J., dissenting).
 Arthrex, 141 S. Ct. at 1978.
 Id. at 1987.
 Patent Trial and Appeal Bd., U.S. Patent and Trademark Off., PTAB Trial Statistics: FY20 End of Year Outcome Roundup IPR, PGR, CBM 4 (2021).
The director has issued interim rules to deal with the pending cases: See USPTO Implementation of an Interim Director Review Process Following Arthrex, U.S. Patent and Trademark Off. (June 29, 2021). According to a recent survey taken by the Federal Circuit in an effort to get a handle on its post-Arthrex docket, a significant percentage of the parties who could seek rehearing before the director chose not to do so. John C. Evans & David M. Maiorana, Most Arthrex Challenges Say “No Thanks” to Director Remand, Jones Day (July 19, 2021). Whether this scenario will apply only to cases already on appeal, or whether it will hold when the case is still at the PTO, remains to be seen.
 See Carr v. Saul, 141 S. Ct. 1352, 1357 (2021) (collecting cases).
 Id. at 1352.
 See generally, 5 U.S.C. § 1302; 5 C.F.R. Pt. 302 (2020).
 Exec. Order 13843, § 3(a), Schedule E, 83 C.F.R. § 32755 (2018).
 Carr, 141 S. Ct. at 1357. The blanket treatment for all ALJs may have been motivated in large part by the Trump administration’s desire to prevent ALJs from having a union on the theory that, if they are officers, not employees, they are not eligible to be in a union. It is not clear that, even if ALJs are inferior officers, that would preclude them from being in a union, but that question is beyond the scope of this essay.
 Id. at 1359–60. Moreover, disability cases are not like the inter partes reviews in Arthrex, in which private parties were battling over the validity of very valuable patents. Compare id. at 1357 with United States v. Arthrex, Inc., 141 S. Ct. 1970 (2021).
 20 C.F.R §§ 404.968, 404.981, 416.1468 (2020).
 42 U.S.C. § 902(a)(4), (7).
 Relying on Seila Law, President Biden has removed the SSA commissioner and his deputy, who are the only Senate-confirmed officers at the agency. Nicole Ogrysko, Biden Fires Saul as SSA Commissioner, Fed. News Network (July 9, 2021).
 Exec. Off. For Immigration Rev., Office of the Chief Immigration Judge, United States Dep’t of Justice (last visited Sept. 1, 2021) (there are 535 immigration judges nationwide); 8 C.F.R. § 1003.10 (2020) (outlining the appointment process for immigration judges).
 For an overview of immigration proceedings, see Stephen H. Legomsky, Restructuring Immigration Adjudication, 59 Duke L.J. 1635, 1641–45 (2010). See also 20 C.F.R. § 404.900 (2020) (describing SSA adjudication conducted in “an informal, non-adversarial manner.”).
 8 C.F.R. § 1003.1(a)(1) (2020). There are some proceedings, such as those in which an IJ rejects a claim that an alien lacks a credible fear of prosecution, in which there is no appeal (by express statutory preclusion) to the BIA/Attorney General. They raise distinct problems and may require a different solution than that proposed in the text. Id. § 1003.1(b).
 Id. § 1003.1(h)(i)-(iii).
 David Hausman et al, Executive Control of Agency Adjudication: Capacity, Selection and Precedential Rulemaking 7 (Apr. 30, 2021) (Unpublished manuscript) (on file with Soc. Sci. Research Network).
 FBA Model Legislation Establishing an Article I Immigration Court, Fed. Bar Assoc. (July 16, 2019).
 U.S. Const. art. II, § 2, cl. 2.
 Id. art. II, § 3, cl. 3.
 For a discussion of Arthrex and the unitary executive theory, see Michael C. Dorf, The Hidden Ideological Stakes of SCOTUS Patent Case, Verdict (Mar. 3, 2021).
 In support of its director-review remedy, the Court cites the recent amendment to the trademark law under which the director can review decisions of the trademark panels that are the equivalent to the panels conducting inter partes review of patents. United States v. Arthrex, 141 S. Ct. 1970, 1987 (2021). The fact that Congress chose to limit the director’s review to trademark cases might also suggest that it did not want to extend that approach to patent cases, for a variety of reasons, such as their greater numbers, their greater complexity, or the desire to see how the experiment worked before extending it further. In that connection, there are only twenty-five trademark judges who issue two hundred decisions annually that are comparable to the five hundred rulings issued by over two hundred APJs. TTAB Incoming Filings and Performance Measures for Decisions, U.S. Patent and Trademark Off. (last visited Aug. 31, 2021) (providing statistics on trademark decisions by fiscal year). On the issue of workload, Arthrex dealt only with inter partes decisions, but APJs and other officials at the PTO make many other decisions, some of which may come within the rationale of Arthrex, thus potentially further burdening the director. A substantial percentage of requests for inter partes review are denied, and it is unclear if the director must review them also.
 Morgan v United States, 298 U.S. 468 (1936); Morgan v. United States, 304 U.S. 1 (1938); United States v. Morgan, 307 U.S. 183 (1939); United States v. Morgan, 313 U.S. 409 (1941).
 Edmond v. United States, 520 U.S. 651 (1997).
 Id. at 653, 666; Welcome, U.S. Court of Appeals for the Armed Forces (last visited Aug. 31, 2021). (“The Court is composed of five civilian judges appointed for 15-year terms by the President with the advice and consent of the Senate.”).
 See Law Professors’ Amicus Brief in Arthrex, supra note 41, at 19–20.
 Compare Edmond, 520 U.S. at 664, 666 with United States v. Arthrex, 141 S. Ct. 1970, 1985 (2021).
 See 35 U.S.C. § 141(a) (“An applicant who is dissatisfied with the final decision in an appeal to the Patent Trial and Appeal Board . . . may appeal the Board's decision to the United States Court of Appeals for the Federal Circuit.”).
 U.S. Const. amend XXII.
 Collins v. Yellen, 141 S. Ct. 1761, 1783–84 (2021).
 N.L.R.B. v. SW General, Inc.,137 S. Ct. 929, 945–949 (2017) (Thomas, J., concurring).
 U.S, Const. art. II, § 2, cl. 3 (“The President shall have Power to fill up all Vacancies that may happen during the Recess of the Senate, by granting Commissions which shall expire at the End of their next Session.”).
 United States v. Eaton, 169 U.S. 331 (1898) (holding that the Appointments Clause allows Congress to vest power in the president to appoint vice-consuls to fill the office of consul in case of disability or absence without the advice and consent of the Senate).
 Justice Thomas did not mention that issue in his concurrence in Collins, perhaps because under his approach there, the plaintiffs would not have prevailed even if the acting director was unconstitutionally serving in that position. Collins, 141 S. Ct. at 1789–99 (Thomas, J., concurring).
 United States v. Arthrex, 141 S. Ct. 1970, 1986 (2021) (quoting Buckley v. Valeo, 424 U.S. 1, 126 (1976)) (noting that Congress has assigned APJs “‘significant authority’ in adjudicating the public rights of private parties, while also insulating their decisions from review and their offices from removal.”).
 Compare id. with Collins, 141 S. Ct. at 1783–84.