PCAOB Debate

This debate will focus on the following questions:

Is the Sarbanes-Oxley Act consistent with the principle of separation of powers insofar as it establishes the Public Company Accounting Oversight Board (PCAOB) to be overseen by the Securities and Exchange Commission (SEC), which in turn is overseen by the President? Is it contrary to the Appointments Clause for the members of PCAOB to be appointed by the SEC and thus not directly removable by the president? How is the Supreme Court likely to resolve the case and upon what grounds?

The participants of this installment of Originally Speaking are Martin Flaherty, Gillian E. Metzger, Donna M. Nagy, Tuan Samahon, Peter M. Shane, and Christian J. Ward.

Questions and Answers:

Peter M. Shane: Appointments: There's no real doubt that members of the PCAOB are "officers of the United States." That is, they have duties regarding the implementation of public law that go beyond the tasks Congress could assign to one of its own committees. Hence, its members must be appointed pursuant to the Appointments Clause. And, under the Appointments Clause, they must be appointed by the President with the advice and consent of the Senate, unless they are "inferior officers," in which case they may be appointed by the president alone, by the head of a department, or by a court of law.

This is the PCAOB's greatest vulnerability. The members of the PCAOB may well not be "inferior" in the constitutional sense. Although members are removable for good cause by the SEC, their jurisdiction is far more wide-ranging than that of the independent counsel upheld in Morrison v. Olson. The Court could leave Morrison and its antecedents intact, and enjoin the enforcement operations of the PCAOB on noninferiority grounds. This is doctrinally the most modest way to overturn the PCAOB, and I predict this will be the result, with hardly any greater implications for separation of powers law.

If PCAOB members are deemed "inferior," then I do not see any other vulnerability on the appointments side. As the Court observed in Morrison, Congress's discretion in choosing among the designated modes of appointing inferior officers is not limited by the text. There would not be anything constitutionally anomalous in giving the SEC power to appoint people with expertise in corporate accounting.

Removal: The more controversial question involves the limitation on direct removals by the President. It is not controversial under Morrison v. Olson. Morrison said that limitations on presidential removal powers are permissible unless they interfere with the President's capacity to discharge his constitutionally assigned functions. The President, of course, is constitutionally obligated to take care that the laws be faithfully executed. If a PCAOB member is derelict in this regard, the President must be able to instigate that member's discharge. Under Sarbanes-Oxley, he cannot do so directly – which was also true in Morrison v. Olson – but the failure of the SEC to correct any such dereliction would presumably be good cause for the dismissal of any recalcitrant SEC Commissioner. Under Morrison, this holds up.

The rub, of course, is that there may well be five members of the Court who would now like to overrule Morrison – Roberts, Alito, Scalia, and Thomas, almost certainly, and quite possibly, Kennedy, who recused himself in Morrison. Reaching out to limit or reverseMorrison, however, would be a conspicuous piece of judicial immodesty, especially since the PCAOB can be invalidated on the less controversial ground of noninferiority.

I thus predict the Court will not attack Morrison – but this may be wishful thinking on my part because (a) I agree with Morrison and (b) modesty on the Roberts Court is, at best, an occasional virtue.

Gillian E. Metzger: Since Peter has started the ball rolling, here's my two cents. I wanted to disclose at the outset though that I filed an amicus brief on behalf of constitutional and administrative law professors in support of the respondents in this case, so I'm not an impartial observer.

I think Peter's right that the way to resolve this case on the merits with least impact on the Court's precedent or for the future is to hold that the Board members are principal officers. The Court has acknowledged that the line between principal and inferior officers is murky, and resolving it on this basis would be a more fact-dependent determination than any ruling on the removal challenge.

I disagree though about how easy it would be for the Court to find the Board members are principal officers. Under Edmond v. United States, the key to inferior v. principal status is degree of supervision, and the Board is truly supervised to an extraordinary degree by the SEC. The degree of oversight is evident from the face of the Sarbanes-Oxley Act creating the Board: the SEC must approve all Board rules (including the rules the Board is required to establish to govern its investigation and disciplinary proceedings), reviews Board sanctions on appeal or sua sponte, controls the Board's budget, appoints and can remove members of the Board, can take away areas of responsibility from the Board, and has broad independent authority itself to issue rules it thinks necessary or appropriate. An amicus brief filed in the case on behalf of several former SEC Chairs demonstrates that the extent to which the SEC has exercised these powers and closely supervises the Board in practice.

Peter's also right that the removal challenge here has much more far-reaching implications. The petitioners try to limit that impact by at times framing their claim as being simply that the President or a presidential alter-ego must be able to directly remove (whether at-will or subject to a for cause limitation) any executive official. That condition is not met here because the SEC Commissioners themselves enjoy for cause removal protections and thus aren't presidential alter-egos. But even this more limited claim holds pretty significant implications for the future of independent agencies if upheld by the Court, given that many employees and inferior officers at such agencies have some form of protection from removal. It's also hard to read a decision adopting a requirement of a direct presidential role in all removals as not casting some doubt on the Court's continued adherence to Morrison andHumphrey's Executor, even if the Court doesn't take those decisions on directly.

I don't have great faith in my predictive abilities, but even so I have a hard time seeing a majority of the Court upholding the removal challenge here given the implications for established precedent and the shape of the modern administrative state. Moreover, the Court would be making such a move in a context of not very supportive facts; given the extent of SEC oversight powers, Board members' for-cause removal protection is a formality that does no work in insulating them from the SEC.

The Court has another route it might take here, however, and that is to not reach the merits at all because jurisdiction is lacking. The jurisdictional problems with this case are quite significant. Instead of pursuing the statutory prescribed path of administrative appeal to the SEC and then judicial review of the SEC's decision, the petitioners went straight into federal court with their constitutional claim against the Board. Not only is there a strong argument under Thunder Basin that this statutory review mechanisms is exclusive, there's no apparent statutory cause of action available to the petitioners here. In particular, as the Board is a not-for-profit corporation and not an agency, the APA doesn't apply. Although the Court has been willing to find jurisdiction over preemption claims even when a cause of action appears lacking, it has yet to do so in the separation of powers context and implying a right of action appears particularly dubious here given the statutory provision of a route through which petitioners' constitutional claim could be heard. It also is worth noting that the petitioners' claims turn to some degree on interpretations of somewhat ambiguous statutory provisions governing the SEC's oversight of the Board, such as the provisions governing the scope of the SEC's removal power. That's exactly the type of question on which the SEC's views would be particularly important, and thus reinforces the reasons to require petitioners to pursue the statutory mechanism of administrative review before going to court.

Tuan Samahon: It is likely that the removal and the broader separation-of-powers issues will interest the Court.

First, as to the appointments issue, Peter fears that the Court might overrule Morrison v. Olson on the inferior officer issue. Like it or not, that already happened in 1997 in Edmond v. United States. This case is simply the vehicle by which the Court may recognize that Edmondoverruled sub silentio Morrison v. Olson's Excepting Clause holding. PostEdmond, to be an "inferior" officer is to be a subordinate to the appointing officer. The sine qua non of inferior officerhood is being a subordinate to the appointing officer. Thus, really powerful officers with broad jurisdiction (e.g. bankruptcy judges) can still be "inferior" if subordinate to a superior (e.g. U.S. Court of Appeals judges). What remains is for the Court to give us further guidance on what it means to be a subordinate in particular contexts. Scalia, writing for an eight justice majority in Edmond(Souter concurring on the Excepting Clause analysis), said that to be a subordinate is to be directed and supervised at some level by the appointing officer. I don't know what construction – further doctrinal implementation – the Court might offer of what it means to be a subordinate, but assuming the Court (likely) concludes that the PCAOB members are subordinate to the SEC commissioners (they are directed and supervised at some level), then their appointments present no problem.

Second, as to the removal issue, the PCAOB arrangement is at odds with the Constitution's creation of a unitary executive and is therefore unconstitutional for that reason. The case will pit formalists – who will find that the executive vesting clause doesn't leave room for guesswork as to who the Constitution vests with the executive power (it's the President) – against functionalists, who likely are inclined to believe that this is "only" a minor, but well-justified, incursion on executive authority that leaves the core of executive power with the President. It's indisputable that an unqualified removal power is a potent tool of control. Indeed, plenary removal power is the gold standard for control. But how useful as a tool of control is a qualified power to remove? Not terribly useful. The Attorney General could remove independent counsel Alexia Morrison only for qualified grounds and thus she could truthfully be denominated "independent." Agencies headed by individuals enjoying such protection from removal are similarly denominated "independent." "Independent" (from the President) is at odds with control by the President.

So what should the Court do? Assuming the Court is unwilling to overrule Humphrey's Executor but is disinclined to uphold PCAOB (likely, the members of the Court did not grant certiorari to affirm the D.C. Circuit), it needs to offer an account of why independent agencies are constitutionally permissible but the twice-removed PCAOB arrangement is not. One approach may be to say "thus far and no further." Acknowledge that presidential control is attenuated by any limitation on removal power, but not permit the (inverted?) squaring ofHumphrey's Executor i.e. that presidential control is "so" attenuated that it's no longer control.

Another possible path that the Court might take – that would allow it to uphold Sarbanes-Oxley – is to recharacterize the removal power over the PCAOB as more robust than has hitherto been thought. There is a surprising dearth of authority in the constitutional law context on removal only upon good cause or other limited grounds. The Court might try to clarify how far we should take the logic of Bowsher v. Synar, viz. the idea in another context that removal power, even when restricted to qualified grounds, can create a "here-and-now subservience" of the removable to the remover. After all, the SEC commissioners may remove the PCAOB members for qualified grounds; the President may remove the SEC commissioners for qualified grounds. Traditionally, most executive unitarists have not thought that independent agency heads are here-and-now subservient to the President (or that the PCAOB members are here-and-now subservient to the SEC). Perhaps, though, the statutory canon of constitutional avoidance requires that such removal powers be interpreted to allow broad presidential removal authority that amounts to removal at will. If such a reading of removal authority has not often been exercised, perhaps it is simply a lack of presidential political will rather than a lack of presidential constitutional capacity.

Martin Flaherty: Assuming history is a source of constitutional guidance, and not merely window-dressing for results achieved through other means, it might be useful to review how the Founding bears on these issues. This is doubly so since original understanding arguments figure prominently in just these types of separation of powers disputes. Ironically for many lawyers, less so for historians, here history fails as a constraining device at precisely the point it is needed.

The Founding materials provide somewhat more direction with regard to the appointments issue. The obvious starting place, for originalist as for textual analysis, is the Appointments Clause itself. As Peter Shane points out, the text reflects fairly precise guidelines. (One might argue they are even more precise should one factor in apparent structural limitations. The option of judicial appointment for inferior officers might be understood as limited to officials involved in judicial administration. Not only did Morrisonreject this, of course, but there is scant originalist support for this limiting principle in the sources). The problem is that the materials of the time cast little light on the standards for determining superior from inferior officers, which both Shane and Gillian Metzger rightly argue, is the key for determining this aspect of the case. Moreover, even if some Founding consensus, or something close, could be gleaned, translating it to the infinitely more complex array of different types of officials and agencies that characterize the modern Federal government would be a daunting challenge, to put it mildly.
 
The Founding offers even less help on removal. Much like the Treaty Clause, which speaks to how treaties are made but not terminated, the lack of any Removal Clause to complement the Appointments Clause among other things demonstrates that the Founders were good at getting things started rather then thinking about how they should end (much like relationships). A review of 18th century political thought, American patriot literature, the first state constitutions and their critics, the Federal Convention, and the ratification debates, fail to yield even a dominant thread on removal authority, much less anything approaching widespread agreement. David Currie's assessment of the famous "Decision of 89" concerning who should remove the newly established Secretary of State, stands as a telling example: about a third of the First Congress believed that it could limit the President's removal authority; about a third believed they could not; and the other third believed that they could but for policy reasons decided not to.
 
This is not to say that the Founding offers no guidance, only that such guidance is pitched at a very general and functional level. Separation of powers, among other things, is to promote a balance among the branches to prevent tyranny, to promote efficiency, and to foster a complex type of joint accountability that each of the branches would where appropriate share. That made for a document that "inks in" separation of powers at the top, e.g. the Appointments Clause, but leaves a shocking number of things to be resolved only by these general principles. Finally, and whatever else, one thing the Founding does not do is support the shortcut of reading the word "executive" as shorthand to a bundle of necessarily related powers, such as removal. Too many of the Founding generation simply did not come to this conclusion for it to stand as an "original understanding."
 
So where does that leave this case. I am inclined to go with Professor Shane and Metzger's analysis. As for a prediction, though, I would not be shocked to see a majority clip backMorrison based upon historical arguments that the historical sources do not support.

Peter M. Shane: I should clarify that I do not fear trimming Morrison on noninferiority grounds. That was the weakest point of Morrison. Tightening up on noninferiority would do little harm to the theory or practice of checks and balances.

Tuan is right that constitutional questions regarding PCAOB could be avoided if "for cause" removal were simply given a more robust reading. If, however, policy disagreements were sufficient to permit removals on good cause grounds, then Congress would have acted pointlessly in insulating any administrators from "at will" discharge.

I agree with Marty's historical conclusions about the unitary presidency. Here's what I wrote in Madison's Nightmare: How Executive Power Threatens American Democracy (University of Chicago 2009):

"[W]hatever commitment the founding generation had to centralized management did not translate–even in the late eighteenth century–into a model of executive policy control over all administration at the state or federal level. In 1789, the first Congress created, by statute, four civil administrative establishments–the departments of War, of Foreign Affairs, of the Treasury, and of the Post Office. . . .The statutes that organized [the War and State] departments sketched their departmental organization and their duties only in broad terms and Congress anticipated that the President would exercise broad independence in supervising these departments.

"It is clear from the debates surrounding the Treasury Department, however, that the character of Treasury was not perceived to be executive in a manner akin to the departments of War and Foreign Affairs. First, unlike the other departments, Congress did not label the Treasury as an 'executive' department. Congress went to great length to specify the Treasury's structure, creating a number of fiscal officers with reporting responsibility to the Secretary. Most tellingly, the first Congress treated the Secretary of the Treasury as if that official's administrative obligations ran to Congress as well as to the President. The most important debate regarding the Secretary centered around a proposed duty 'to digest and prepare plans for the improvement and management of the revenue, and for the support of public credit.' This wording is nearly identical to the charge to the financial officers authorized under the Articles of Confederation. . . .

"The implications of this analysis were borne out in practice. For several years, Alexander Hamilton, the first Secretary of the Treasury, performed the functions of the House of Representatives Committee on Ways and Means, which the House created in 1789, but suspended the same year. The House re-appointed the Ways and Means Committee only upon Hamilton's retirement from the Treasury. Congress's newly asserted independence in fiscal affairs was not thought to imply any intensification of presidential responsibility for the Treasury. As late as 1823, the Attorney General advised the President that, far from the President being constitutionally obligated to correct an allegedly improper settlement by the Treasury Department of an individual's claim for reimbursement, it would be legally impermissible for the President to substitute his judgment for that of the accounting officer. The vision of an air-tight executive establishment under unlimited presidential control that is conjured by presidentialists thus cannot be matched by the early history of administrative practice."

Donna M. Nagy: Like Gillian, in the interest of full disclosure, I must mention that I filed an amicus brief with the Supreme Court. I filed on behalf of 15 corporate and securities law professors in support of the petitioners. The brief makes clear that we applauded Congress's decision in the Sarbanes-Oxley Act to create a new regulator to oversee the conduct of the auditors of public companies. But it expresses a concern that Congress accorded the PCAOB substantial governmental power without imposing constitutionally sufficient accountability. Although I taught constitutional law early in my teaching career, almost all of my teaching and writing over the last decade has been in the area of corporate and securities law. 

As I see it, the DC Circuit's 2-1 decision upholding the constitutionality of the PCAOB was predicated on the majority's determination that the PCAOB was a "heavily controlled component" of the SEC. If the majority's depiction of the SEC-PCAOB relationship is incorrect, then the validity of its Appointment Clause and separation of powers analysis falls with it. 
 
I argue in the brief (and in a forthcoming article in the Pittsburgh Law Review), that the PCAOB is not a "heavily controlled component" of the SEC. It's clearly not a "component" of the SEC – even the respondents do not go that far in making that argument. The PCAOB was established as a private corporation – it's an entity separate and apart from the SEC. It's that separateness that justifies the PCAOB's exemption from the FOIA and the APA. And it is that separateness that allows the PCAOB members to command their substantial private sector salaries.
 
But we can leave semantics aside to focus on the question of whether the PCAOB is subject to the SEC's "extraordinary" control, with the PCAOB "subject to a vast degree of Commission control at every significant step." This finding is essential to the claim that PCAOB members are inferior officers, because as others in this debate have emphasized, to be considered "inferior" under the Court's decision in Edmond, an officer must be directed and supervised at some level by a principal officer. If PCAOB members operate with substantial discretion and autonomy, then notwithstanding the SEC's oversight and enforcement authority, they are principal officers who must be appointed by the President with advice and consent from the Senate. And as principal officers performing significant executive functions, PCAOB members must be removable by the President for cause. 
 
If the Supreme Court chooses to take this more fact-based approach to the constitutional questions, it will likely focus on three areas: (1) the text of the Sarbanes-Oxley Act (with respect to the PCAOB's powers and the SEC's role as overseer), (2) the legislative history evidencing Congress's intentions in creating the PCAOB with members who are removable only for-cause (with "cause" defined restrictively), and (3) the securities industry's self-regulatory (SRO) model on which the PCAOB is based. I've exceeded my word limitation, so later posts will have to discuss how all three areas support the conclusion that PCAOB members act with substantial discretion and autonomy and are thus principal officers underEdmond.

Christian J. Ward: I filed an amicus brief in support of the PCAOB on behalf of the Council of Institutional Investors and other signatories. The views expressed in this debate are my own and do not necessarily reflect policy positions taken by the Council or other signatories.

I agree with Gillian that it is far from clear that the Court would have an easy time declaring the PCAOB members principal officers rather than inferior officers under established precedent. To the degree the Court is willing to adhere to Edmond v. United States, it would have a hard time finding that the Board members were other than inferior officers given the extensive control over their activities by the SEC. The PCAOB's budget and rules are subject to SEC approval. The SEC can review any disciplinary action taken by the Board on its own motion and can enhance, reduce, otherwise modify, or completely overrule any sanction imposed by the Board. 15 U.S.C. §§7211(g), 7217(c)(3), 7219(b). 
 
For all intents and purposes, it is the SEC regulating the accounting industry through the PCAOB, and that's in line with the well-trodden precedential path we have traveled to the modern administrative state. For close to a century now, the Court has been upholding congressional determinations to vest regulatory power in independent agencies. E.g., Mistretta v. United States, 488 U.S. 361 (1989) (Federal Sentencing Commission); Wiener v. United States, 357 U.S. 349 (1958) (War Claims Commission); Humphrey's Ex'r v. United States, 295 U.S. 602 (1935) (Federal Trade Commission). The SEC's oversight of the accounting industry through the PCAOB is very similar to the SEC's long-established supervision of SROs like FINRA and the NYSE and the Financial Accounting Standards Board (FASB), except that the SEC has vastly greater control over the PCAOB than it has over any of those entities.
 
I agree with Peter that under Morrison the removal issue doesn't present a hard question. I won't hazard a guess as to whether the Court has a majority to overturn Morrison. If it does, it could—and should—however preserve the PCAOB by severing the removal provision and upholding the rest of the statutory scheme creating the Board.
 
Cases including Regan v. Time, Inc. have made clear that the presumption is in favor of severability even when a statute contains no severability provision. Under Regan, the touchstone for severability is legislative intent. The statutory scheme and—if necessary to resort to it—the legislative history make clear that Congress's overriding intent in creating the PCAOB was not to insulate it from the Executive Branch but rather to make it independent from the accounting industry it would be regulating.
 
The removal provisions could be excised from the statute in either of two ways without disrupting the rest of the statutory scheme. If the Court deems that generic for-cause removal power for the SEC is acceptable, it could make appropriate strike-outs. If the Court deems it necessary for the SEC to have plenary at-will removal power, it could strike the removal provisions entirely, leaving a statute silent as to removal that is presumed to vest the appointing authority with at-will removal power under cases like Shurtleff v. United States, 189 U.S. 311 (1903).

Tuan Samahon: Donna: Here is a little friendly fire. The last line of your post intrigues me. You said "PCAOB members act with substantial discretion and autonomy and are thus principal officers under Edmond." AssumingEdmond, and not Morrison, now controls, I wonder what implications would follow from reading the Edmond requirement of supervision and direction "at some level" to preclude, as you suggest it would, substantial discretion and autonomy in officers such as the members of the PCAOB. For example, Congress has vested the U.S. Courts of Appeals with the power to appoint bankruptcy judges pursuant to the Excepting Clause. In vesting their appointments with the Courts of Law, Congress characterized the bankruptcy judges as "inferior" officers. Bankruptcy judges exercise substantial discretion and autonomy (especially for core proceedings in which their orders are self-executing, subject only to appellate review). Significantly, bankruptcy judges are removable ONLY for cause by the U.S. Courts of Appeals. Thus, the construction you advance of what it means to be subordinate would likely sink the appointments of every U.S. bankruptcy judge, because surely they too could be said to be "principal" rather than inferior officers. Although I teased out exactly such a challenge in a recent article (60 Hastings L.J. 233 (2008), my suggestion was that the subordinate interpretation of inferior is consistent with a more remote level of control, or as Scalia's Edmond opinion emphasized, a departmentalized hierarchical supervision and direction that could be remote, i.e. emphasis on "at some level."

Donna M. Nagy: Taun: You raise a number of interesting points. In deciding whether an officer is "subordinate" to a principal officer, I would say that the amount of discretion and autonomy he or she exercises is an important variable in analyzing whether that officer is "directed and supervised at some level" – but that variable needs to be placed in the context of the officer's overall responsibilities and powers as well as the oversight to which the officer is subject.

As others have noted in this debate, SOX charged the SEC with oversight responsibilities in a number of well-delineated areas. SOX provides for SEC review of PCAOB final disciplinary sanctions, and that review power could well be sufficient to render PCAOB members, acting solely in their adjudicative/disciplinary capacity, subject to SEC direction and supervision at some level. SOX also provided that no PCAOB rule can become effective over the SEC's objection. So the SEC's role as a "reviewer" of PCAOB rules could well make PCAOB rulemaking subject to the SEC's direction and supervision at some level.

But in addition to significant rulemaking and adjudicative/disciplinary powers, SOX also vests PCAOB members with substantial investigatory and enforcement powers. SOX, however, assigns the SEC no specific statutory oversight role with respect to PCAOB investigations and enforcement determinations (excepting subpoena requests to 3rd parties outside the auditing industry, and a notification entitlement when a PCAOB investigation involves a possible violation of the federal securities laws). PCAOB members possess enormous discretion and autonomy with respect to investigative and enforcement determinations, and function much like U.S. Attorneys – and SEC Commissioners – in their ultimate responsibility for investigating and prosecuting alleged violations. The PCAOB decides the scope and duration of any investigation (and the investigative phase can often last a year or more); the PCAOB decides whether enforcement action is warranted; the PCAOB decides which firms or officials should be named as respondents and the specific violations that will be alleged; and the PCAOB decides whether (and how) to litigate or settle. SOX assigned a role for the SEC – the role of a reviewer – only at the completion of a disciplinary proceeding when a sanction is imposed. And if the PCAOB determines not to initiate an enforcement action against a firm or official, the SEC has no review role at all.

To be "inferior" officers, PCAOB members must be directed and supervised at some level by the SEC with respect to all of their significant statutory responsibilities. With respect to the PCAOB's investigative and enforcement authority, SOX did not provide for SEC control. And I'd say that notwithstanding the D.C. Circuit's insistence, Congress certainly did not ensure that the PCAOB's investigative and enforcement powers would be "subject to check by the Commission at every significant step."

Gillian E. Metzger: I think Tuan's emphasis on considering the implications of finding the Board members to be principal officers is important—and that doing so suggests that the Board's powers to investigate and initiate enforcement proceedings are very unlikely to suffice to create principal officer status.

As an initial point, I take a different view from Donna of the SEC's powers in the investigatory context. The SOX Act requires the Board to promulgate rules to govern its investigations and disciplinary proceedings, and those rules can only take effect upon approval of the SEC. The Board needs to apply to the SEC for a subpoena to compel document production or testimony. Moreover, the fact that the SEC has the broad powers previously outlined over the Board's budget and jurisdiction as well as de novo review of any sanction means that the Board is going to be very attentive to the SEC's concerns about how the Board undertakes enforcement.

But in addition, discretion over initial investigations and enforcement is hardly unique to the Board. Many administrative agencies are structured on the assumption that lower level personnel may make such initial decisions, subject to de novo review at the agency level. Thus, for example, the SEC is authorized to delegate any of its functions to a single commissioner, division, employee, or ALJ, subject to subsequent opportunity for review by the full commission. See 15 U.S.C. §78d-1. This model of initial decision by lower level officials, often ALJs, subject to subsequent de novo review at the agency head level is famously incorporated in the Administrative Procedure Act as the basic model for formal agency adjudication. See 5 U.S.C. § 557(b). If Donna's right and the ability to make initial discretionary policy choices about enforcement is enough to render an executive official a principal officer, notwithstanding that higher level officials can review any final decision emerging from the proceedings, then a substantial number of executive officials thought to be at most inferior officers become principal officers—and correspondingly a vast array of administrative arrangements become unconstitutional. I find it very implausible that the Constitution mandates any such result.

It also merits noting here that the SEC's ability to exercise subsequent de novo review means that decisions by the Board are subject to greater review than decisions by the judges of the Coast Guard Court of Criminal Appeals found to be inferior officers in Edmond; those judges' decisions were subject to review by the Court of Appeals for the Armed Forces only for whether "there is some competent evidence in the record to establish each element of the offense beyond a reasonable doubt." 520 U.S. 651, 655 (1997). Similar limited review was present in Freytag. Another important feature is that public accounting firms have a right to appeal Board sanctions or inspection reports to the SEC. This gives the firms a mechanism for forcing the SEC to actually exercise the oversight powers it is granted, and marks a difference from other arrangements. For example, although US attorneys have been held inferior officers because of the broad array of supervisory authority the Attorney General is formally granted over them, only in rare instances (e.g., wiretaps) is AG approval of US attorney decisions required.

Donna M. Nagy: I'd like to build on Gillian's point that administrative agencies are structured on the assumption that lower level personnel may make initial decisions regarding investigations and enforcement, subject to de novo review at the agency level. If the PCAOB's investigative and enforcement decisions were subject to de novo review by the SEC, then the PCAOB could well fit this model. But SOX did not provide for SEC review of PCAOB investigative and enforcement determinations. Rather, SOX authorized the PCAOB's five members ("the Board”) to make these decisions, and subjected only the Board's final disciplinary actions to de novo review by the SEC. See SOX Section 107(c) ("Commission Review of Disciplinary Action Taken by the Board”). 

The Board thus functions in both a prosecutorial and adjudicative capacity. Based on an investigatory record presented by its staff, the Board acting in its prosecutorial capacity decides who does (or does not) have to defend against specific charges, and what those charges shall be. The Board in its adjudicative capacity then decides whether the facts establish that the charged rules, standards, or laws have been violated, and determines the appropriate sanction if violations are found. SOX provides for SEC review only at this adjudicative stage. And the SEC's decision is itself subject to review by a federal circuit court.

No inferior officer or employee at the SEC possesses this degree of prosecutorial discretion. To be sure, the Director of the Enforcement Division exercises significant discretion in determining who gets investigated and for what. But that Director serves at the pleasure of the SEC Chairman and Commissioners, and his discretion is checked by the fact that he can be asked to step down at any time. Moreover, he and his staff cannot initiate an administrative proceeding (or federal court action) against any entity or person. Rather, the Enforcement Division recommends enforcement action, and SEC Commissioners decide whether such action is warranted based on the investigative record presented by the staff. ALJs at the SEC generally conduct administrative proceedings that require a hearing, but ALJs oversee only those proceedings that the SEC Commissioners have authorized. ALJ rulings and findings are subject to de novo review by the SEC Commissioners, and the SEC's decision is subject to review by a federal circuit court. 
 
One final thought on subpoenas – while Gillian is correct that the PCAOB needs to apply to the SEC for a subpoena to compel document production or testimony, subpoenas are only necessary when a third party outside the auditing industry refuses to cooperate voluntarily. SOX Section 102(b)(3) provides that as a condition of registration, accounting firms and associated persons must consent to cooperate with any PCAOB request for testimony or documents. (We can save for another debate the question of whether this mandatory condition of cooperation raises constitutional questions – though I should note that PCAOB rules and policies provide for assertions of privileges, including the 5th Amendment privilege against self-incrimination).

Peter M. Shane: This hugely thoughtful and nuanced debate on the subordinate status or lack of it of the PCAOB is, for me, a healthy reminder just how completely divorced from any important principle of checks and balances is the textualist squabbling over who is an "inferior" officer. The stakes here, we should recall, are whether PCAOB members must be appointed by the President with advice and consent or whether Congress may vest appointment authority in some other agency of the executive branch. In operational terms, I can't see this making one whit of difference to the actual performance of the agency – that is, I don't see that a presidentially appointed PCAOB is likely to be any more or less effective, well managed, well qualified, democratically responsive, attuned to the rule of law, or virtuous in any other way we expect of administrative agencies. Am I missing something? As I said in my opening salvo, I see much less at stake in the inferiority issue than in the removal issue, but if there is a plausible case for regarding PCAOB members as "inferior," my personal inclination would be to leave Congress with discretion over the appropriate mode of appointment.

Donna M. Nagy: Two brief responses to Peter's post:

First, this isn't a case where the Supreme Court can simply defer to Congress's judgment that newly created federal officers are "inferior" rather than "principal" within the meaning of the Appointments Clause. The Sarbanes-Oxley Act states specifically that that the PCAOB "shall not be an agency or establishment of the United States Government" and that no member of the PCAOB shall be deemed an "officer or employee" of the Federal Government. As I see it, Congress was operating under the misimpression that PCAOB board members would be private sector officials akin to the officials who head NYSE Regulation and the NASD (now FINRA). However, because of the precedent in Lebron, the Supreme Court now has to place the PCAOB into an organizational chart of the federal government, at least for purposes of constitutional law.
 
Second, in terms of practical ramifications with respect to the PCAOB, I see the Appointments Clause issue as more important than the removal question. Presidential appointment and Senate confirmation of PCAOB members would likely raise the public stature of the PCAOB. And increased public attention would likely prompt questions about why, exactly, the PCAOB is a "private" corporation exempt from FOIA, the APA, and other federal statutes. Given its congressionally-created mission, responsibilities and powers, why isn't the PCAOB a part of the federal government for all purposes? Is any public policy served apart from private sector salaries? The removal/separation of power issue may have more far reaching consequences for administrative law and the operation of other federal agencies. But the Appointments issue, in my view, will have direct consequences for the PCAOB. I'm very interested in the thoughts of others on this.

Peter M. Shane: Donna's is a very thoughtful response, although I don't think presidential involvement in the appointments process is always sufficient to raise public consciousness in the manner for which she hopes. For the record, however, I do want to throw in the perhaps obvious point that, even if the current structure of the PCAOB is constitutional, its design might be wrongheaded.

Gillian E. Metzger: On future implications: I think Donna is correct that the claim the Board members are principal officers, if upheld by the Court, has the greatest significance for the PCAOB; it is hard to see how the PCAOB can continue in that case absent new congressional legislation, whether that legislation provides for presidential appointment or expands the SEC's oversight still further. By contrast, as an earlier blog noted, if the Court upholds on the double for cause removal issue, then a logical approach for the Court to take, particularly given the SEC's broad oversight powers and the Court's recent remedial jurisprudence, is simply to sever the for cause removal protection for Board members.

I remain unpersuaded, however, that the degree of prosecutorial discretion that Board members exercise is of such a nature to render them principal officers. To be sure, they do enjoy some discretion here, and that discretion and their other policy-setting responsibilities suffices to make them government officers as opposed to employees. But it does not, to my mind, afford them the degree of final, unreviewable policymaking authority that Edmond identifies as distinctive of principal officers. Prosecutorial decisions that result in sanctions are clearly reviewable by the SEC, the rules and procedures governing investigations need to be approved by the SEC, and on top of that nothing the Board does impairs the ability of the SEC to itself regulate and take enforcement action against accountants and auditors for purposes of enforcing the securities laws.

Finally, on the value of presidential appointment, Donna may be right that presidential appointment and senate confirmation might raise the Board's public profile, however I am much more doubtful that it would provoke public discussion of the Board's status as a private entity, its exemption from the APA, etc. Those questions of institutional design do not strike me as the types of questions that have a lot of public traction—though the Board members' private sector salaries might well generate some popular outrage. More importantly, I share Peter's view that, as a practical matter, presidential appointment or not here will have very little impact on how the Board functions, given the degree of SEC oversight—except that given the delays and burdens of the appointments process, a requirement of presidential appointment might well create more leadership gaps at the Board and thereby impair its effectiveness as a regulator.

Tuan Samahon: I want to return to an earlier thread in this friendly exchange. Although I question Donna's ultimate position that the PCAOB members are principal officers, I do not think that the interpretation of "inferior" as subordinate is far removed from any separation of powers concern.

Peter: Correct me if I'm wrong, but I don't understand you to be saying that the consequence of the subordinate interpretation of "inferior" is unimportant. After all, it means that post-Edmond, cross-branch appointments under the Excepting Clause are unconstitutional – one can scarcely be subordinate to an appointing officer from another branch of government without raising serious separation of powers concerns. Instead, I understand Peter's point to go to construction (i.e. doctrinal implementation), not interpretation, and the question of what it means to be subordinate. This interpretation-construction distinction is important because, here, I think the principal dispute is not about the interpretation of "inferior" as "subordinate" but the construction of that interpretation. For an originalist employing this distinction, the Court ought to police aggressively interpretations of the Constitution that are faithless to the written document, but as to the implementation of the subordinate principle, we ought to be cautious about judicial review that displaces congressional and judicial constructions of what it means to be subordinate.

In my view, Justice Scalia's opinion in Edmond did two different things. First, it offered an interpretation of inferior as subordinate. The historical and textual evidence of that interpretation is quite strong (including contemporaneous Virginia ratification statements by Madison on this precise point). Second, it offered a construction of that interpretation. The Court said that to be subordinate means to be "directed and supervised at some level by others who were appointed by Presidential nomination with the Senate's advice and consent." (One minor point: Scalia erred when he said that the officers must be appointed by someone appointed with presidential nomination with the Senate's advice and consent. After all, the President alone can be given the power to appoint inferior officers, and the President is NOT appointed with the Senate's advice and consent. I take, then, that Scalia would revise his position to be that the inferior officer would have to be directed and supervised at some level by the appointing officer.)

Did Scalia intend by his construction to articulate a standard or a rule? Scalia may believe this is a rule, i.e. find a hierarchically superior appointing officer who "directs and supervises at some level" the officer in question and that officer is inferior for Excepting Clause purposes. It, however, is not so clear to me that this construction is as plain as it is being characterized. Does the Edmond opinion's language "at some level" mean to describe the manner of management employed (e.g. high-level direction, day-to-day direct oversight, etc.) or does it mean to describe the hierarchical description of where the supervising officer may be located (e.g. 1-layer of supervision, 2-layers of supervision, etc)? If it is the first version, i.e. the manner of management employed, it sounds like a fairly deferential standard and I don't think a challenge would carry the day.

Like Peter, I think there is much more at stake with the removal question. I will post a separate remark shortly on removal.

Peter M. Shane: I do not think that inferior-as-subordinate voids cross-branch appointments, as long as one is subordinate to a removing official. As it happens, independent counsel had some formal accountability to both the AG (and through the AG, to the President) and to the appointing panel, although the latter would surely have been exercised very, very rarely, even if the law had been permanently reauthorized. I would say that, "As to the implementation of the subordinate principle, we ought to be cautious about judicial review that displaces congressional determinations of what it means to be subordinate in a particular administrative context."

Tuan Samahon: The question boils down to subordinate to whom? In the context of the appointments clause, it is the appointing officer, not the removing officer. The majority in Morrison v. Olson recognized the difficulty with cross-branch appointment under the Excepting Clause and cross-branch removal, if the good cause provision were interpreted to require the Special Division of the U.S. Court of Appeals for the D.C. Circuit’s approval. I gather that is why the Court: (1) rejected the subordinate analysis for the appointments issue (because to do so would have meant invalidating Morrison’s appointment as an executive officer, who may not appropriately be subordinate to a judicial officer); and (2) abused the canon of avoidance to settle on an interpretation of the Ethics in Government Act that would have read the removal provision as wholly intrabranch, all while acknowledging the problems of cross-branch removal ("it is the duty of federal courts to construe a statute in order to save it from constitutional infirmities…and to that end we think a narrow construction is appropriate here. The termination provisions of the Act do not give the Special Division anything approaching the power to remove the counsel while an investigation or court proceeding is still underway – this power is vested solely in the Attorney General. As we see it, ‘termination’ may occur only when the duties of the counsel are truly 'completed' or 'so substantially completed' that there remains no need for any continuing action by the independent counsel. It is basically a device for removing from the public payroll an independent counsel who has served his or her purpose, but is unwilling to acknowledge the fact. So construed, the Special Division's power to terminate does not pose a sufficient threat of judicial intrusion into matters that are more properly within the Executive's authority to require that the Act be invalidated as inconsistent with Article III.").

What this comes down to is whether Morrison v. Olson continues to be valid. I don’t think it is post-Edmond. If you take the subordinate interpretation and apply it to Morrison, the IC was not subordinate to the appointing judicial officers, or if she was, that would create another separation-of-powers problem.

Yes, there was a question of whether the IC was subordinate to the Attorney General, but that was in the context of the other challenge Olson raised, i.e. the unitary executive issue of whether the President was in control of officers exercising purely executive power.

Martin Flaherty: The exchange has been not just friendly, but useful and informative, especially in its close consideration of the Appointments Clause issue. Apart from the participants, one reason this has probably been true is that this aspect of the case does not as readily lend itself to the usual battle between those seeking checks on Executive authority and advocates of the so-called, "unitary" Executive.

Remaining friendly, I would take a couple of steps back and approach the case from a somewhat more broad-brush perspective. My perspective is principally informed by how the Founding mostly doesn't, but sometimes does, address these issues. Earlier I'd argued, and not surprisingly Peter concurred, that original understanding casts little light on the specifics of either the appointments or removal issues. The scope and complexity of the Executive Branch is simply too radically different to permit much traction for Founding views on superior v. inferior officers. The materials are even less helpful with regard to removal, as the lack of any text at all indicates.

So what guidance might the Founding cast? First, as I noted, the text runs out very quickly, and even when it's there, it often is of little help (see e.g. the Appointments Clause in superior/inferior). Second, original understanding does suggest fairly wide agreement on three functional values. The branches needed to be balanced to prevent tyranny. They needed to be sufficiently independent for efficiency. Each, finally, had a claim to accountability to the people, including in insulated fashion, the judiciary. Third, and not necessarily finally, evolving custom through interbranch accommodation and case law provides a further source for resolving specific issues.

The question then is how these interpretive principles apply here. On appointments, I would opt for treating the PCOAB board members as inferior, though like Peter I am not convinced anything momentous turns on the point. Just to focus on the balance point, whether the President and Senate appoint, or whether Congress vests in the President or heads of departments, does not seem to matter much. Either way, Congress is in some way involved. I take the point treating them as superior would heighten their profile, but this seems more a policy benefit than a constitutional imperative. Though I claim no expertise, it also seems that the PCOAB is more analogous to bodies whose members have been deemed inferior, so custom seems to cut in this direction as well.

By comparison, removal appears to me to be more clear-cut and more important. With regard to original functional values, permitting Congress to provide some check on administrative and even "pure" executive officials (a la Morrison), better comports with the idea of balance in the age of a gargantuan Executive. Nor do I see any particular cost in efficiency, given the constitutional availability of "for cause" dismissal. Custom here too seems to confirm this result.

Finally, the implications of the case I believe are also related to the type of reasoning the Court employs. If it concentrates on the quirks of the PCAOB, then the effects obviously will be minimal. If, however, it employs broad formalist or functionalist analysis (sort of as I've done!), then we are likely to see the jurisprudence lurch away from Morrison and back toward Myers.

Tuan Samahon: On the question of removal, is there a way to save SOX and still preserve the traditional unity of the executive, at least that unity allowed since Humphrey's Executor? Here's an attempt.

Assuming, arguendo, that the Court ought to be deferential to a congressional decision to restrict presidential control of executive officers, is it really even clear here that Congress has erected double barriers to presidential removal? Gillian's amicus brief points out that no statute explicitly provides the SEC commissioners with good cause tenure. To be sure, there is a statute that creates a fixed term, but that does not indicate tenure removable only for cause. For example, magistrate and bankruptcy judges are appointed for fixed terms, but separate statutory provisions specify their removability for cause. The fact of a fixed term might be read, as in agency law for an agent with a fixed contractual term, as merely establishing the quantum of injury (i.e. a basis for damages calculation in, say, the U.S. Court of Federal Claims), not as making it impossible for the principal to dump an agent he no longer wants. As in agency law, the principal (here, the President) retains a power to remove the agent, but might have to pay money damages to the agent. If that is the case, then the President is separated from the PCAOB by only 1, not 2, layers of good cause protection. The Court could invoke the canon of constitutional avoidance to dodge answering the constitutional question of double removal and conclude that it's unclear that any statute protects the SEC commissioners. After all, the court of appeals cases about the SEC's status appear to mostly assume or state in dicta that the SEC is an independent agency. If there is a Supreme Court opinion squarely addressing the question of whether the SEC is indeed an independent agency, I'm unaware of it. Therefore, PCAOB could be upheld on the rationale of Humphrey's Executor (the FTC did explicitly provide statutory good cause tenure) WITHOUT testing the permissibility of a double layer of good cause protection. On the removal question, SOX would survive (for now) because the PCAOB would turn out to be really only at 1-remove from presidential control. Alternatively, the Court could read the President's power to remove SEC commissioners for "good cause" broadly enough to create a here-and-now subservience, per Bowsher v. Synar.

Does the Court's opinion in Wiener v. United States preclude such an approach to *SEC commissioner* removal (again, NOT PCAOB member removal)? Wiener is readily distinguished. The SEC commissioners are unlike the judges of the War Claims Commission, a body with an "intrinsic judicial character" that adjudicated compensatory monetary claims, which is more akin to the U.S. Court of Claims. Moreover, Wiener sued only for money damages in the U.S. Court of Claims, not for injunctive relief to force his reinstatement. The Court never forced President Eisenhower to reappoint Wiener. In that way, the fixed statutory term in that case created nothing more than a right to monetary damages for those with terms fixed by statute. None of the rest of the language implying congressional intent to create tenure against removal is necessary to the holding, and even if it were, it's doubtful the Court would extend Wiener beyond its facts.
 
Thus, SOX is saved (for now) but the power of the President to control executive officials – at least at the level recognized since Humphrey's Executor – is preserved.

Peter M. Shane: Tuan's points are elegant, but the canon of constitutional avoidance is typically invoked as a tool of judicial modesty. Declaring that the SEC, which is universally regarded as an independent agency – indeed, Congress so defines it in 44 U.S.C. sec. 3502(5) and Presidents have insulated the SEC from OMB rulemaking review precisely on this ground – to be something other than an independent agency would hardly qualify as modest. I also worry about making too much out of the nice distinctions among the removal provisions or lack of them in independent agency statutes. The FCC, for example, also has no express protection against at will discharge.

With regard to removals, I think it important not to take one's eye off the real ball. The "ball" is the President's alleged constitutional entitlement (or lack of it) to direct the exercise of whatever policy discretion Congress confers on administrative officials, whether or not Congress has purported to limit the President's authority in this respect and regardless of the subject matter at issue. In my view, the President has no such blanket constitutional authority beyond the purview of those subjects on which he has direct Article II authority – which do not include the regulation of securities markets. Trimming Humphrey's Executor in any respect would be viewed, I am sure, as suggesting that such inherent and unregulable presidential authority does exist, which would be legallyerroneous and institutionally dangerous.

Tuan Samahon: My approach of plausibly interpreting the SEC organic act to avoid the constitutional question is more modest than might be supposed because it reinforces the democratic process. If it was Congress's considered judgment that it meant to provide for good cause tenure for commissioners at the SEC or elsewhere, but it just failed to do so, it could always still legislate it. This is not a requirement of magic words in the nature of a plain statement; it's a requirement of ANY statement. Shockingly, there's been none. Of course, if Congress did grant the SEC commissioners good cause tenure, that would then squarely present the question that is supposedly now presented by PCAOB, viz., whether double layers of safeguard against presidential removal impermissibly attenuate presidential control over officers exercising executive power.

If Justice Kennedy and a majority were looking for a non-constitutional ground of decision but wanted to signal this arrangement might be constitutionally problematic if pressed, this approach might be the minimalistic way to do it. It allows the Court to signal clearly its inclination on the constitutional question but gives Congress an opportunity to respond legislatively to make its intent clear. Assuming, arguendo, that a majority is unprepared to overrule Humphrey's Executor, it still permits one layer of removal safeguard, but forces Congress to think long and hard about whether it really wants to create two layers of removal safeguard. It could also give Congress the chance to make that removal safeguard severable from the rest of SOX.

That 44 U.S.C. 3502(5) calls the SEC an "independent regulatory agency" (relating to paperwork and publication) does not create SEC commissioners protected against presidential removal. The SEC can be said to be "independent" for many reasons without any necessary implication that Congress intended to limit the President's power to remove officers, e.g., one might consider them "independent" because no more than three SEC Commissioners may belong to the same political party or because the commissioners serve for 5-year staggered terms. To be sure, on my view commissioners could be removed early and for a broad array of reasons, but there are good political reasons why the President might be disinclined to go that route, even if constitutionally capable. In that sense, Congress might term the SEC an "independent regulatory agency" without necessarily implying the President has no power to remove the SEC commissioners.

Ultimately, the question will return to whether our Constitution requires that the President must be able to control other executive officers.