Yesterday, I started to make the case for why the en banc D.C. Circuit Court of Appeals should use PHH Corp. v. CFPB as an opportunity to reject the idea that the novelty of the CFPB’s structure makes it unconstitutional. It turns out, I’ve got a bit more to say on the point…. specifically, 85 law review pages if you're really interested!
Using novelty as a limiting principle, it turns out, has other problems besides the kind of arbitrariness I described yesterday. The panel opinion in PHH also illustrates several other kinds of arbitrariness that are caused by judges relying on a statute’s novelty to conclude a statute is unconstitutional.
Levels of Generality. Using novelty as a limiting principle requires judges to define what the prior traditional practice is. That is, under the reasoning of PHH, courts must identify whether a challenged statute is different from statutes that Congress has enacted previously, and different from statutes that the Supreme Court has upheld previously. Courts must, in other words, identify what the tradition to date has been.
But as numerous Justices and scholars (including Take Care’s own Mike Dorf and Larry Tribe) have explained, traditions can be explained at different levels of generality. A tradition can be defined at a high level of generality, such as “there is a tradition of insulating agencies from Presidential control.” Or a tradition can be defined at a more specific level of generality, such as “there is a tradition of Congress creating multi-member commissions of five to seven individuals who can be removed by the President only for malfeasance.” What’s the right level of generality at which to define a tradition?
That problem was on display in Free Enterprise Fund v. PCAOB, the case that held the PCAOB’s structure unconstitutional. According to the majority in Free Enterprise Fund, the relevant tradition was agencies with a single layer of for-cause removal—that is, the tradition was a pattern agencies led by individuals who are removable by the President only for cause. And, according to the majority, the PCAOB fell outside of this tradition because the PCAOB had a double layer of for-cause removal—the PCAOB was led by individuals who were removable only for cause by a group of officials who were themselves only removable for cause by the President. The dissent, however, defined the relevant tradition as agencies that are insulated from Presidential control. The PCAOB, which is likewise insulated from Presidential control, comfortably fit within that definition of tradition.
In PHH, the problem is not just that there are different levels of generality at which the relevant tradition might be defined. That is a problem, to be sure: Is the relevant tradition “multi-member commissions of five to seven individuals who can be removed by the President only for cause” or is the relevant tradition “commissions led by some number of persons who can be removed by the President only for cause”?
The problem unique to PHH is that the way the panel distinguished the CFPB from other independent agencies suggests that the way the panel elected to define the relevant “tradition” somewhat silly. In PHH, the panel was forced to explain why a single-director independent agency is unconstitutional but a multi-director independent agency is not. According to the panel, the reason is this:
In the absence of Presidential control, the multi-member structure of independent agencies acts as a critical substitute check on the excesses of any individual independent agency head.
In plain English: the multi-director structure will limit an agency’s power because obtaining consensus is hard. As I wrote yesterday, this difference between the CFPB and other agencies does not explain why the President has less power over the CFPB than over an independent agency headed by a multi-member commission.
But this purported difference between single-director and multi-member-commission agencies also calls into question the way Free Enterprise Fund had defined the relevant tradition. Recall that Free Enterprise Fund defined the relevant tradition (that the PCAOB fell outside of) as “independent agencies insulated from Presidential control with single layers of for-cause removal restrictions.” But the panel in PHH reasoned that mechanisms other than presidential removal provide a meaningful and sufficient means to limit the power of and legitimize independent agencies. If that’s right, the relevant tradition in Free Enterprise Fund should not have been “agencies insulated by single layers of for-cause removal restrictions”; it should have been “agencies with sufficient limits on their authority.”
Finally, if the panel in PHH seriously believed that an agency’s multi-member structure provides such a meaningful of limit on the agency’s authority, then the relevant constitutional tradition against which to judge the CFPB is “independent agencies with meaningful limits on their authority.” The CFPB has many meaningful limits on its authority, even if a multi-member commission isn’t one of them.
Defining Novelty. But the arbitrariness in defining the relevant tradition at the right level of generality is not the only way in which PHH’s use of novelty was arbitrary. I’ve (occasionally) been putting air quotes around the word “novelty” when I describe the CFPB. There’s a reason for that, which is that I’m not sure the panel was right to say the CFPB is in fact “new” or different from many other agencies. To begin with, there are the slew of executive agencies that are headed by single individuals (like the Environmental Protection Agency, Health and Human Services, the Department of Education, or the Department of Energy).
There are also other independent agencies that are headed by single individuals. The panel distinguished three of them, the Social Security Administration, the Office of Special Counsel, and the Federal Housing Finance Agency, all on iffy grounds. The panel threw a bunch of distinctions at the SSA (unclear which one mattered): The SSA was originally an executive agency (but has been independent for awhile); the President issued a signing statement when the SSA was made independent; and the SSA doesn’t have unilateral enforcement authority. The panel likewise threw up a bunch of distinctions with the OSC, none of which explained why the President had any more control over that agency than the CFPB—the panel noted that the OSC has narrow jurisdiction and is a (relatively) recent creation (in 1978).
The panel dismissed the FHFA on the ground that it was created too recently (in 2008). Thankfully, the administration has written and signed a brief explaining why the single-director FHFA is perfectly constitutional! (And the reasons the administration gave for why the single-director FHFA is constitutional also apply to the single-director CFPB.)
My larger point is this—how many prior agencies does it take before an agency is not considered “new”? Apparently more than three. Would four work?
Treatment of Precedent. It is also worth pausing over one of the premises of PHH’s reliance on the CFPB’s purported novelty. The panel in PHH had to conclude (and it was apparent from the opinion that it did conclude) that the Supreme Court’s decisions in Humphrey’s Executor v. United States, Wiener v. United States, and Morrison v. Olson were all wrong. On Morrison, the panel wrote:
“[T]he independent counsel experiment ended with nearly universal consensus that the experiment had been a mistake and that Justice Scalia had been right back in 1988 to view the independent counsel system as … unconstitutional.”
(Morrison addressed whether the independent counsel statute was constitutional.)
On Humphrey’s Executor, the panel wrote:
“In 1935, however, the Supreme Court carved out an exception to … Article II by permitting Congress to create independent agencies that exercise executive power.”
The panel elsewhere described Humphrey’s Executor as “notwithstanding Article II.” The panel did not describe Wiener v. United States above the line.
Maybe if you think constitutional decisionmaking should proceed anew issue by issue, the panel’s treatment of precedent might not trouble you. I do not share that view. I think there is a reason for respecting precedent (be it judicial precedent or congressional precedent). The decision in PHH discarded decades of precedent with nothing more than a turn of phrase.
I hope that the en banc court treats precedent as something to be taken more seriously.
The Legitimacy of Novelty. I won't go into the full detail here, but in the article, I outline a litany of valid reasons why Congress might enact a novel statute. Some of those reasons are totally mundane. For example, if Congress hasn't enacted a statute in a given regulatory sphere, the forms of regulation that had been used in statutes to date might not be well suited to the new area of regulation. Other times, Congress might enact a new statute to tackle a new problem, or to address a perennial one that has not yet been solved. In the case of the CFPB, for example, Rachel Barkow has written how Congress created the CFPB with an eye toward precedent to make sure the CFPB's structure was lawful, but that Congress also elected to use a different structure because prior regulatory structures had failed.
Forward and Backward Looking Implications. I wasn't sure where this point should go, so I put it at the end. The point is this: Let's assume that the CFPB's structure is novel, or different from, other independent agencies. So what? Independent agencies contain multitudes. Ricky Revesz and Kirti Datla wrote an article disaggregating indepedent agencies from one another, and showed that independent agencies come in many different varieties. So, if you squint hard enough, it is almost always possible to identify a way in which one independent agency is different from the others. And the difference the challengers have identiifed between the CFPB and other independent agencies isn't a particularly relevant one--that is, it doesn't explain why the President has any less control over the CPFB than the President has over other independent agencies.
The point can be generalized further still: Agencies contain multitudes. In Bureacracy at the Boundary, Anne O'Connell showed that the executive-versus-independent-agency dichotomy itself fails to capture the true variety in the federal administrative state. Agencies come in many different shapes and forms: Some agencies are formally housed in the legislative branch even though the head of the agency is chosen by the President (like the Government Accountability Office); agencies have different structures (like the Federal Reserve Board, whose commissioners are, by statute, removable only for cause, and is led by a Chairman and Vice Chairman); there are agencies that blur the boundaries between the federal government and tribes, or the federal government and states, or the federal government and the private sector.
If the CFPB is constitutionally suspect because it differs from other (independent) agencies, then are all or most independent agencies (which differ from one another) unconstituitonal? Are the many shapes and forms of the federal bureacracy unconstitutional as well? The idea that a federal statute's novelty is a sign that it is unconstitutional did not catch on until the 1990s. What if it had caught on, or been used, a bit earlier--say in the 1930s, or 1960s? These implications should give one pause before accepting the constituitonal challenge to the CFPB. I hope that the en banc D.C. Circuit rejects that challenge, and that the Supreme Court does as well (should it come to that).
Disclosure: I’m not sure if this does not make me neutral, but just in case: I wrote an article that argues a federal statue’s novelty is not a sign that a statute is unconstitutional. Two amicus briefs, one that was filed by scholars of financial regulation, the other by scholars of separation of powers cited the article in the en banc proceedings before the D.C. Circuit.
Follow Leah on Twitter @LeahLitman