The PHH Case Part III: Fair Notice

This is the third in the series of posts about the PHH case. To review, the Consumer Financial Protection Bureau determined that PHH had violated section 8 of the Real Estate Settlement Procedures Act by means of a practice in which PHH referred mortgage insurance business only to companies that then bought reinsurance from a PHH subsidiary. The Bureau ordered PHH to disgorge $109 million in payments, apparently all the reinsurance payments received in this practice between mid-2008 and the Bureau’s order. In the last two posts I discussed the possibility that PHH shouldn’t be liable for payments before July 2011 or perhaps even July 2013.

Today’s post is about a different issue, although one that also involves temporal limits on liability. PHH has argued that the practice the Bureau criticized was widespread in the industry, and that industry participants generally thought it was lawful based on certain guidance documents from the Department of Housing and Urban Development. (HUD was the Bureau’s predecessor in enforcing RESPA.) Even if the Bureau now decides the insurance-reinsurance arrangements are improper, PHH says, it’s not fair to make PHH pay for what it did when the practice was regarded as acceptable.

As I noted in my first post, PHH’s challenge to the Bureau’s order is pending with the D.C. Circuit. The panel hearing arguments seemed sympathetic to PHH’s argument, and it wouldn’t be surprising if the court relieves PHH of its liability on this basis. The question for this post is, what happens the next time, in the next industry? The “fair notice” argument may grow less effective over time.

PHH’s argument derives from a principle that a person shouldn’t be penalized unless the person had “fair notice of the conduct” that is “prohibit[ed] or require[d].” This is an old idea, but the Supreme Court has given it renewed vigor in recent years. In FCC v. Fox Television Stations, the FCC changed its policy about indecency standards. Under the new policy, it imposed penalties for several broadcasts that had predated the policy. The networks argued that the broadcasts were not indecent under the prior policy, so it was unfair to penalize them. The Court, focusing on the complicated history of FCC indecency enforcement in the area, noted that the networks had lacked “fair notice” that their broadcasts would be considered improper. The regulatory change was too “abrupt.” In Christopher v. SmithKline Beecham, the Department of Labor filed an amicus brief interpreting its Fair Labor Standards Act regulations to make the position of “pharmaceutical detailer” a non-exempt position. (Thus, FLSA wage regulations would apply for pharmaceutical detailers.) But the industry had treated detailers as exempt salespersons for decades, without objection from DOL. The Court refused to defer to DOL because DOL’s position would entail “massive liability . . . for conduct that occurred well before that interpretation was announced.” Coming after decades of DOL’s evident acquiescence in the industry practice, the opinion says, “the potential for unfair surprise is acute.”

Some language in the SmithKline Beecham opinion almost suggests that an administrative agency should never order a respondent to be liable on the basis of an interpretation announced for the first time in the same proceeding. But it would probably be a mistake to read that language too broadly, because another longstanding principle is that an agency can make its policy decisions on a case-by-case basis, even if that means applying new policies to prior conduct. Even since SmithKline Beecham—and even in cases imposing serious consequences for past conduct—courts have reiterated that an agency has discretion whether to announce new principles via adjudication or rulemaking. Besides, it would go a bit far to insist that an agency must spell out an interpretation in unmistakably clear language before it can hold people to it. The concept is “fair notice,” not knock-down-the-door notice. Or, as FCC v. Fox put it, “a person of ordinary intelligence” should have “fair notice of what is prohibited.”

What irritated the Court in FCC v. Fox and SmithKline Beecham was clearly a sense that the agencies involved had reversed course from positions that the regulated parties had relied on. Each agency had given those companies some reason to think their conduct was lawful, and no hint to the contrary until the adjudications that imposed liability.

Suppose instead an agency announced that it was seriously reconsidering its past position and didn’t know what the decision would be; or had disavowed a past interpretation and said the issue was open for future adjudications. Could the agency then announce a new interpretation (assuming it were reasonable) and enforce it in the same adjudication? Maybe yes, maybe no. At any rate the Bureau may, in the PHH case, have made a disavowal of this sort.

Recall that PHH argued it had relied on past HUD guidance. PHH says its case is like SmithKline Beecham and FCC v. Fox because the Bureau reversed the HUD interpretation. Of course, the Bureau is not HUD. And, given the circumstances of the Bureau’s founding as an agency meant to improve on the regulation of consumer finance that went before, it would not be surprising to see it disagree with at least some of its predecessor agencies’ policies. But, as PHH points out, the Bureau said in 2011 that, pending any further action, it would apply “the official commentary, guidance, and policy statements” of its predecessor agencies.

The Bureau also said it would “give due consideration to the application” of “other written guidance” from predecessor agencies, depending on factors including the persuasiveness of the guidance. Presumably, by contrast the statement about “official” guidance, this sentence meant the Bureau would not necessarily apply other guidance documents. Still, before the Director’s PHH decision, one might have wondered whether the Bureau really would depart significantly from its predecessors’ “other” guidance without engaging in some process. One might also have wondered what would distinguish “official” from “other” guidance, because the Bureau’s statement didn’t say. That ambiguity itself might have provided some reassurance; had the Bureau intended to depart abruptly from some category of prior guidance, one imagines, it would have demarcated that category more clearly.

The PHH decision goes some way to disabusing us of these notions. PHH principally cited a 1997 letter from HUD staff, responding to industry inquiries about the lawfulness of “affiliated-reinsurance arrangements.” The Director’s decision criticizes the 1997 letter as internally inconsistent, but at bottom it acknowledges that on the key point—that if the value of payments is no greater than the value of the reinsurance there is no section 8 violation—he simply disagrees. Evidently the Bureau is quite ready to reverse the guidance of one of its predecessors.

Moreover, the PHH decision provides a significant clarification about the difference between “official” and “other” guidance. The 1997 letter was not “official” guidance because it wasn’t published in the Federal Register and therefore didn’t rank as a “regulation, rule, or interpretation of the Secretary” under HUD’s RESPA regulations at the time. The agencies implementing many other federal consumer financial laws had similar concepts in their regulations. So, for those laws, only the formal published interpretations might be official guidance. Anything else, evidently, is “other” guidance. Taking the PHH decision together with the Bureau’s 2011 statement, the Bureau seems to have indicated that a large swath of the guidance its predecessors provided is truly open to reversal.

That may or may not be enough to provide the “fair notice” that FCC v. Fox speaks of. Still, it seems to me it has become riskier to rely on pre-Bureau guidance. PHH may prevail in its “fair notice” argument. After PHH, future respondents might not fare as well.

The PHH Case, Part II: Retroactivity and Liability Exposure Continued

In my last post, I talked about the possibility that enforcement by the Bureau is a change in law that shouldn’t have retroactive effect. That is to say, conduct that predated the change in law shouldn’t be subject to it. The last post assumed that July 21, 2011, was the date of the change; so according to this retroactivity theory, conduct earlier than that date wouldn’t incur liability to the Bureau. (As I also noted, the same conduct might expose a person to enforcement from other corners. For example, the Department of Housing and Urban Development could, on July 22, 2011, have sued PHH on the same theory the Bureau eventually used. No guessing here how that suit would have turned out.)

In this post I’ll talk about an even later date: July 16, 2013. As seasoned Bureau-watchers know, that was when the Senate confirmed Director Cordray. While a confirmation of the agency head would be important for any agency, it was particularly significant for the Bureau because of the staged way in which the agency came into existence and assumed its powers. The Bureau existed as of July 21, 2010, when the Dodd-Frank Act was enacted. For the first year of its existence, it engaged only in managerial operations: hiring employees, preparing budgets, negotiating service contracts, etc. The substantive regulatory provisions of the Consumer Financial Protection Act became effective on a date chosen by the Treasury Secretary, which the statute called the “designated transfer date.” That ended up being July 21, 2011. Also on that date, a range of pre-existing regulatory functions already being performed by other agencies transferred to the Bureau. But there was no Director yet. Under section 1066 of the Act, the Treasury Secretary exercised the Bureau’s authorities “under this subtitle” until the Senate confirmed a Director.

“This subtitle,” the one to which section 1066 refers, is subtitle F. That subtitle describes the transfer of pre-existing functions; it does not encompass the Bureau’s new authorities. There are various ways to read section 1066. One could take it to mean that the Treasury Secretary would be acting “under this subtitle,” but exercising all the Bureau’s authorities. Or one could accept that “under this subtitle” modifies “authorities,” so that the Treasury Secretary could only exercise subtitle F authorities. I’m not sure I agree with that second reading. But the Inspectors General of the Treasury Department and of the Federal Reserve adopted it, and the Bureau has not to my knowledge disagreed. (On April 22 this year an administrative law judge opined that under section 1066 the Treasury Secretary was able to exercise all the Bureau’s administrative enforcement authority. However, the reasoning seems dubious, and it seems unlikely the Director would adopt it. The opinion observes, correctly, that section 1061(a) defines “consumer financial protection function” to include “all authority to . . . issue orders” under “any Federal consumer financial law.” But section 1061(a) doesn’t actually confer any authority; the only powers that section 1061 actually provides are those consumer financial protection functions that already existed at predecessor agencies.)

So it seems reasonable to assume that under the prevailing view of section 1066, the Bureau was unable to exercise the new authorities created by the Consumer Financial Protection Act until it passed the section 1066 trigger. One could also debate when that happened, but for illustrating the ideas in this post I’ll assume the trigger is exactly what section 1066 purports to say: Senate confirmation.

So, on July 15, 2013, the Bureau could sue to enforce RESPA section 8—that was a transferred authority under subtitle F—but it couldn’t enforce through an administrative proceeding. HUD never had the option of administrative proceedings, and the Bureau’s ability to use them comes from subtitle E. In the last post, I identified two ways in which Bureau enforcement could constitute new law that shouldn’t operate retroactively. First, the Bureau is a new plaintiff with priorities and incentives different from its predecessor agencies. Second, and really more importantly, the Director as an adjudicator is different from a court. Following through that idea from the last post, if the Bureau commenced an administrative enforcement proceeding on July 17, 2013, that proceeding oughtn’t to cover conduct from before July 16. (At this point it might be worth repeating a caveat from the last post. There are arguments on either side of this issue, and I don’t know how it would really come out.)

Defendants and respondents in Bureau enforcement cases have been quite alert to the significance of section 1066. But many of the arguments have not quite hit the point straight on, in part because they are excessively constitutionalized. Respondents and defendants have argued, for example, that the Bureau didn’t have standing under Article III until it had a Senate-confirmed Director; or that the Appointments Clause precluded the Bureau from exercising new authority until the Senate confirmed an agency head. Arguments like these are a bit discomforting, because it’s not uncommon for this or that agency in the federal government to be run by someone filling a vacant seat on an acting basis. If some aspect of the Constitution would preclude the Bureau from doing some of its work while it lacked a Senate-confirmed head, wouldn’t the same argument block other agencies too? As a constitutional matter there wouldn’t seem to be anything different about the Bureau just because it was new.

As a matter of statutory interpretation, by contrast, the newness of the Bureau was perhaps quite important. Congress chose to phase in the new agency in several steps, beginning (as explained above) with the first in which the Bureau existed but had no regulatory authority at all. It would be within the prerogative of the legislature, I’d think, to set things up so that for some time the Bureau could only exercise the existing authorities, and then eventually could (among other things) enforce all federal consumer financial laws by administrative order. The question is simply one of statutory interpretation: Is that in fact how the Dodd-Frank Act works?

To answer that question, how to read section 1066 is obviously quite important. Set against that, the fact that various subtitles besides subtitle F were “effective” as of the designated transfer date in 2011 seems less significant. To be sure, various provisions in subtitle C made certain conduct unlawful as of that date. And it would be odd to interpret the Dodd-Frank Act to immunize that conduct unless and until the Bureau received a Senate-confirmed Director. But it didn’t; under section 1042 states were able to sue immediately on the transfer date; and they could obtain remedies provided by section 1055 (in subtitle E, the general enforcement subtitle). The fact that the section 1066 trigger depended on administrative action (nominating a Director) also doesn’t seem too important. After all, the Consumer Financial Protection Act made the effective date of many of its provisions depend on administrative action, namely the Treasury Secretary’s choice of the designated transfer date.

The Director did eventually ratify the Bureau’s pre-confirmation actions. But that also doesn’t necessarily resolve the issue. The prerequisites for a ratification are generally that the person had the authority for the act at the time of ratification, and would have had the authority had she been in place when the act originally occurred. One purpose for these requirements seems to be to protect individual rights; “[t]he intervening rights of third persons cannot be defeated by [a] ratification.”  One would think the rules for ratification would also protect against retroactivity.  If Bureau administrative enforcement of a given law is indeed a change that shouldn’t apply retroactively, it’s not clear that the Director could simply ratify administrative proceedings from before his confirmation.

The bottom line of the arguments here would be: For many federal consumer financial laws, the Consumer Financial Protection Act did not permit the Bureau to engage in administrative enforcement until it had a Senate-confirmed Director; when that point came, the Director became a potential decisionmaker for administrative enforcement proceedings; that change in decisionmaker is a substantial change in law that should not apply to conduct from before the confirmation. So if the Bureau wants to impose liability for conduct before July 16, 2013, it needs to sue in court.

In the PHH case, that conclusion would have erased all but perhaps $2 million of the $109 million in disgorgement that the Director ordered. We can expect respondents to take serious effort to make some version of this argument work.