The Trump Treasury Department's Dodd-Frank Act report spends more pages on the CFPB (including mortgage regulation) than on any other issue. There's a whole bunch of blog posts that one could write about the Treasury report, but I want to limit myself here to one item that has long been on the GOP/industry complaint list about the CFPB: that its power to proscribe "abusive" acts and practices is a problem because the term "abusive" is novel and undefined and that this creates uncertainty that is chilling economic growth. Total hooey. The Treasury's report is a lazy document is totally unconnected to the realities of how the CFPB has operated. It's a shame that some commentators are buying into it.
Here's the story of the "abusive" power in a nutshell: it's the dog that didn't bark. The CFPB's critics have been complaining about the vagueness of the "abusive" power ever since the Dodd-Frank Act was in the legislative process. Those arguments didn't hold a lot of water then because the term is defined by statute and has a history (namely HOEPA, the FDCPA, the Telemarketing Sales Rule, and the FTC's interpretation of "unfair" from 1962 to 1980), and the codification of "unconscionability" in the Uniform Consumer Credit Code. But we now have the advantage of six years of CFPB enforcement activity to understand how the agency has used this power and what it means. Unfortunately, it seems that no one at Treasury bothered to look through the CFPB's enforcement actions to see how the agency has actually used its power to prosecute "abusive" acts and practices. I did. Here's the two things that stand out.
First, the CFPB has been very sparing in alleging that acts and practices are "abusive". The CFPB has brought around 185 enforcement actions to date. Only 22 of these (less than 12% of all enforcement actions) have included counts alleging "abusive" acts and practices. In all but one instance in these 22 cases, the very same behavior alleged to be "abusive" was also alleged to be "unfair" and/or "deceptive." Unfair and deceptive are not new standards. They have been around in the FTC Act since 1935. While these standards weren't applied to banks for half a century (Regulation AA was from 1985), no institution, bank or non-bank, should be wholly surprised at what might be alleged to be unfair or deceptive. And indeed, when the CFPB has brought unfairness charges, they have generally been in situations in which there is no consumer benefit whatsoever from the practice (e.g., Wells Fargo false accounts). What this means is that the CFPB has not actually been surprising anyone when it has alleged "abusive" acts and practices because to date, the "abusive" power has been little more than a belt to go with the suspenders of "unfair and deceptive".
Second, the behaviors alleged to be abusive are almost all in the context of pre-existing customer relationships. (I include in this things like private student loans through the student's school.) This is the angle about consumer finance that is almost always ignored. The real problems in consumer finance are less about consumers getting snookered into a product in the first place than by financial institutions taking advantage of existing consumers by changing terms, applying undisclosed fees, or, my favorite, Citizens Bank's "[we] keep the change" policy. In other words, "abusive" is getting applied to function as a publicly enforceable duty of good faith and fair dealing, an implied term in all contracts.
All of this suggests that there's really no crisis of uncertainty about what is "abusive". To suggest, as the Trump Treasury Department's Dodd-Frank Act report does, that "Without meaningful standards that provide fair notice, many consumer financial firms are reluctant to innovate or offer new financial products or services," is utterly unsupported, and is really the result of lazy analysis. The types of behavior that the CFPB has targeted are not behaviors that anyone would think are OK: collecting debts that are unenforceable under state law or requiring servicemembers to litigate debt collection suits in a distant and inconvenient forum with which they have no connection, resulting, of course, in default judgments. That's why you don't see any examples ever cited of legitimate business behavior getting improperly tagged as "abusive." And the fact that almost all of the cases deal with the treatment of consumers in existing relationships with the financial institution means that the "abusive" cases have nothing to do with the offering of "new financial products or services." They're all about mistreatment of consumers who bought into existing financial products or services. And if innovation is the concern, well, there's the CFPB's Project Catalyst. Oh yeah, that's not mentioned in the Treasury report. Treasury was too busy citing the Congressional testimony of the lawyer for PHH in its constitutional suit against the CFPB to actually look into the operations of the agency.
The Home Mortgage Disclosure Act of 1975 is a key piece of fair lending legislation. It requires mortgage lenders to report data on loan applications and loans funded that enables both government and private groups to monitor lending patterns for violations of the Fair Housing Act and Equal Credit Opportunity Act (as well as state fair lending laws). In 2015 the CFPB adopted a new HMDA rule that would expand the number of data fields collected by some 25 fields, effective Jan. 1, 2018. This is being decried as an unreasonable burden on small institutions and a bipartisan collection of Senators on the Senate Banking Committee have proposed a bill that would exempt financial institutions that made less than 500 open-end loans or 500 close-end loans in each of the previous two years from the new HMDA reporting requirements.
There's no question that the new HMDA requirements add something to financial institutions regulatory burden. But a look at what these requirements are shows that the burden is really de minimis. It's not going to make-or-break a small financial institution. Below is a list of all 25 new data fields. As you will see, after each one I have indicated whether it is data that is already required for the TILA-RESPA Integrated Disclosure (TRID) or would normally be in a loan underwriting file. If it is in either, then it is simply a matter of having adequate software to plug that data into HMDA reporting. Asking a bank to have integrated mortgage underwriting and reporting software doesn't seem like an unreasonable request, but none of this is stuff that should take very long to do even by hand-entry of data (something I've done plenty of). I've dotted all my i's and crossed my t's here, but the bottom line is this. Almost every piece of information required under the new HMDA rule is already being collected by the lender for either its own underwriting purposes or for compliance with other regulatory requirements. In other words, this just ain't a big deal. My guess would be that the total additional compliance costs is a few thousand dollars per year for this.
The CFPB itself estimates (see p. 66308) per the Paperwork Reduction Act requirement that for truly small banks total HMDA compliance costs (which includes existing costs) will be between 143 and 173 hours of time annually. Even at $100/hr (which is far more than a compliance staffer at a small bank makes), this would total, at most, $17,300 annually. Around half the fields are new, so we're looking at around $8,650 annually in additional costs for small banks as the high-end estimate. So this leave me wondering why the pushback against the new HMDA rule.
Am I missing something here? This just doesn't seem to be a game changer for small financial institutions, and it will cause some serious damage to HMDA data in some communities and even some entire states in which large financial institutions don't have much of a presence.
- Street address of the collateral property. Required for TRID. 12 CFR § 1026.37(a)(6); 12 CFR § 1026.38(a)(3)(vi). It's also going to be in the underwriting file--lenders want to know where their collateral is located.
- Borrower’s age. This is potentially new. It won't be in TRID, but it might be in the underwriting file, at least for a reverse mortgage. Doesn't seem like a hard piece of data to obtain and report.
- Credit score. This is almost certainly going to be in the underwriting file, and if the lender takes adverse action based on the score it has to be produced to the consumer under the Fair Credit Reporting Act, 15 U.S.C. § 1681m(a)(2)(A)
- Total loan costs or total points ands fees. Required for TRID. 12 CFR § 1026.37(f)(4);12 CFR § 1026.38(f)(4). This is also likely to be in the underwriting file.
- Origination charges. Required for TRID. 12 CFR § 1026.37(f)(1);12 CFR § 1026.38(f)(1). Also in the underwriting file.
- Discount points. Required for TRID. 12 CFR § 1026.37(f)(1);12 CFR § 1026.38(f)(1). Also in the underwriting file.
- Lender credits. That will be in the TRID disclosures already, and probably in underwriting file.
- Interest rate. Required for TRID. 12 CFR § 1026.37(b)(1). Also in the underwriting file.
- Prepayment penalty term. Required for TRID. 12 CFR § 1026.37(b)(7). Also in the underwriting file.
- Debt-to-Income Ratio. In underwriting file. Also required for QM Rule compliance, 12 CFR § 1026.43(e)(2)(V)(B).
- Cumulative Loan to Value Ratio. In underwriting file (this gets reported in securitization data).
- Loan term. Required for TRID. 12 CFR § 1026.37(a)(8). Also in the underwriting file.
- Introductory Rate Period. Required for TRID. 12 CFR § 1026.37(a)(10)(iv). Also in the underwriting file.
- Non-amortizing features. Required for TRID. 12 CFR § 1026.37(a)(10)(ii). Also in the underwriting file.
- Property value. This is going to be in the underwriting file. And for some loans an appraisal is actually required by statute under 15 U.S.C. § 1639h.
- Manufactured home secured property type. This is whether the lien is on just the manufactured house or also the land. I'm guessing (but not sure) that this would be in the underwriting file. It's pretty important for a lender to know if it wants to perfect its security interest.
- Manufactured home land property interest. This is about whether the loan applicant owns or rents the land on which the manufactured home is situated. I'm guessing (but not sure) that this would be in the underwriting file.
- Total units. Again, I would expect this to be in the underwriting file, as a 1 family property has a different risk profile than a multi-family. There would also need to be a rider to the mortgage if it covers multiple units.
- Multifamily affordable units. I'm guessing that this is in the underwriting file because it affects the property value, but not sure. Unlikely to be much of an issue for rural lenders, because multi-family is not common in rural America.
- Application channel—should be in underwriting file; a loan that is done through a broker has different risk characteristics than one that isn't, etc.
- NMLSR Identifier. Required for TRID. 12 CFR § 1026.37(k); 1026.38(r)(4). Should also be in the underwriting file for SAFE Act compliance.
- Automated Underwriting system used. This ought to be in the underwriting file.
- Reverse mortgage indicator. In underwriting file.
- Open-end Line of credit indicator. In underwriting file. The lender needs to know because it will necessitate different disclosure forms.
- Business or commercial purpose. In underwriting file.
Jeff Sovern has an excellent new article about arbitration clauses and class action waivers that uses the Wells Fargo fake account scandal as a test case. He also does a monster job knocking down the Johnston-Zwyicki arbitration study. As Sovern points out, the Johnston-Zywicki study makes a big deal out of some data on a Texas bank's voluntary refunds of fees in consumer disputes. But as Sovern observes, Johnston and Zywicki aren't able to differentiate between fees due to bank misconduct and fees due to consumer behavior (account inactivity, overlimit, etc.), much less why the bank refunded the fees in some cases. Highly recommended and relevant in the run-up to the anticipated CFPB arbitration rulemaking.
It was not at all surprising that, for his first (traditionally unanimous) opinion, in Henson v. Santander, the new Justice Gorsuch took on the relatively simple and low-key issue of the definition of "debt collector" in the Fair Debt Collection Practices Act. It was also not surprising that he hewed quite closely to the approach of his predecessor in basing his decision on the "plain meaning" of the words in the statute, complete with grammatical analysis of past participles and participial adjectives (the example adduced, "burnt toast," might describe how the consumer protection industry will view this latest ruling). The FDCPA is as simple as it appears, the Court confirmed: if you're collecting a (consumer) debt owed to someone else, then you're a debt collector; if you're collecting on a debt owed to you, for your own account, you're not a debt collector, even if, as in Santander's case, you bought the debt from the original creditor with the intention of collecting it for an arbitrage profit later. The notion that Congress did not foresee the debt buying industry and its explosive growth when it wrote the FDCPA in the 1960s, and it certainly would have wanted to constrain abusive collections practices by debt buyers as much as by debt collectors was ... wait for it ... a matter for the present Congress to clarify. You can almost see Scalia whispering in Gorsuch's (or his clerk's) ear as the opinion is drafted. Well, at least there's something to be said for predictability.
I'm testifying before the Senate Banking Committee this week about "Fostering Economic Growth: The Role of Financial Institutions in Local Communities". It's the undercard for the Comey hearing. The big point I'm making are that the problem is not one of economic growth, but economic distribution. While the US economy has grown by 9% in real terms since Dodd-Frank, real median income has fallen by 0.6%. That's pretty grim. The gains have all gone to the top 10% and particularly the top 1%.
None of the various deregulatory proposals put forward by the financial services industry have anything to do with growth, and they have even less to do with ensuring equitable growth. For example, changing the CFPB from a single director to a commission or switching examination and enforcement authority from CFPB to prudential regulators shouldn't have anything to do with growth. It's a reshuffling of regulatory deck chairs.
The banking industry has been doing incredibly well since Dodd-Frank, outperforming the S&P 500, for example. You'd never know it, however, from their trade association talking points. It really takes a certain kind of chutzpah to demand the repeal of consumer protection laws and laws designed to prevent the privatization of gains and socialization of losses when you are already doing so much better than the typical American family.
My complete written testimony can be found here.
Statutory interpretation enthusiasts: prepare to nerd out on an issue on which the Court has a cert petition pending. The question involves the federal jurisdictional bar to Medicare challenges. Let’s start with the text:
“No action against the United States, the [Secretary of Health and Human Services, see 42 U.S.C. § 1395ii], or any officer or employee thereof shall be brought under section 1331 or 1346 of title 28 to recover on any claim arising under [the Medicare Act, see § 1395ii].”
Bankruptcy types—and quite frankly, all lucid readers of English—might well think that this jurisdictional bar does not apply to bankruptcy jurisdiction under section 1334. Not so, say a surprising number of courts.
What could possibly justify reading what’s clearly only a bar of federal question jurisdiction (1331) and federal officer jurisdiction (1346) as a bar to bankruptcy jurisdiction (1334) as well? In In re Bayou Shores, the subject of the pending cert petition, the Eleventh Circuit offered an argument based on congressional intent. Until 1984, the precursor to the above-quoted statute did bar bankruptcy jurisdiction. But it did so obliquely, by way of a statutory cross-reference to a subsequently dismantled omnibus jurisdictional grant in the Judicial Code. So for a few decades after the dismantlement, the Medicare bar was a cross-reference to a no-longer-existing passage of the Judicial Code that formerly encompassed virtually all grants of jurisdiction to the district courts, including bankruptcy.
When Congress finally got around to updating the bar in the 1980s, it enacted the text above. Without more, that would just be Congress changing the law from a broad bar cross-referencing all jurisdictional grants to a specific bar covering only two grants. But there was more: Congress did so as part of a package of self-styled “technical corrections.” And of these corrections, Congress wrote a proviso that “none . . . shall be construed as changing or affecting any right, liability, status or interpretation which existed (under the provisions of law involved) before [their effective] date.” To read the amended bar as written would appear to violate this admonition. The Eleventh Circuit (and others) thus read the proviso to justify junking the statute’s text in favor of the perceived congressional intent not to change anything.
What to make of a statute that unambiguously changes the law on the one hand and instructs courts it isn’t to be construed to change the law on the other is, to say the least, a vexing paradox that cannot be addressed here. (If you want to see my thoughts on the matter, see pages 15–24 of the amicus brief I co-authored with Asher Steinberg in a case I argued at the Seventh Circuit two weeks ago.) But here’s the punchline: that heady issue is irrelevant, because the proviso has been misread by the Eleventh Circuit and every other court to read it as precluding substantive change. The full proviso is actually just a banal timing rule, not the paradox-creating mindwarp many courts have thought it was. Here’s the proviso’s complete text, which for some reason has never been fully quoted or addressed in any of the court opinions (hence the need for amici to jump in):
SEC. 2664. (a) Except as otherwise specifically provided, the amendments made by sections 2661 and 2662 shall be effective as though they had been included in the enactment of the Social Security Amendments of 1983 (Public Law 98–21).
(b) Except to the extent otherwise specifically provided in this subtitle, the amendments made by section 2663 [the section containing the amendment to the Medicare jurisdictional bar] shall be effective on the date of the enactment of this Act; but none of such amendments shall be construed as changing or affecting any right, liability, status, or interpretation which existed (under the provisions of law involved) before that date.
Read in context, the underlined language, on which Bayou Shores all but exclusively relies, is nothing more than a mere grandfathering clause, one which locked in vested rights “which existed” in particular cases in the pipeline on the date of enactment, notwithstanding the immediate effect language of subsection (b) (and retrospective effect language of subsection (a)). It did nothing to undermine the substantive text of other parts of the statute. This is plainly apparent for four reasons:
- Congress addresses timing rules in clauses entitled “EFFECTIVE DATES.” It does not bury substantive interpretive commands in such provisions.
- The word “but” suggests that the underlined language is a carve-out from the amendments’ immediate effect expressed in the prior clause. A grandfathering clause provides an exception to immediate effect, and hence naturally links with a “but,” “however,” etc., as here.
- Grandfathering clauses are often drafted in terms of locking in some right or entitlement that vested “before that [effective] date” or “before such [effective] date” or similar timing language; on the other hand, disclaimers of substantive change usually contain more explicit wording like “without substantive change.”
- Locking in a series of four kinds of things “which existed” in some concrete sense on the date of enactment is an incredibly odd, roundabout, and riskily underinclusive way to say that the law itself wasn’t intended to change. But it’s a perfectly sensible way to grandfather vested rights, liabilities, statuses, or “vested” interpretations in the form of opinion letters or interpretations in pending cases.
If this conclusion is right, Medicare hasn’t a leg to stand on; its sole basis for asking courts to take the extraordinary step of atextually reading § 1334 jurisdiction back into the Medicare jurisdictional bar 33 years after Congress amended it to take § 1334 out is nothing more than a grandfathering clause that has been repeatedly misread.* There is a circuit split, so the Court may take this petition, but other cases are still bouncing around the circuit courts in the appellate pipeline, so the Court may wait for a bit. But it will surely have to confront the issue.
* There is a non-jurisdictional question whether Medicare providers must exhaust their administrative remedies before pursuing Medicare in bankruptcy court. Bayou Shores, in petitioning for certiorari, takes the position that no exhaustion’s required when a provider gets into court through a jurisdictional hook, like § 1334, that’s independent of the Medicare Act’s Medicare-specific jurisdictional grant, which comes packaged with various exhaustion procedures. I have argued in my Seventh Circuit briefing that the fairest reading of the statute and case law is that debtors do have to exhaust—subject, however, to the normal excuses for dispensing with administrative exhaustion that the Supreme Court has recognized in a century’s worth of precedents. One of these excuses, the futility of raising an issue before an agency that lacks the competence to address it, has particular force in bankruptcy disputes with Medicare. (Medicare, for its part, denies that the normal excuses apply.) Practitioners fighting off Medicare motions to dismiss could try making both Bayou’s aggressive argument that exhaustion is not required and my more moderate one that exhaustion is required but readily excusable under the traditional tests.
Yesterday I purchased a travel alarm clock through Amazon. This morning, the manufacturer emailed me with instructions for its use, including a very important point about switching the travel lock button off to activate the clock. The clock apparently arrives in the locked condition, which has caused some customers confusion and led them to think that the clock was defective when it was not. The email made me think of a recently published book, The New Handshake: Online Dispute Resolution and the Future of Consumer Protection, by Professor (and former Slips guest blogger) Amy Schmitz and Colin Rule, who is the former Director of Online Dispute Resolution for eBay and PayPal.
The New Handshake surveys that current landscape of online dispute resolution and sets out a blueprint for how the Internet can help consumers worldwide deal with disputes arising from their e-commerce transactions. With more and more consumer transactions moving online (ten years ago, I likely would have purchased that travel alarm clock at the-somehow-still-semi-alive Radio Shack), the book's detailed ideas for how to design an effective dispute resolution system is increasingly important for businesses and for consumer advocates. As Schmitz and Rule note, largely gone are the days when transactions were sealed in person with a firm handshake, and class actions seem less and less effective overall -- which leaves both challenges and space to innovate for business and consumers. For my own interests, two parts of the books stood out.First is data that Rule shares about customer loyalty from eBay's experience establishing its online dispute resolution system. eBay tracked its system's "return on resolutions," calculation of which includes how the resolution process impacts customer loyalty and retention. As a baseline, it calculated that buyers who did not file a complaint (that is, did not use the system) in a 3 month period used eBay about 8% more in the subsequent 3 month period. But buyers who filed complaints that were resolved amicably with the sellers through the system during a 3 month period used eBay 17% more in the next 3 month period. Similarly, buyers who filed complaints that were settled outright during a 3 month period used eBay 14% more in the next 3 month period. Even those buyers who filed claims that were settled in the sellers' favor because the buyers were found to be at fault increased their usage of eBay. The only group of buyers who used eBay less were those who filed claims, but whose claims remained pending for longer than six weeks. (For more details, see chapter 4 of the book.)
Which leads to one of The New Handshake's most interesting and important insights. Consumers want to be heard, not necessarily win (though that's nice) or be paid off every time they have a problem, such as with coupons and gift cards (though those are nice too). This results aligns with research regarding procedural justice in civil and criminal law. People often are more concerned with the fairness of the adjudicatory process than the process's ultimate outcome. They want a voice and for their voice to be heard by people they view as trustworthy. And this is essentially what eBay found when it looked at its data.
Second is one of the fictitious case studies that Schmitz and Rule include at the end of the book to illustrate the basic premises of their global redress system, which they name newhandshake.org. The email from the manufacturer of my new travel clock reminded me of this case study. The case study's specifics are less important than its outcome. In reaction to its interaction with a customer via newhandshake.org (the online dispute resolution system), the fictitious business adds important information to its website. I have a feeling that my travel clock manufacturer can tell a similar story. Via Amazon, it likely heard lots of complaints. The email was its equivalent of updating the website.
The power of global redress system that The New Handshake outlines is to aggregate complaints about goods or services that businesses otherwise would not hear in sufficient numbers to prompt them to update their website or other materials to head off future complaints. Although businesses have the advantage, as Schmitz and Rule aptly point out repeatedly, proceeding from the premise that the vast majority of businesses want to provide quality goods and services to their customers (which is easy to forget when transactions break down), a truly global system will channel information to businesses so that they can provide customers what they expect, decreasing complaints. And although some businesses necessarily will fail, and some customers simply will not be satisfied, The New Handshake shows that, in the future, it is possible to leverage the Internet to increase consumer protection.