Friday, October 2, 2015

Hawkins v. Community Bank of Raymore - A Sleeper Chevron Case?

Hi.  I'm an erstwhile law clerk for a couple federal judges; my most recent tour of duty with the judiciary ended Wednesday.  While I apply to jobs, I thought I would entertain myself by blogging on the upcoming Supreme Court term, along with the occasional circuit opinion or law review article that catches my attention.  I won't blog every SCOTUS case, but only those where I think I have something smart to add to what's been written on Scotusblog and other outlets/law blogs, though I do intend to recap most arguments.

The first case I'll write about is Hawkins v. Community Bank of Raymore, which will be argued Monday.  Hawkins is a fun, seemingly straightforward interpretive dispute, though I think it has an interesting wrinkle in it or two that the lower courts or the parties haven't recognized.  Perhaps most intriguingly, Hawkins may raise whether an agency exercise of discretion to blend two permissible readings of a statute, when that blend isn't itself a permissible reading of a statute, can receive Chevron deference.  More broadly, it may aggravate tensions on the Court as to Chevron's scope and even its continuing vitality.

The question Hawkins presents is this: can only a borrower sue to enforce the Equal Credit Opportunity Act's (ECOA) prohibition against banks requiring spousal guaranties, or can a spousal guarantor also sue to enforce that prohibition?  Since 1985, the agency with rulemaking power under ECOA has said guarantors can sue.  ECOA, however, only gives a cause of action to "applicants" for credit, and defines "applicants" in relevant part as "any person who applies directly for an extension, renewal, or continuation of credit."  The defendant bank in this case, which won in the courts below, argues that a spousal guarantor is unambiguously not an applicant, and that the regulation defining applicants to include guarantors fails at Step One of Chevron.  Plaintiffs and the SG's office, which has filed an amicus brief, predictably argue that the regulation offers a permissible and reasonable reading of "applicant."

Background, and Why Hawkins Matters

Before getting into the substance of this dispute, a bit of background on why it matters, at least within the context of ECOA.  (I'll say something about why Hawkins could matter outside the ECOA context later in this post.)  ECOA, enacted in 1974, makes it "unlawful for any creditor to discriminate against any applicant, with respect to any aspect of a credit transaction," on the basis of, in addition to the usual protected characteristics, marital status.  And, it allows aggrieved applicants to sue if violations occur and seek actual damages, punitive damages, and equitable relief.  

ECOA doesn't elaborate on what discrimination against applicants for credit on the basis of marital status might look like, but legislative history shows the 1974 Congress was concerned about the obvious - banks requiring married persons, particularly married women, to get spousal guaranties on loans.  Consequently, only a year after ECOA was enacted, the Federal Reserve Board, exercising its rulemaking power under ECOA, promulgated a regulation which barred banks from requiring guaranties if a married person satisfied the bank's standards for creditworthiness, and which barred banks from specifically requiring spousal guaranties in instances where a borrower's credit was such that a bank legitimately required a guaranty from someone. In 1985, the Federal Reserve Board amended its ECOA regulations to provide that, solely for purposes of the spousal-guaranty rule, guarantors were applicants.

The bank in this case doesn't challenge the regulation's interpretation of marital-status discrimination to reach requiring spousal guaranties.  It only argues that the spousal guaranty regulation's definition of applicant to include guarantors is faulty.  One might wonder why this dispute matters; on the bank's position, while spousal guarantors can't sue to enforce the prohibition against requiring spousal guaranties, their borrower-spouses can.  

The parties' briefs are fairly vague on this point, but it appears the issue is one of damages, and that disabling guarantors from suing would make a crucial difference.  When a borrower's request for credit is turned down for failing to get a spousal guaranty, he may suffer actual damages, though they will often be slight, e.g., a difference in interest rates between the rate the borrower would have gotten from the discriminating lender and what he ultimately got elsewhere.  But when - as was the case in Hawkins - a borrower is required to get a spousal guaranty and gets one, it's difficult to see what damages he suffers.  (One possibility is the association with the guarantor's credit profile with respect to the loan, which could cause some damages.)  If he has no actual damages, he won't be able to seek punitives, and he almost certainly can't seek the equitable relief of voiding his spouse's guaranty.  The injured party, who may suffer substantial damages and who may be able to void her illicitly obtained guaranty (though courts that hold guarantors have causes of action under ECOA disagree on this remedial question), is the guarantor.  So if the bank prevails, lenders will largely avoid liability for requiring spousal guaranties so long as borrowers and their spouses acquiesce to the lenders' demands.

The Eighth Circuit below, and the bank in its brief, makes a pretty simple argument.  "Applicant" is defined by ECOA as a person who applies for credit.  Guarantors don't apply for credit; first, though they desire to facilitate the extension of credit, they don't request credit for anyone, neither themselves nor their spouses, and second, to "apply" means, according to some dictionaries, to seek something for one's own benefit, and applying for credit thus means to apply for credit for oneself.  This argument has substantial surface appeal, and will likely sway those Justices who are (a) textualists and (b) have relatively expansive notions of what interpretations are clear for Chevron Step One purposes, or are skeptical of Chevron altogether.  But I think the SG puts enough holes in the bank's argument, at least to my way of thinking, to survive Step One and win at Step Two.

First, the SG argues that guaranties are implied requests to extend credit to the primary borrower, and that the common law has traditionally regarded them as such.  The SG also notes that some guaranty documents expressly request the extension of credit, or, as was the case of some of the guaranties in this case, recite that they were intended to induce the lender to extend credit.  Second, the SG argues that an applicant for credit need not request credit for herself, and that all the bank has to suggest that "apply" is limited to requests for one's own benefit is one dictionary definition.

The claim that a guaranty is an implied request to extend credit to a borrower seems persuasive enough; a tougher question in my view is whether a guarantor is therefore an applicant for credit.  This is a point on which all the parties might have done better to put down the dictionaries and look more broadly at examples of common usage.  

On the one hand, the SG is quite right that people commonly talk about applying for someone else to receive a benefit; "I applied for my son to get into a selective preschool" or "I applied for my father to receive veterans benefits" are ordinary-enough usages of apply.  On the other hand, note that these examples take the following form: X applied for Y to get something.  When we simply say that X applied for something, we almost always, I think, mean that X applied for himself.  That implication is defeasible if a Y is inserted into the sentence, but the ECOA's definition of applicant is simply a person who applies for an extension, renewal or continuation of credit.  The most and maybe only natural reading of that language is a person who applies for credit for himself.  Further, applying for credit might be a special case.  A guarantor may request a bank to extend a borrower credit, and a guarantor might say that they asked a bank to extend their spouse credit, but would a guarantor ever say, "I applied for my husband to get credit"?  Applying for credit, as a matter of ordinary usage, seems particularly bound up with the person getting the credit. That being said, while guaranties are obviously not the prototypical sense in which applying for credit's meant, it's much harder to show that guaranties aren't a sense in which applying for credit can be meant.

What I expect will sway the less textualist and/or more Chevron-friendly members of the Court are the many improbable results the bank's position would create.  As discussed above, if the bank wins, the prohibition on requiring spousal guaranties becomes pretty toothless in cases of acquiescence - a consideration which speaks powerfully to the reasonableness of the SG's contrary view at Step Two.  Banks would potentially pay a small price in cases where a borrower or a spouse turned down a demand for a spousal guaranty and obtained credit elsewhere, but in cases where banks succeeded in illicitly requiring guaranties, they'd face little to no liability, even though a successful spousal-guaranty demand imposes far greater harms than a failed demand.  

More troublingly, the bank's interpretation would free banks to blatantly discriminate against guarantors in other ways.  ECOA only forbids discrimination against applicants; if guarantors aren't applicants, ECOA doesn't protect them.  A guarantor's request to extend credit to a borrower could be rejected because of the guarantor's race; the guarantor couldn't sue because, on the bank's view, he isn't an applicant, while the applicant-borrower couldn't sue, because he wouldn't have suffered discrimination on the basis of race.  A bank could charge higher rates for loans guaranteed by spousal guarantors than equally credit-worthy non-spousal guarantors; again, the borrower couldn't sue because it would be the non-applicant guarantor, not the applicant-borrower, against whom the bank discriminated on the basis of marital status. Maybe ECOA just fails to reach guarantors, but it seems odd that a statute enacted to stamp out discrimination in lending would actually permit it in a fairly integral aspect of the lending process.  Note, however, that the agency has taken most of the wind out of the sails of this argument by defining "applicant" to include guarantors only for purposes of the spousal guaranty rule (see note below*).  That said, a win for the bank at Step One of Chevron would permanently lock guarantors out of all ECOA protections, which currently isn't the case.

Why Plaintiffs Could (and Possibly Should) Lose, Either Now or on Remand, Even If They're Applicants

It seems fairly clear to me that the agency's interpretation of applicant to cover guarantor is both barely permissible and eminently reasonable - or at least, would have been had the agency interpreted applicant to cover all guarantors (more on that later) - and I would bet there are at least four and maybe five votes for that position.  Yet I'm still not sure that plaintiffs have a claim.  Why?  Because they're not the ones being discriminated against on the basis of marital status; their spouses are.  When a bank requires a spousal guaranty, it reacts to the borrower's marital status and treats him differently from unmarried borrowers by requesting that his spouse guarantee the debt.  How does that request discriminate against the spouse, as applicant, on the basis of marital status?  

In the first place, the request can't discriminate against the spouse as applicant, because the spouse isn't an applicant at the time of the request; the spouse only becomes one when she offers to guarantee the borrower's debt.  The request, as to the spouse, is a solicitation to become an applicant.  If discrimination against applicants encompasses discriminatory solicitation of prospective guarantors who later become actual guarantor-applicants, ECOA's coverage expands quite a bit.  For example, suppose a borrower provides a list of four potential guarantors, three of which are in their 20s, and the lender says it would only accept the older potential guarantor solely because of age and requests that he execute a guaranty.  That potential guarantor could sign a guaranty, thereby become an applicant, and subsequently sue under the ECOA to void his guaranty, claiming that he suffered age discrimination.  That seems wrong.  Yet such a hypothetical is exactly analogous to the claims here, where plaintiffs say they were asked to become guarantors, and became guarantors, solely because they were married to their borrower-spouses.

But even if discrimination against applicants can be stretched to include discrimination against persons who later become applicants, it's difficult to see how a spousal guarantor is discriminated against on the basis of marital status.  It's true that, but for the spouse's being married to the borrower, the lender wouldn't ask the spouse to enter the transaction.  It's also true that the spouse is singled out vis-a-vis others who have close relationships with the borrower.  But does that amount to marital-status discrimination?  I don't think so, for two reasons. 

First, unlike the borrower, who is discriminated against purely by virtue of being married, the spouse suffers disparate treatment not because of marital status, but because of who she's married to.  Were she married to someone else, she wouldn't be solicited.  So to say she suffers marital-status discrimination, "marital status" has to include spousal identity.  Once one sees that plaintiffs' claims are claims of discrimination on the basis of who precisely their spouses are, they start to take on a peculiar class-of-one flavor that's difficult to square with the statute's listing marital status among a series of classic protected traits.  That is, plaintiffs' theory is really that they were discriminated against because they happen to be the single persons in the world who were married to their borrower-spouses, not that they were discriminated against because they're married (unlike their borrower-spouses, who were).  Treating marital status as inclusive of spousal identity would also expand ECOA in other factual contexts; a bank that denied a woman a loan because she was married to a well-known fraudster would potentially be committing marital-status discrimination.

Second, where is the disparate treatment here?  Usually, discrimination involves disparate treatment among similarly situated persons on the basis of some protected trait; for example, if a bank denies guaranties executed by African-Americans but not by whites, that's racial discrimination.  But when a spouse is solicited to execute a guaranty on the basis of her being married to a prospective borrower, who are the comparators to whom we compare her?  The millions of persons who weren't solicited because they're not married to the borrower?  The trouble with that, among other things, is that the discrimination there is a discrimination as between non-applicants, all but one of whom never became applicants.

I tend to think, then, that while the agency regulation permissibly and reasonably defines applicants to include guarantors, spousal guarantor-applicants don't suffer marital-status discrimination when they're solicited to become guarantors on the basis of being married to a prospective borrower.  It doesn't necessarily follow, though, that they can't sue.  

Recall that ECOA gives a cause of action to "the aggrieved applicant" if a creditor violates ECOA.  A creditor discriminates against an applicant-borrower when it demands a spousal guaranty; if the spouse executes the guaranty, she becomes an applicant on my view, and is arguably aggrieved by the demand.  In Thompson v. North American Stainless, the Court unanimously held that where an employer retaliates against an employee by firing her spouse, her spouse is a "person claiming to be aggrieved" who can sue under Title VII for retaliation against his wife.  ECOA limits its cause of action to aggrieved applicants, but a guarantor is, on my view, an applicant, and is aggrieved by the discrimination against her borrower-husband in a pretty direct way.  That said, ECOA's reference to "the aggrieved applicant" might be read to refer only to the applicant who suffers discrimination, not other applicants who happen to be aggrieved by discrimination against an applicant.  The definite article isn't so much the problem here as the seeming linkage between who the substantive prohibition protects (applicants) and who gets to sue (applicants), which suggests the people who get to sue are just the discriminated-against applicants.

Is Any of That Before the Court?

Whether plaintiffs are aggrieved applicants, assuming they're applicants, probably isn't before the Court.  The district court and Eighth Circuit held that plaintiffs weren't applicants, and the Court granted certiorari on whether or not they were. The bank's brief in opposition to certiorari raised a number of alternative grounds for affirmance; none were the argument sketched above.  And the bank's merits brief only engages with the definition of applicant. 

The reason this issue just might be before the Court is this.  The regulation defining applicants to include guarantors only defines them as guarantors for purposes of the spousal guaranty requirement rule, so as to give guarantors standing to sue to enforce that rule. (See note below.)  So what plaintiffs and the SG are asking the Court to defer to is not a rule that generally defines applicants as guarantors, but a rule designed to give them standing to make one particular claim that there's a pretty good argument guarantors can't make.  Whether that rule is reasonable may turn on whether guarantors can ever have claims for violations of the spousal-guaranty regulation.  So I wouldn't be incredibly surprised to see the Court decide whether spousal guarantors are aggrieved applicants, and I wouldn't be surprised at all to see the Court reverse and remand with instructions to consider that issue.

Larger Implications?

Whether the agency regulation reasonably construes applicants to include guarantors only for purposes of the spousal guaranty requirement rule may also turn, more generally, on a question about Chevron deference that's come up in some recent D.C. Circuit cases.**  If a statute can be permissibly read to mean both X and not-X (here, that guarantors are applicants, and guarantors are not applicants), does that mean it can be permissibly read to mean X some of the time and not-X some of the time?  There's a pretty plausible argument that, whatever applicant may mean, it can't mean that guarantors are applicants only when they're the victims of spousal-guaranty requirements, but not when they're the victims of any other prohibited form of discrimination. 

The counter to that argument, inspired by reasoning in the D.C. Circuit opinion linked above, is that the agency can permissibly read applicants to include guarantors and then exercise its discretion to preclude guarantor enforcement in certain contexts, just as the agency could, were it charged with enforcing the ECOA, decline to enforce it in certain contexts.  The counter to that argument is that, once having determined that guarantors are applicants, the agency can't deny them a cause of action that the statute gives them, even if it itself could to decline to enforce the ECOA as to certain guarantor-applicants.  

The counter to that counter, which I'm generally inclined to accept, is that it misunderstands Chevron.  When the Federal Reserve Board decided it could treat guarantors as applicants for a single purpose, it didn't do so because it concluded that guarantors legally were applicants; it did so because it concluded that "applicants" could be permissibly read to include guarantors, and could not be, and because treating a particular subset of guarantors as applicants made the best policy sense.  Congress, in defining "applicants" ambiguously enough to not clearly dictate the treatment of guarantors, delegated to the agency a policy space in which to operate, not a binary choice, or a binary guessing-game as to what Congress meant.  

My reservation about that final counter is that, while it gets Chevron right, it begs the question as to the details of particular ambiguities.  If an ambiguity, as all ambiguities do, can have outer limits, why can't it also have an excluded middle?  Indeed, having an excluded middle - that is, having a small, discrete set of possible and distinct meanings with nothing in between, not a continuous range of possible meanings, as a word like "tall" does -  is arguably one of the properties of ambiguity, as opposed to vagueness.  (This isn't to suggest that because Chevron talks about ambiguity, all interpretive disputes to which Chevron applies have excluded middles.  Chevron can apply to vagueness too - indeed, may well make more sense as applied to vagueness than as to ambiguity, for reasons I don't have the space to get into here.) 

I tend to think Chevron's skeptics on the Court, who seem to be growing both in number and skepticism by the year,*** will not be enthused with a rule that (a) reads a statutory term in such a way as to draw distinctions that traditional interpretive processes never could, and that (b) exists only to manipulate statutory standing to enforce a prohibition that would exist no matter how the agency defined the term in dispute.  Some members of the Court may doubt whether, under Mead, the Federal Reserve Board and now the CPFB ever were delegated the authority to control statutory standing in civil litigation; in the Texas Housing case last term, the Court declined, for reasons left unstated, to defer to a regulation addressing the scope of, and evidentiary burdens associated with, disparate impact liability under the Fair Housing Act, even as the Court essentially sided with the agency.

Justice Thomas is particularly important to watch here.  On the last day of the 2014-15 term, he wrote a concurring opinion in Michigan v. EPA doubting Chevron's constitutionality under an expansive theory of the non-delegation doctrine, after having previously doubted the constitutionality of Auer deference earlier in the term.  In Michigan, Chevron's continuing validity was irrelevant to his vote because he concurred in an opinion holding the EPA lost at Step One.  What if, in this case, Justice Thomas concludes the agency's reading of applicant isn't the best reading of ECOA, but is permissible?  Will he refuse to apply Chevron?  His non-delegation concerns about Chevron should be aggravated by an agency that claims authority to enact its own compromise between dueling permissible readings of what Congress wrote.  Hawkins may be remembered, if at all, as a flashpoint in the Court's escalating deference debates. 

* Under the the regulation guarantors are only defined as applicants for purposes of the rule prohibiting spousal guaranty requirements, leaving guarantors who suffer racial discrimination, gender discrimination, age discrimination, or even marital-status discrimination in some fashion other than being required to execute a spousal guaranty, out in the cold.  See 12 C.F.R. 1002.2(d); 50 Fed. Reg. 48,020 (explaining that the definition of applicant was modified to include guarantors solely for purposes of the spousal-guaranty rule, so that guarantors could sue for violations of it, but not for purposes of anything else in the ECOA or enforcing regulations, so as to protect creditors from liability for unspecified "technical violations"). 

** The case in the link involved whether an agency could define a "mine fire" to include some smoldering fires, i.e. fires without flames, given that "fire" is sometimes used to encompass smoldering fire, sometimes isn't, but is never used to only encompass the precise subset of smoldering fires the agency defined "mine fire" to include.  The D.C. Circuit said the agency could define "mine fire" to include only that subset, making essentially an enforcement-priorities argument.

*** A group which would include (1) Justice Thomas, who has come to doubt the constitutionality of Chevron, see discussion supra, (2) Chief Justice Roberts, who most recently has held for the Court that Chevron doesn't apply to major questions, written an opinion arguing that Chevron doesn't give agencies authority to resolve self-contradictions in statutes, and authored a dissent in City of Arlington calling for case-by-case determinations as to whether Chevron applies to questions seemingly within the scope of an agency's general rulemaking authority and raising vaguely constitutional concerns about the power of the administrative state, (3) Justice Kennedy, who joined that dissent, and (4) Justice Alito, who joined that dissent and has written several opinions of his own doubting the wisdom of Auer deference.  Justice Breyer is a skeptic of Chevron doctrine and shares Roberts's view (perhaps even more aggressively than Roberts) that whether deference is appropriate is a case-by-case issue, see, e.g., his concurring opinion in City of Arlington (which is only the latest iteration of an argument he's been making for thirty years), but he is not a skeptic of deference or the administrative state writ more largely and I would suspect he is likely to find the Federal Reserve's construction of "applicant" eminently reasonable.  

There theoretically could come a case where Justice Thomas, for constitutional reasons, concurs in a Breyer-Roberts-Alito-Kennedy judgment against an agency holding that Chevron doesn't apply to a given ambiguity on a case-by-case theory - a development, which, were it to occur, would rock the world of administrative law and roughly return us to the pre-Chevron days of permissive deference regimes.  I doubt that that case is on the Court's docket this term, or that that majority will ever form, as it's unlikely that that particular set of Justices would agree on a regulation's being bad, even if they agree on standards of review.  But it could happen, and if it does happen it will probably be in a case with low ideological valance - like this one, though this case may not be low enough on ideological valance and probably doesn't present an issue where Justice Breyer and the City of Arlington dissenters would say the agency doesn't get Chevron deference.

No comments:

Post a Comment