Consumer Law & Policy Blog

Credit Report Accuracy

by Jeff Sovern

A couple of weeks ago, Ira Rheingold and I had an op-ed in the Times about issues with credit reports.  Almost on cue, the Associated Press reports Jury awards Oregon woman $18.6M over credit report.  It seems she had been trying to get Equifax to correct errors for two years. 

In the op-ed, we mentioned some of the changes we would like to see, including that Congress should require "greater accuracy from credit bureaus. . . ."  At present, under 15 U.S.C. § 1681e(b), credit reporting agencies are required "to follow reasonable procedures to assure maximum possible accuracy . . . ."  I would like to see Congress raise that to  “best procedures to assure maximum possible accuracy.” Courts have been too willing to excuse errors in credit reports by finding the credit bureau procedures that caused them or failed to prevent them reasonable.  For example, in Sarver v. Experian Information Solutions, 390 F.3d 969 (7th Cir, 2004), a credit bureau erroneously reported that some of the consumer's accounts had been involved in bankruptcy even though the consumer's other accounts did not carry that notation.  The court found that it was reasonable for the credit bureau not to have picked up the discrepancy between the different accounts.  The court also found it reasonable for the bureau not to know what's in its files or that its files contained inconsistencies, given that the credit bureau receives over 50 million updates a day.  I am told that credit bureaus have software that flags some anomalies, such as apparent attempts by children to take out car loans; surely it would be possible to create similar software to pick up anomolies of the sort in Sarver.  I think credit bureaus should be obliged to use such software, but I fear that decisions like Sarver, based on the old standard, reduces their incentive to incur the cost of the software.  By contrast, a standard that obliged credit bureaus to use best efforts would cure that problem, and by increasing report accuracy, help both consumers and lenders.

Dazzling Times Op-Ed: To Catch a Creditor

Here. 

WSJ Op-Ed: Regulators Have Created a Mortgage Minefield

by Jeff Sovern

Former governor, HUD general counsel, and current head of the American Bankers Association Frank Keating has penned an op-ed for the Wall Street Journal (behind paywall) with the headline above.  Keating complains that HUD has adopted the disparate impact test for discrimination; that the CFPB's qualified mortgage rule (QM rule) will restrict lending to borrowers who would otherwise get mortgages; and that banks which comply with the QM rule risk falling afoul of the disparate impact test.

Keating explains: "The QM requirements will result in an immediate tightening of credit, with banks substituting a one-size-fits-all federal mandate for their own good judgment and sound underwriting." That sounds scary, until you remember that the banks' "good judgment and sound underwriting" led to the making of millions of mortgage loans on which borrowers defaulted and were foreclosed upon, which wrecked the economy and led to enormous bailouts to the banks which exercised such good judgment.  Keating never mentions any of that. Personally, I prefer the QM rule to that kind of "good judgment and sound underwriting."

Keating says use of the disparate impact test "represents a radical shift in how the government enforces fair housing law."  Somehow he forgets to note that more than 40 years ago, the Supreme Court approved use of the "radical" disparate impact test, albeit in the employment discrimination context.  Courts have used the disparate impact test in lending discrimination for some time.  To be sure, the Supreme Court has taken the Mount Holly case on the issue, but that was not because of a circuit split on this issue--because there isn't one.  And even if there were, that wouldn't make use of the test "radical." 

 

Supreme Court Grants Cert on Disparate Impact Test

by Jeff Sovern

The Supreme Court today granted cert in Mount Holly v. Mount Holly Gardens Citizens in Action, which raises the issue of whether courts can use the disparate impact test in federal Fair Housing Act (FHA) cases. Because the language of the FHA is similar to that of the Equal Credit Opportunity Act (ECOA), the Court's decision is also likely to have an impact on whether the disparate impact test can be used in ECOA cases.  ECOA bars discrimination in lending to consumers.  Under the disparate impact test, courts can find discrimination when the discrimination is unintentional, as long as the conduct has a discriminatory effect, though under some circumstances, lenders can continue to engage in even discriminatory conduct.  You can read more about the case in the SCOTUS blog, Bloomberg, the LA Times, the Associated Press, and the Ballard Spahr web site. I had previously explained why I think it is desirable for courts to use the disparate impact test here. 

One thing that is strking about the case is that the circuit courts are not split on this issue. The Courts of Appeals to address the issue and the relevant administrative agency all bless the use of the disparate impact test.  It is difficult to see why four justices would have voted to take the case under those circumstances unless they harbored significant doubt about the validity of the test. 

Daniel Schwarcz Paper on Post-Sale Monitoring of Financial Products

Daniel Schwarcz of Minnesota has written Monitoring, Reporting, and Recalling Defective Financial Products, University of Chicago Legal Forum (2013).  Here is the abstract:

In recent years, innovations in consumer financial protection have drawn heavily from the law governing the safety of tangible products.   This short essay, prepared for a symposium entitled "Frontiers of Consumer Protection," extends this approach by evaluating whether the law governing post-sale product safety for tangible consumer products offers potential lessons for consumer financial protection. In particular, it considers the extent to which the law could require providers of consumer financial products to affirmatively monitor, report or correct regulatory problems with their products that come to light after sale.  Such post-sale self-policing by financial firms could offer benefits similar to those that justify standard product recall law: the sellers of financial products may be best situated to identify unintended negative consequences of their products that come to light after sale and to minimize the resulting costs through proactive market interventions.  The essay illustrates this approach by reconsidering the Consumer Financial Protection Bureau’s rules implementing the Dodd-Frank Act’s “ability to repay” requirement for mortgages.  It also suggests that the analytical frame embedded in post-sale safety requirements offers potential lessons for a number of other consumer financial protection issues, such as disclosure requirements and prohibitions on impermissible discrimination.

When Consumer Reports Are Not Used for Credit

by Jeff Sovern

We frequently write here about the Fair Credit Reporting Act because of its application to credit reports.  But in fact, the FCRA applies to many non-credit transactions.  Those transactions typically take one of two forms. In one form someone uses a credit report for something other than a lending decision. For example, many employers use credit reports to decide whether to hire someone. Thus, yesterday's Times included an article, The Long Shadow of Bad Credit in a Job Search, about how people lose their jobs, default on debts because of the job loss, end up with a bad credit report because of the defaults, and then can't get a new job because of the bad credit reports. Some states have responded by limiting the use of credit reports for employment decisions.

The second type of non-credit transaction arises because the FCRA, notwithstanding its name, applies to "consumer reports," and defines "consumer report" as including many reports that have nothing to do with credit. For example, payday lenders would like to know if someone has committed check fraud and landlords want to know if someone has violated a lease provision in the past. "Specialty consumer reports" provide that information.  Because the FCRA applies also to these reports, consumers have the same rights to see the reports and corect errors as they do for credit reports. The CFPB has compiled a list of such specialty consumer report providers here.

UPDATE: Lea Krivinskas Shepard of Loyola of Chicago, has written an article, Toward a Stronger Financial History Antidiscrimination Norm, 53 Boston College Law Review (2012), which discusses the normative debate surrounding employers' use of credit reports.

More Evidence That Consumers Can't Provide Consumer Protection

by Jeff Sovern

Traditional consumer protection rules drew on law and economics models that assumed that consumers were rational and that when consumer markets functioned poorly, all that needed to be done was give rational consumers the ability to protect themselves. For example, the Truth in Lending Act's focus on disclosures presupposes that rational consumers would read them and use them to make appropriate decisions.  More recently, the 2003 FACTA Act amended the Fair Credit Reporting Act to allow consumers to obtain a free copy of their credit report each year from each of the big three credit bureaus.  Again, rational consumers should take advantage of that opportunity to protect themselves from identity thieves. 

But behavioral law and economics has shown that consumers are often irrational in predictable ways.  To the extent that law-makers fail to take these irrationalities into account in formulating rules, the rules will fall short of what is needed to protect consumers.  More recent laws, like 2009's Credit CARD Act and 2010's Dodd-Frank Act have used some of these insights to create rules which fit real consumers better (my co-author, Dee Pridgen, wrote about this in her paper, Sea Changes in Consumer Financial Protection: Stronger Bureau and Stronger Laws, 13 Wyoming Law Review (2013)). 

As for the specific examples I noted above, I have previously written about failures of disclosure, as have plenty of others.  But what about the right to obtain free credit reports?  It turns out that less than 20% of consumers check their credit report annually, according to the CFPB.  In other words, more than 80% of consumers are not taking an easy, free opportunity to determine if they have been victims of identity theft (I wonder how many readers of this blog post fall into that category). 

So if the vast majority of consumers do not protect themselves, the government has a choice in making rules: it can continue to depend on consumers to protect themselves, knowing that many will not and therefore will not be protected, or it can provide protections that do not depend on consumers to act.   

An Item on My Wish List

by Jeff Sovern

Tonight marks the start of the NFL draft, a time when hope blooms for all fans of an NFL team, because they can always suppose that their team has used the draft to get better--and many teams will.  Many fans create wish lists--what they want for their teams.  In that spirit, I have a wish list of what I want in consumer protection. One item would shed significant light on whether discrimination actually occurs in lending.  I wish the CFPB would work with the credit bureaus to create records on imaginary individuals.  These individuals would each be paired with another fictitious person with an identical or nearly identical record, except that one of the two would appear to be a white male, and the other either a person of color or a woman or both.  The credit files would not actually indicate race or gender, but would signal such characteristics through zip codes, names, etc.  Then the Bureau would have the imaginary people apply for credit.  By comparing the applicants' grant and denial rates, and, for those who are offered credit, the terms of the offer, the Bureau would go a long way to showing that discrimination does or does not occur in the granting of credit.

To be sure, this would require making false statements to lenders, which might violate federal law. Perhaps the relevant laws have exceptions for such experiments; if not, the remedy would be to seek enactment of such an exception, which seems unlikely to happen in the current environment, but you never know. The idea is not very different from the testers that have been used to determine if realtors and others are engaged in housing discrmination.

What's on your consumer protection wish list?

Another Reply to Chris Willis: This Time on Disparate Impact

by Jeff Sovern

Chris Willis of Ballard Spahr's CFPB Monitor recently blogged about whether disparate impact can properly be termed discrimination and therefore prohibited by ECOA.  For those who don't know, courts generally recognize three ways to prove discrimination in violation of ECOA: direct evidence, disparate treatment (except in the Seventh Circuit) and disparate impact, or as it is sometimes called, disparate effect. Direct evidence would exist, for example, if a lender said we don't lend to old people.  For obvious reasons, that doesn't happen that often.  Disparate treatment requires that the lender have discriminated intentionally.  That turns out to be very difficult to prove, so difficult in fact that it almost doesn't matter that the Seventh Circuit doesn't use the disparate treatment test. That leaves disparate impact. Disparate impact requires the plaintiff to show that using a particular criterion has a disproportionate effect on a particular group.  That can happen even if the lender did not intend to discriminate against anyone.  If the plaintiff carries that burden, the lender has the burden of showing that the criterion has a manifest relationship to creditworthiness.

To take a somewhat fanciful example, but one that I think makes the rule fairly clear, suppose the cattle organization that sued Oprah Winfrey some years back for disparaging beef decides to open a bank and that it will lend only to people who eat meat.  Some religions require their adherents to be vegetarians.  Accordingly, the lender's policy would discriminate against members of those religions, which I believe includes Budhism, at least in some forms.  The lender would not have intended to discriminate against Budhists, but its policy would have a disparate impact upon them. Consequently, the lender would have to show that its criterion enabled it to avoid lending to non-creditworthy borrowers. 

So the disparate impact test prevents unintentional discrimination against members of protected groups.  Why should we use such a test?  I confess I haven't studied the literature on this as much as I would have liked, but I can think of several reasons (I think Chris knows more about this than I do, but I wanted to respond anyway).  First, the legislative history of ECOA supports the use of the disparate impact test.  Some lawyers and judges oppose the use of legislative history, but I don't intend to repeat that debate here; it happens that I believe in using legislative history, at least under some circumstances. Second, the administrative agencies that have interpreted ECOA, such as the Federal Reserve and HUD, have also endorsed using the disparate impact test.  Courts usually defer to administrative agencies in interpreting statutes.

There also are policy justifications for using the disparate impact test.  if Budhists can't borrow from a particular lender because of their religion, does it really matter that the lender didn't intend to discriminate against Budhists but just insists on lending to meat-eaters?  Whether the lender intended to discriminate or not, the Budhists who refuse to eat meat can't borrow from that lender because of their religion.  They are being treated as second-class citizens by that lender.  And, again, if the lender can demonstrate that meat eaters are less likely to default, then the lender could continue insisting that its borrowers eat meat.

Another reason to recognize the disparate impact test is because the other tests have not been very effective. If courts don't block disparate impact, ECOA won't have very much impact at all.  In other words, lenders could continue to treat people differently based on criteria that don't relate to creditworthiness, but that produce different effects on protected and non-protected groups.  And just suppose the lender really did want to discriminate against Budhists, and is willing to accept as the price of doing so that it also discriminate against non-Budhist vegetarians, to avoid creating evidence of disparate treatment. If the lender managed to avoid providing direct evidence of animus towards Budhists, disparate impact would be the only avenue for stopping the lender.  Take that away, and the lender could discriminate with impunity. 

Chris provides an engaging example about two car loans he and his wife took out.  I think his example is beside the point because, unless I am misinformed, disparate impact cases are not based on one or two examples, but on enough examples to create statistical evidence that different groups are being treated differently.  If a car dealer systematically charges one group higher rates than another, it becomes much more difficult to say that the explanation is innocent (but maybe I'm not understanding Chris's example properly, because I don't see why the car price, as opposed to the loan terms, are relevant).  Similarly, Chris's point that the different treatment is not that different also misses the mark, in my view.  If the differences are not that great, then maybe there hasn't been a disparate impact.  But that doesn't mean that in other cases, in which the differences between the treatment of two groups are statistically significant, we should ignore those differences. 

I agree with Chris that unintentional discrimination doesn't carry the same moral objection as intentional discrimination.  But treating people differently because of race or religion, etc. is still deeply troubling. In any event, I agree with the courts and administrative agencies that have concluded the disparate impact test is an appropriate test of whether ECOA has been violated.

More on the CFPB Auto Financing Rules: A Reply to Christopher Willis

by Jeff Sovern

Earlier this week, I commented on car dealer opposition to the CFPB's auto financing rules.  Chris Willis replied on Ballard Spahr's informative CFPB Monitor Blog.  I want to respond to two of Chris's points: first, that compliance, policing, and later enforcement steps will increase the cost of credit; and second, that dealers will respond to the loss of revenue from financing by increasing the cost of cars.

As to the first point, Chris may well be right that compliance, policing and enforcment efforts will increase the cost of credit; I don't know enough to say.  No doubt efforts to eliminate discrimination in other contexts--such as employment, housing, and consumer credit transactions--have also increased costs.  But the country has been willing to accept those costs as the price to pay for blocking odious discrimination.  In addition, sometimes warnings of harms caused by regulation turn out to be off the mark.  For example, back in 2000, before the Federal Reserve Board adopted regulations lowering the APR trigger for coverage under the Home Owners Equity Protection Act (HOPEA), see Regulation Z, 12 C.F.R. § 226.32(a)(1)(i), a banking industry association predicted that the regulation would drive legitimate lenders from the subprime market. See Sandra Fleishman, Fed Favors Tougher Loan Rules; Abuses in Subprime Lending Are Targeted, Wash. Post., Dec. 14, 2000 at E01. We now know that lenders did not leave that market, and that the economy might be in much better shape if they had.  That doesn't mean that Chris is wrong to call attention to the possible cost increases. But, in my opinion, it does mean that we need to know more about the likelihood that such predictions will come true and the amount of the cost increases before allowing them to derail rules which serve the important purpose of blocking discrimination. 

Moreover, lenders have a history of predicting higher costs when opposing regulation than they later incur when the regulation takes effect.  I don''t say that to imply dishonesty; rather, at the point at which the lenders predict the higher costs, they are unlikely to devote resources to minimizing costs.  Once the regulation is adopted, the lenders have an incentive to minimize costs and find that they are able to do so when they put their ingenuity towards such an effort.  For example, before the amendment of the Equal Credit Opportunity Act to require disclosure of reasons for an adverse action, which in its current form appears in 15 U.S.C. § 1691(d), Congress was told:

Sears Roebuck and Company stated that its annual estimated cost for such compliance would be approximately $5 per letter. * * * Even if all creditors could operate as efficiently as Sears, the aggregate annual cost of this requirement could easily amount to hundreds of millions of dollars. The requirement is staggeringly inflationary.

Testimony of National Retail Merchants Association, Hearings on S. 483, S. 1900, S. 1927, S. 1961, and H.R. 6516, Subcommittee on Consumer Affairs, U.S. Senate Committee on Banking, Housing and Urban Affairs 399 (1975).

After enactment, the Federal Reserve Board surveyed large creditors to determine the actual costs of furnishing rejected applicants with statements of reasons for adverse actions under Reg. B § 202.9. Board of Governors of the Federal Reserve System, Exercise of Consumer Rights Under the Equal Credit Opportunity and Fair Credit Billing Acts, 64 Fed.Res.Bull. 363 (1978). Sears reported that the average cost of providing rejected credit applicants with the reasons for rejection was 59 cents.

Chris also argues that dealers will increase the price of the cars they sell to recoup the revenues they would otherwise get from financing.  This assumes that dealers are not already charging as much as they can get for their cars (and if they aren't, why aren't they? Car dealers are not known for kindness).  It also assumes that dealers will not simply absorb the lost profits or find another source of revenue.  But let's suppose that dealers across the board find that they need to raise car prices because of the CFPB's efforts. We still don't know how much that increase would be.  In addition, given the past willingness of consumers to absorb price increases to eliminate discrimination, why would we assume they would not be willing to do the same here?