Rotten Rule
The Fed’s silly little boxes.

Mr. Clegg is general counsel at the Center for Equal Opportunity.
February 14, 2002 9:30 a.m.

uppose that a financial institution handled home-loan applications over the phone, as many do. Suppose further that the law prohibited racial discrimination in such lending, as it does. Would it be a good idea or a bad idea, then, if the federal government required lenders to ask phone applicants about their race? This is the question raised by a proposed rule published on February 7 by the Federal Reserve board of governors.

According to the proposed rule, the Fed's current policy is to permit, but not require, such questions for "an application made entirely by telephone." The Fed already requires that this data be collected when applications are made in-person, over the Internet, or by regular mail. The new proposal would extend the data-collecting to over-the-phone loans, too, and the Fed now "solicits comment on the benefits and burden of this proposal."

Well, here are my comments. There are three possible situations when a lender is talking with an applicant over the phone: (1) he can't guess the applicant's race; (2) he thinks he can guess the applicant's race, but really can't; or (3) he really has correctly guessed the applicant's race.

In the first instance, it clearly does not advance the cause of nondiscrimination to require the lender to ask the applicant's race. If he doesn't know the race, he can't discriminate on that basis. By giving him that information now, you have made racial discrimination possible when before it wasn't. That's not progress.

It also makes no sense in the second instance to require the lender to ask the applicant's race. Either the lender was going to engage in discrimination based on an incorrect assumption — or he wasn't. If he wasn't, then you haven't helped matters by setting him straight. And if he was, why should we be helping him to discriminate more efficiently?

Which leaves us with the third instance. I will note at the outset that it assumes we can tell a person's race by his voice. Articles of politically correct faith to the contrary notwithstanding, this is certainly true sometimes (though it is also frequently false). The Fed has urged lenders to train employees to better evaluate race and ethnicity by "visual observation"; perhaps the Fed contemplates similar training for "auditory identification." But if the lender already knows the applicant's race, why should we require him to ask about it?

The answer is, I guess, that this will "provide data that assist in identifying possible discriminatory lending patterns and enforcing antidiscrimination statutes" — to quote from another Fed rule in this area. Is this plausible?

Well, let's suppose that the lender is deliberately discriminating. He knows this is illegal, and he won't want to be caught. So if the Fed tells him he has to document the case against himself, what will he do? He might stop discriminating — or he might just lie. In other words, he'll cover his tracks by claiming that some of the applicants whom he knows to be black and plans to turn down had told him that they were white. If he's caught — not likely — he's no worse off than before. And it's his word against the applicants' (who, after all, would have an incentive to lie about their race if they suspected they were being discriminated against). Or he could claim he just made an honest mistake — you know how easy it is to make a writing error when you're trying to carry on a conversation over the phone.

Alternatively, suppose the lender is unintentionally discriminating. That is, he intends to treat blacks and whites equally, but subconsciously he views applicants of one race as better than applicants of the other. Such subconscious discrimination is certainly possible, but it seems rather unlikely in the lending context, where the criteria are overwhelmingly objective: amount of savings, income level, credit history, size of the loan, and so forth.

So we are left with one other possibility, and I suspect that this is what many of the rule's proponents are really after. They're not trying to stop discrimination at all. Rather, they want to encourage it.

That is, they want to pressure lenders into making loans on a racially "proportionate" basis. If lenders have selection criteria — such as wealth, income, credit history, etc. — that have a disproportionate effect on one racial group or another, they want the lender either to get rid of those criteria, or to apply them to applicants of one race differently than they apply them to applicants of a different race.

Either way, the result is discrimination. If a lender were to jiggle his selection criteria with an eye toward reducing the percentage of loans given to blacks, the civil-rights establishment would rightly object. But it would violate the same principle if lenders jiggled their criteria with an eye toward increasing the loans given to blacks. Either would necessarily mean that some other folks who would have gotten loans before will not get them now.

Or, of course, if the criteria are to be left in place but applied differently to different groups, then that itself would be discrimination, by any reasonable person's definition.

The civil-rights groups will object that criteria are being used now that lenders think make sense but that really don't. Sorry, but I'm pretty dubious that lenders are likely to be acting irrationally in lending out their money, making loans they shouldn't and missing better opportunities elsewhere. And I'm even more dubious about letting Jesse Jackson — or even a civil-rights bureaucrat or a federal judge — set the criteria for other people to use in lending their money.

So the benefits of the Fed's proposed rule seem pretty small and the potential mischief pretty big. And, last but not least, there is one other potential cost: that we'll be further institutionalizing race-consciousness.

We will be doing so, moreover, at a time of international struggle in which our skin-color differences seem quite petty, and where the nation has a unique opportunity to move decisively away from bean-counting. Indeed, to judge from the Fed's own statistics, more and more Americans are refusing to claim any race other than human.

The major reason for the proposed rule, according to the Fed, is that more and more home-loan applications are being submitted without racial and ethnic data on them (thank goodness, you can still get the loan even if you won't check a box). Maybe this is a result of there being more applications by phone — or maybe it's simply because more and more people are refusing to fill out what Ward Connerly calls "the silly little boxes."