Patterico's Pontifications

6/19/2006

L.A. Times and Liberal Organization Claim: Study Shows Race Affects Mortgage Rates

Filed under: Race — Patterico @ 6:01 am



The L.A. Times reports:

[A] new two-year investigation in six metropolitan areas suggests that mortgage quotes are not always colorblind.

. . . .

• Brokers discussed loan fees with 74% of the white shoppers but only 31% of the minority shoppers. Yet loan fees — points and a wide variety of other charges — can add significantly to the out-of-pocket costs of one mortgage compared with another, even if the interest rates are the same.

The story goes on to list other ways in which whites were treated better than similarly situated (or better situated) blacks and Latinos.

I am generally critical of such studies for lacking controls — and of the L.A. Times for failing to report the obvious flaws with such studies. The usual study finds disparities based on race, without controlling for the factors that lenders look at. But ths article claims that at least some of these controls were in place:

The six markets investigated were Baltimore, Washington, Chicago, Los Angeles, St. Louis and Atlanta. In each area, African American and Latino couples or individuals visited the same mortgage brokerage firms as white shoppers, all purporting to apply for home loans of similar amounts. All the applicants were assigned specific income, credit and employment profiles to present to loan officers. African American and Latino applicants had slightly higher incomes, credit scores and longer employment backgrounds than their paired white colleagues making separate applications at the same brokerage firms.

That sounds good, if you fully trust the organization that conducted the study, and if you further trust the newspaper to accurately report any potential flaws in the study.

I don’t. Here’s why:

The organization that conducted the study is called the National Community Reinvestment Coalition. It is a group devoted to increasing the availability of credit in minority communities. As the organization’s mission statement says:

NCRC was formed in 1990 by national, regional, and local organizations to develop and harness the collective energies of community reinvestment organizations from across the country so as to increase the flow of private capital into traditionally underserved communities.

So, the whole purpose of the group would be undercut by a study that showed minority communities are being served just fine by traditional lenders. I’d be as quick to take the word of these people as I would be to accept scientific studies by cigarette companies showing no link between tobacco smoking and lung cancer.

Somehow, the fact that the study was conducted by a group with an axe to grind doesn’t make it into the article about the study. And the L.A. Times has a history of printing uncritical articles touting studies that reinforce liberal assumptions — and somehow, the agenda of the organizations conducting these studies always gets buried.

I can’t seem to find the study itself on the organization’s web site. I see only a press release about it, and a complaint filed with HUD that references it. For some odd reason, the study itself is AWOL.

Without seeing the details of the study, I’m not trusting that the proper controls were in place. Lenders look not just at your income, but also your total debt compared to income. They look not just at the length of your employment background, but at your particular job, your employment status, your job history, and your income stability. They look not only at your credit score, but also at your ability to document financial information, your total financial assets, your loan-to-value ratio, and many other factors.

An experiment that does not control for *all* these factors is unreliable.

Nor do I have any idea whether the study used “double-blind” techniques to eliminate the observer-expectancy effect, defined as “a cognitive bias that occurs when a researcher expects a given result and therefore unconsciously manipulates an experiment or misinterprets data in order to find it.” Did the purported credit applicants know that they were participating in a study designed to determine whether race affects credit opportunities? Were they getting paid by the NCRC? Did they do anything, in dress, attitude, or otherwise, to (consciously or not) skew the results?

I have my suspicions. But again, without the details, it’s impossible to say.

Has the study been subjected to peer review? I have no idea — but I rather suspect not.

If you trust a liberal group to conduct the study properly, and a liberal newspaper to expose any flaws in the study, then voila! we have our proof. Lenders are discriminating on the basis of race.

But if aren’t quite that trusting — and I suggest you shouldn’t be — then the jury is still out, at least until the study itself is made public. The press release has contact information for a person who can provide more information. I have asked that person for a copy of the study. I’ll let you know what I hear back — if I do.

24 Responses to “L.A. Times and Liberal Organization Claim: Study Shows Race Affects Mortgage Rates”

  1. Lenders want to maximize profit. Discrimination reduces profit. q.e.d.

    lincoln republican (6ec212)

  2. LR, you seem to be confusing profit maximatization with profit uniformity. They are not the same thing, and are frequently at odds with each other. Suppose you are a mortgage broker. If you charge everyone the same price (after adjusting for income, credit score, and anything else that bears on the cost or value of the product to you), then you will lose potential profits. Guaranteed. If you price your product at the top of the market, you’ll make a killing on the few suckers willing to do business with you, but forgo profits from everyone else. If you price at the bottom, you’ll gain market share, but you’ll lose profits you could have made on the consumers who would have been willing to pay more.

    At the one extreme, if your price is the highest

    Now, suppose you are not just a mortgage broker, but a ruthlessly profit-driven one who cares nothing about lender laws, racial harmony or antitrust. To you, finding that “sweet spot” that maximizes lawful profits by chasing away the “right” number of cheap customers and forgoing the “right” amount of windfall profits from big spenders is a fool’s errand. Why not really maximize profit by charging every individual the highest price you think he will be willing to pay?

    From there, it’s not hard to imagine (note I said “imagine,” as I have no idea whether any real data bears this out) that market studies may find that all races are not equally savvy, equally cheap or equally extravagant. If it could be shown that the average black family (who may not own a computer, and if they do it’s mostly for games) with a given combination of credit score, income, etc. is willing to pay 100 basis points more than the average white family (whose geek son has already scoured the Internet to find out what rates are out there), then a purely rational, profit-maximizing businessman might well decide that it makes sense to offer overpriced loans to blacks, and minimally profitable ones to whites.

    Xrlq (23419a)

  3. Lenders want to maximize profit. Discrimination reduces profit. q.e.d.

    Looks like we don’t need any civil rights laws at all.

    actus (ebc508)

  4. Looks like we don’t need any civil rights laws at all.

    That is one philosophy: that the marketplace will eventually reward businesses that discriminate with the boot.

    Patterico (50c3cd)

  5. In dealing with certain businesses, you have to work very hard to get a good deal (mortgage brokers and car dealers come to mind.) Could it be that a higher percentage of blacks are trained to be victims, are used to letting governmental agencies represent their interests, and thus do not fair as well in negotiating their own best deal?

    Lincoln’s point is valid. It doesn’t make sense that the lender’s would let discrimination hurt their profits. It makes more sense that there is something different on the other side of the table.

    Mike S (d3f5fd)

  6. Or could it be that the observer-expectancy effect caused the minority applicants in the study not to push very hard, and the whites to ask a gazillion questions?

    I was pleased with the mortgage I got from my mortgage broker, who did much better for me than any of the numerous lenders I had contacted on my own. But I had to push and push — and ultimately get him in a bidding war against someone else — to get a rate and fees I could live with. It was ugly, drawn-out, and highly unpleasant.

    Patterico (50c3cd)

  7. communication occurs on a number of planes beyond the verbal. like most lawyers, loan officers talk to strangers every day for a living. i can’t empathize very well with loan officers, but i can tell you what it’s like for lawyers: we become skilled at discerning lies from experience by analyzing the signal on all of those planes, not just the verbal. did the axe-grinding minority testers somehow cue the loan officer that they were lying?
    follow the money. these organizations live off government grants. professional malcontents feed their families with grants. if one of them suddenly announced that america had finally become a peachy keen, non-racist lovezone, wouldn’t that threaten the flow of grant money?

    assistant devil's advocate (a27f31)

  8. Right on target.

    Patterico (50c3cd)

  9. LR is right, discrimination does reduce profits… but in this case, a bit different than in the way one might think..

    Since nobody forced the blacks (in this study) to accept the terms being offered, let’s stipulate that these terms represented the fair market price (the price point at which a willing buyer will purchase from a willing seller). If so, then the allegation should be reframed: blacks weren’t being overcharged for loans, whites with similar credit factors were being undercharged.

    Yet, why would mortgage brokers lower their profits by offering whites a lower price than the price they would presumably be willing to pay? Having a relative in that business, I know how brokers try to maximize their return from each loan they sell (a prime reason for the mysterious fees that pop up at settlement… a great way of boosting the return per loan). Brokers don’t like to ‘leave money on the table’. It would be one thing for them to do so for a relative, a friend or someone they were trying to suck up to… but to forego profits for an average white guy? I don’t see it happening.

    steve sturm (b5aa23)

  10. Institutional lenders package loans for the secondary mortgage market. It’s a cookie cutter business, and they don’t collect racial information, it’s not only illegal, it’s irrelevant.

    To get a loan, you need a subject property which is worth the appraised value, you need to have enough cash to make the down payment and cover the closing costs, and you need to show a record of financial responsibility (the credit report), and at least two years of proved earnings (usually a job) adequate to make the mortgage payment including taxes and insurance, and to meet all your other long term debts (the ratios).

    Meet those requirements and you get a loan no matter if your uncle is J Fred Muggs.

    Now, loan brokers are a bit different. They get paid extra if they can get you to accept a higher interest rate or to pay more garbage fees. That’s the negotiable end of the loan business. You might be able to get a better deal, but you’re up against someone who can see all your cards and whose direct interests don’t fully coincide with yours.

    The simple fact is that its your cold hard cash which is up for grabs, race just doesn’t count. Caveat emptor.

    Black Jack (d8da01)

  11. Xrlq’s example wasn’t very persuasive to me on an ethical basis (I’m sure he’s right that it would be a legal problem though); if blacks are willing to pay more for loans and banks are willing to charge more for loans, it just sounds like capitalism in action. However, if the reason blacks are willing to pay more for loans is that they have a harder time getting loans (rather than just a differing set of priorities), then in this case it strikes me as unethical to charge them more on that basis. In that case, you aren’t exploiting the priorities of blacks to charge them more, you are exploiting some racial inequities in the system, and at the same time re-enforcing the racial inequities.

    (Still, while capitalism isn’t a guarantee of raical equality in the market, Patterico is correct that this study can’t be taken at face value.)

    Doc Rampage (47be8d)

  12. Doc: but wouldn’t it be discriminatory for a broker to give black customers better terms than they ask for, while not doing the same for white customers?

    steve sturm (b5aa23)

  13. From earlier testimony from this organization:

    NCRC has recently completed a Department of Housing and Urban Development Fair Housing Initiative Program (FHIP) Private Enforcement Initiative Grant. Through this initiative, NCRC conducted subprime fair lending testing of large lenders in six major metropolitan areas throughout the United States. The results provide detailed and vivid examples of disparate treatment and pricing in subprime lending based on race and gender.

    NCRC conducted forty-eight tests of 12 subprime lenders with retail outlets serving the metropolitan areas of Atlanta, Baltimore, Chicago, the District of Columbia, Los Angeles, and New York City. We conducted this national testing project with the assistance and cooperation of local NCRC members, community organizations, civil rights activists, and consumer protection organizations. (emphasis mine)

    I think these remarks may refer to a prior study by the NCRC, but if this one was conducted in a similar manner, the data could obviously be skewed by any bias of the individuals and organizations who are collecting the data. From the list provided here, I think there’s a good chance that they are biased.

    Regret (22cb76)

  14. I think it has been covered here, but banks, S&L’s, and other institutional lenders pretty much have a floor and a ceiling on rates, dependent on the type of laon (conventional, VA, FHA), conforming or non conforming (secondary market analysis). It really doesn’t matter if you are blue, black, green or purple, you’ll get the rate that you fall in the slot for.

    Now, mortgage brokers, they’re a different breed. Most mortgage brokers are paid on commission, based on a percentage of the income that the brokerage gets. The higher the brokerage income – the more they make. Brokerage companies do not fund their own loans and therefore do not do their own underwritimg. Some exceptions for FHA & VA loans where the brokerage has an “in house” approved underwriter.

    If you took a loan package to four different brokers and never pitted one against the other, you would probably end up woth four different rates or loan costs.

    I really don’t see much to be alarmed about without seeing the sources of the loans.

    RLS (0516f0)

  15. xrlq makes a good point. Price discrimination is one of the tools used by businessmen, particularly monopolists and oligopolists, to maximize profits, and where the price discrimination is based upon the lender’s perceptions of the economic acuity of the borrower, such a strategy may well have a disproportionate racial impact.
    But does this tell us anything about racial discrimination? Maybe, but I still think the answer is probably no, since the victims of discrimination suffer from a discrimination, not against their race, but against their perceived lack of financial sophistication. When car dealers stick it to women, the elderly, and professional basketball players by getting them to pay sticker price, are they guilty of sex, age, and athletic ability discrimination(s)? They’re being selfish bastards, but not racially discriminatory selfish bastards. I think there’s a difference.

    lincoln republican (0ab8c9)

  16. The bottom line is that we need to know more about the study. The article seems to leave out any and all information relevant to figuring out whether there is a problem here and, if so, what we might do to fix it. It appears to be fueled by the sponsor’s desire for publicity and a big legal settlement, which does nothing more than cast doubt on those few “controls” that are pointed out. That said, you can’t blame partisan organizations for conducting studies that advance their agendas, and trying to publicize them. But you can and should blame journalists that blindly reprint allegations of widespread and outrageous (if not illegal) behavior, without taking the time to dig a little deeper. Instead, we are left with yet another topic where, notwithstanding our having slogged our way through the story, we don’t feel we have enough to form a remotely intelligent opinion.

    James E. (631e78)

  17. I am always mistrustful of studies that report only percentages instead of actual numbers of people involved. OK, so 74% of white shoppers and 31% of minority shoppers were told about loan fees, but 74% of what? Thirty-one percent of how many? As James E. put it above, there isn’t enough information to form any kind of informed opinion, but that won’t stop many from assuming the worst, which, as Patterico would point out, is exactly that the Times wants.

    Rick (381fd7)

  18. There’s an obvious point. If mortgage brokers can charge non-white customers more, and make more money off them, then they would compete to get more non-white customers. Suppose the brokers stick black customers for an extra $500 each; then one broker would be able to get more black customers by overcharging only $300. He wouldn’t make as much off each one, but he’d get a lot more. Other brokers would respond in kind, and the differential would disappear.

    If it persists in the face of such competition, then there’s a reason.

    Mass ingrained prejudice? Possibly. Or maybe brokers tend to have a lot of bad experiences with black borrowers, such that none of them want the extra business.

    There’s a reason why cabbies don’t want to pick up single black youths, and it’s not prejudice. Liberals pass laws forcing cabbies to pick up anyone who waves – but they aren’t the ones who get robbed and murdered. There was a cabbie in Chicago who led protests against such a law – and he lived in a 90% black neighborhood.

    For a study like this to be conclusive, it would have to be conducted by a genuinely impartial organization, using impartial operatives. Then I would want to see follow-up interviews with brokers who appeared to demonstrate prejudice – with the content anonymized and the broker guaranteed against any sort of consequences for his comments.

    It might be useful to interview a lot of brokers beforehand, get a general attitude picture for each, and see if that predicted whether they had a
    discrimination pattern.

    Rich Rostrom (4e26d1)

  19. lincoln republican is wrong and xrlq is right, some discrimination is “rational” and will increase not reduce profit. Many stereotypes are true statistically and it may be profitable (although sometimes illegal) to act upon them. The issues are similar to those surrounding racial profiling in law enforcement which is also sometimes “rational”.

    James B. Shearer (48d074)

  20. Doc Rampage, it is not so much that blacks have a harder time getting loans, it is that they have a harder time getting people to offer them a good deal on loans (or any other negotiated price such as for a new car). Partly because of discrimination (rational and irrational) by vendors and partly because they are not as good at negotiating.

    James B. Shearer (48d074)

  21. In addition to the obvious flaws in this kind of “study”, there is the question of whether sufficient data are available on actual loans that have been written and have background information on the borrowers. By comparison to such a study, this is guerilla theatre. If there is racial bias in lending, the default rates of the groups subject to bias should be lower: if it’s harder to get a loan, the bar for qualifying will be higher, so the affected minorities will tend to have better credentials than whites. Of course, such a study wouldn’t be as much fun as playing dress up and set up.

    David (871a52)

  22. xrlq makes a good point. Price discrimination is one of the tools used by businessmen, particularly monopolists and oligopolists, to maximize profits, and where the price discrimination is based upon the lender’s perceptions of the economic acuity of the borrower, such a strategy may well have a disproportionate racial impact.
    But does this tell us anything about racial discrimination? Maybe, but I still think the answer is probably no, since the victims of discrimination suffer from a discrimination, not against their race, but against their perceived lack of financial sophistication.

    It depends on how they assess that. Assuming that a given mortgage broker does not have perfect knowledge of each individual customer’s maximum price, but does know that he can squeeze 100 extra basis points out of the average black customer, he might well use race as the sole criterion for deciding how much to charge everyone within a given credit profile. He won’t be right every time, of course, as some whites would have willingly paid the higher price, and some blacks will turn down the higher price who would have accepted the lower one. In the aggregate, however, he’ll make more sales, and average a higher profit margin per sale, than if he had quoted any one price to everybody.

    Think, by analogy, to auto insurance companies, who openly discriminate by sex (and surely would by race, as well, if it were legal to do so). It’s not because actuaries think penises (or pigment) cause accidents; it’s because penises (or pigment) correlate with risk factors that cannot be measured directly (i.e., there’s no easy way to predict which 17 year old boys will engage in high risk behavior, and which ones will drive like 17 year old girls).

    Xrlq (f52b4f)

  23. That is one philosophy: that the marketplace will eventually reward businesses that discriminate with the boot.

    It worked like a charm.

    actus (ebc508)

  24. Think, by analogy, to auto insurance companies, who openly discriminate by sex (and surely would by race, as well, if it were legal to do so). It’s not because actuaries think penises (or pigment) cause accidents; it’s because penises (or pigment) correlate with risk factors that cannot be measured directly (i.e., there’s no easy way to predict which 17 year old boys will engage in high risk behavior, and which ones will drive like 17 year old girls).

    Penises may coorelate with risk factors (although the stats are changing somewhat — not that boys are getting to be better drivers, but that girls are becoming worse!). Anyway, how would an insurance company determine “race”? It is illegal to ask such a question. If an agent indicated on an application in a remark that an individual was from a particular racial group, the underwriter would come down hard on the agent. Why? Because not only is it illegal (and unethical), but it also opens the company up to adverse action by the various state Departments of Insurance, who audit the companies. Believe me, those fines would more than make up for any “additional money” a company might make by doing this on a routine basis.

    [Personal experience — I came down hard on an agent about 10 years ago who told me (sotto voice) that the insured was black. And our company no longer does business with that particular agent.]

    In any case, in the 15 or so years I’ve been in insurance (13 as an underwriter), I have yet to see any statistics regarding risk factors associated with race. Race just isn’t a factor I consider in my underwriting decisions. It is immaterial to the exposure. Age is a factor, gender is a factor (for younger drivers), location is a factor, insurance score is a factor, type of vehicle is a factor, driving record is a factor, loss history is a factor. What do I need race for? (I can come up with a similar list for homeowner’s insurance.)

    Several years ago, a certain state wanted to see if there was discrimination by race in the auto insurance industry. So they asked for the racial make-up of the industry’s business. We esentially answered, “Huh, we don’t track that and don’t care about it.” So the state proposed to require the companies to ask about race so they could track it to see if we were discriminating. As an industry, we “laughed” at them. The state was asking us to violate the law to gather information about race so they could then use the information that we didn’t care about to penalize us for having it. Yeah, right! We managed to put a stake through that one’s heart. But I suspect it will come back one day.

    Anyway, long-winded response to XRLQ’s comment about the insurance industry. We get enough bad-press. Need to point out when we’re being used as an example. We don’t fit this one!

    Bill M (d9e4b2)


Powered by WordPress.

Page loaded in: 0.0708 secs.