In an article on page two of today’s Washington Post, staff writer Kirstin Downey reports on “disparities” in sub-prime housing loans, those loans given to the worst credit risks at a higher than average interest rate. In what I think is an egregious but not uncommon error in articles of this type, she writes:
Lender groups do not dispute that there are disparities based on race and sex, but say borrowers pay higher rates if they pose a higher credit risk for any one of a variety of reasons, such as low income, poor neighborhood, bad credit history or low down payment.
Really? In that case, show me a single lender group, or a single spokesman for any lender group, who does “not dispute that there are disparities based on race and sex” (emphasis added). If a reporter can’t tell the difference between “based on” and, say, “associated with” or something similar, she shouldn’t be covering stories involving race. It’s one thing for “[a]dvocacy groups such as the National Community Reinvestment Coalition” to assume and assert that every “disparity” is based on discrimination, but it’s quite another for news reporters to incorporate such assumptions in their writing.
William Raspberry’s column in today’s WaPo also discusses the racial gap, much more intelligently. To his credit, he says that more attention should be devoted to examining how successful blacks have succeeded rather than continuing to blame whites for all those who have not.
Look at those who have flowed through those widened gates [of opportunity], and you will see an unusual commitment to education, to civic involvement, to clean records and sterling reputations, to long-term thinking, to responsible parenthood.
Along the way, however, Raspberry does miss one bit of unintended humor in the National Urban League’s new report on the State of Black America. Discussing the Urban League’s new “Equality Index,” Raspberry writes that it is
designed to measure everything from health and employment to civic engagement and volunteerism puts blacks at 73 percent of parity. But both the general finding and the recommended solutions are the same as they have been for as long as I’ve been following these reports: The black/white gap shows no signs of closing, and in some ways may be worsening, and the general society — the government — needs to do more.
So, if there is a “volunteerism” disparity, perhaps “the government” should draft more blacks to volunteer? Oh well, Raspberry is good enough that he can be excused for missing opportunities like this.
More seriously, Raspberry quotes Urban League chief executive Marc Morial’s statement that, despite progress,
many have been left behind…. [O]ne out of every four black Americans lives in poverty, and almost half of those who live in poverty live in extreme poverty.
But if this is true, to go back to the first article I discussed, how can housing advocates claim that a “disparity” in sub-prime loans “corroborates [their] long-held belief that lenders discriminate”? Does it believe that a community with 25% of its members living in poverty, half of whom live in extreme poverty, should receive good mortgages at the same rate as communities with a much smaller percentage of its members living in poverty?
Of course, if all you have is a hammer/disparity, then everything you see looks like a nail/discrimination.
I like how the lenders say the poor terms are due to risk, but are not turning over scores that would allow us to verify that. Cuz the lenders care about privacy.
actus, it ought to be extremely easy to check this. If lending practices are not in fact tracking risk, but are instead discriminating arbitrarily by race, then black default rates will be lower than white default rates, because black borrowers will have been held to higher standards than white borrowers were. I seem to recall one or more studies some years back that tried to find evidence of this effect and could not.
THE CASE (AGAIN)OF THE MISSING ASIANS:
Oh great, another liberal diatribe against the evil white man discriminating against the poor minorities. But…aren’t Asian Americans minorities??? If there is discrimination by the evil white-dominated banks of America, they’re actually discriminating in FAVOR of a minority group, Asians. But, of course, that obvious point (shown by their graph showing that the “sub prime” rates) will never be acknowledged in the mainstream media since it contradicts their agenda.
Asian 3.8%
White 10.4%
Hispanic 15.2%
Black 29.4%
Thank god all these asians are in the USA, or else more people would buy into the pathetic excuses of all of the race-card hustlers who say that the “white power structure” makes it impossible for non-white people to succeed in America.
THE CASE OF THE (AGAIN)MISSING ASIANS
Great, another article from the mainstream media about white establishment “discriminating” against “minorities”. But like the articles on alleged discrimination in college admissions, the word “Asian” never seems to be mentioned. I guess they’re more “white” than all the “minority” “Hispanics” , some of whom are blonde, blue-eyed Caucasions.
But if you look at the graph for sub-prime loans:
Asian 3.8%
White 10.4%
Hispanic 15.2%
Black 29.4%
the conclusion would be that the banks are discriminating against WHITE people in favor of ASIANS. But that doesn’t fit the liberal media’s agenda, so it’s just overlooked.
Besides, any elementary student knows that to find a cause-effect correlation in a science experiment, you need to isolate all the other variables (in this case, poor credit, bad records for paying bills on time, troubled neighborhoods with high foreclosure rates and falling home values, etc)It’s basic economics, but the race-hustlers don’t know basic elementary school science, so they probably don’t know basic economics either.
“If lending practices are not in fact tracking risk, but are instead discriminating arbitrarily by race, then black default rates will be lower than white default rates, because black borrowers will have been held to higher standards than white borrowers were.”
It goes the other way too. If they’re not being denied credit, so much as being denied cheap credit. So they face higher payments than their credit risk tells us they should, and thus might have higher defaults.
actus, so compare black and white debtors with the same payment rates, and see which cohort defaults more often. If they are the same, I think the lender has taken equal risk with each cohort.
I don’t really understand your point. Do you think lenders want more defaults? Surely if there are good risks to be had in the African-American community, it would be silly of businessmen to force them to pay rates that they couldn’t sustain.
“actus, so compare black and white debtors with the same payment rates, and see which cohort defaults more often.”
sounds good. lets ask the lenders for the data, hopefully they won’t object like they do with teh credit scores.
“I don’t really understand your point. ”
my point is the proper way to figure if the disparate lending is due to disparate creditworthyness is to look at creditworthyness data. And for some reason, must be privacy, the lenders don’t want to give that up.
actus,
my point is the proper way to figure if the disparate lending is due to disparate creditworthyness is to look at creditworthyness data.
No, the “proper way” is to see whether different risk cohorts all result in roughly the same ratio of money paid in to money paid out. If two cohorts with the same “creditworthiness” stats nonetheless default at different rates, the obvious conclusion is that your “creditworthiness” measures are missing something. What you want is the same failure rate across the board — small enough that you can eat it, large enough that you don’t lose all your clients — and if you aren’t getting it, you are missing some risk factor somewhere.
I repeat: Look at black and white cohorts paying the same rates. If the black cohort defaults significantly less than the white cohort, you have strong circumstantial evidence of discrimination. If not, you don’t.
Look, suppose you’re a careful driver who’s never so much as run a yellow light. If you are male, 22, and own a flashy sports car, you’re nonetheless going to pay in insurance for things you never did. And if you suspect that the insurance companies just have it in for young men with flashy cars, the proper way to check would be to compare the money in vs. money out for insurance claimants in that cohort, vs. money in/money out for other cohorts. If the ratios are all pretty much the same, no one is being gouged.
“No, the “proper way” is to see whether different risk cohorts all result in roughly the same ratio of money paid in to money paid out.”
So banks expect the same profitability / return accross all risks? Thats weird. I wouldn’t have thought.
actus,
So banks expect the same profitability / return accross all risks? Thats weird. I wouldn’t have thought.
No? I would imagine that lenders operate very much like insurers — trying to minimize risk and maximize return. I would think that they would try to equalize the expected return as percentage of risked capital across investments, but then I’m not a banker, so I might be wrong. Possibly you make more money by gouging your rich customers. I cannot really imagine that you make more money by gouging your poor ones — partly for the very reason you cited earlier, that usurious rates cause more defaults. If you’re a lender, are you in favor of more defaults?
How can something be egregious and not uncommon? Egregious MEANS uncommon!
It comes from ex (out of) and grex (flock).
Sheesh, after all the comma stuff…
“If you’re a lender, are you in favor of more defaults?”
Ceteris paribus, no. But if a certain policy will cause more defaults but also more returns, and those returns are better, then you would be, in some sense, “in favor of more defaults.”
I still think its odd that cohorts more likely to default are giving the same return as the less likely. Usually with higher risk (variance) we expect a risk premium, on top of the expected values.