WASHINGTON - A new Federal Reserve Board study takes aim at the widely held view that banks discriminate against marginally qualified minority loan applicants.
The study contends the exact opposite is true - banks are extending more loans to risky minority borrowers than to comparable white applicants.
The study is one of a handful of recent reappraisals of lending discrimination that challenges the conventional wisdom that banks discriminate against black Americans.
Because it plays into the views of the Republicans who now control Congress, it has the potential to change the way regulators and lawmakers look at fair-lending rules.
Behind the study, conducted by four government economists, is a simple idea about loan default rates and marginal borrowers.
If bankers discriminate, the researchers argue, then default rates should be lower among blacks and other minorities. That's because marginal borrowers - those most likely to default - would be rejected more often if they hail from minority groups and approved more often if they are white.
On the other hand, if lenders do not discriminate, default rates should be roughly equal among all groups.
When the economists examined loan data, they discovered that black borrowers defaulted more frequently. From that observation, they concluded that banks could not be discriminating by forcing black borrowers to meet more stringent financial conditions.
"This is consistent with the notion that at least within (Federal Housing Administration), there isn't any discrimination against minorities," said Anthony Yezer, a George Washington University economist who focuses on fair-lending issues. "When you discriminate against somebody, you ought to make more money off them, not less."
Federal Reserve Board economists Glenn B. Canner, Freddie Mac economist James Berkovec and their two colleagues based the study on 220,000 FHA loans made from 1987 to 1989. Both the Fed and the Department of Housing and Urban Development underwrote its cost.
The study found default rates for blacks were 2% to 4% higher than default rates for whites. Default rates for other minorities were the same as default rates for whites.
Mr. Canner said he also found that losses to banks on loans extended to blacks who live in predominately black neighborhoods were greater than the losses on defaults to whites in predominately white neighborhoods.
The study reaffirms the industry's view that it does not discriminate, said James McLaughlin, director of agency relations at the American Bankers Association.
"It is another little piece to help the industry prove a negative, which is almost impossible to do," Mr. McLaughlin said.
The study supports similar findings from the early 1980s.
"They did it on new data," Mr. Yezer said. "But what they are getting today is exactly what people have gotten in the past, including myself."
Mr. Canner acknowledged one major caveat: the study's technique cannot be used to clear individual banks of fair-lending charges. That's because there are not enough defaults at any single institution to create a statistically valid sample size.
D. Jean Veta, a partner at Covington & Burling, said bankers should not relax. Any number of groups can produce different studies to buttress their particular arguments, she said.
In fact, this is one of a number of studies released within the past six months. Several civil rights and community groups have released reports analyzing Home Mortgage Disclosure Act data. They have found wide disparities between how banks treat minority and white loan applicants.
And, late last year, Mr. Yezer released a study showing that the formula used to prove discrimination can show fair-lending violations where none exist.
While the Fed study may not exonerate individual banks, the report, released just after Republicans took control of Congress, provides community reinvestment foes with added ammunition, industry analyst William M. Cunningham said.
"I think it is bound to have an impact on policy," Mr. Cunningham said. "What it will say is that while it is nice to have these ideas that you can do things for minority groups, the evidence shows that whatever you do doesn't work."
He said the study fits the Republican pattern of criticizing government programs that don't achieve their goals, such as a welfare system that encourages dependency or affirmative action initiatives that harm whites. Now, it is logical for foes to say that community reinvestment programs cause banks to lose money, he said.
"Don't get me wrong," Mr. Cunningham said. "I don't believe any of that. But, the potential is there for those types of arguments to be used. The timing is just all wrong for this type of study to come out."
The study questions the usefulness of any report based upon rejection rate disparities, Mr. Cunningham said.
"What this study says is (disparities in rejection rates) may not be an indication of discrimination," Mr. Cunningham said. "That may just be bankers acting to protect their shareholders."
And one community activist said the report misses the point. Anyone who travels to a low-income or a minority community quickly realizes that these people do not have equal access to credit, National Community Reinvestment Coalition president John Taylor said.
Studies such as this do nothing to solve the credit problem, Mr. Taylor said. "We need to figure out why some institutions do a much better job at serving the needs of traditionally underserved people and why some institutions do a horrible job at it," Mr. Taylor said. "
But American Bankers Association economist Michael A. ter Maat said Mr. Canner's and Mr. Berkovec's reputations as meticulous researchers make this study material.
"This paper is important enough to get everyone's attention," Mr. ter Maat said. "This is not a paper to be treated lightly."
The study does have some limitations, Mr. Cunningham said. The biggest problem, he said, is the lack of data from past years.
"I think part of the problem is that they are looking at a very short time period and they are looking at loans that defaulted very early in the loan life," Mr. Cunningham said. "That might bias the answer."
Mr. Canner said using FHA data, as opposed to conventional loan data, shouldn't matter. The reason: if discrimination exists, then bankers will reject for conventional loans qualified minority applicants. This should flood the FHA program with qualified minority borrowers, which lower, not increase, default rates.
But, Mr. Canner readily concedes the limitations of his study. First, he said it assumes bankers discriminate only against the least qualified minority applicants. If bankers discriminate equally, regardless of income, then the results would be skewed, Mr. Canner said.
