WASHINGTON -- A study by the Federal Reserve Bank of Boston found that the extent of racial bias among mortgage lenders has been exaggerated, but confirmed that discrimination is a serious problem.
The study took a comprehensive look at many variables that affect mortgage lender's decisions in the Boston area, seeking to isolate when race appeared to be the key factor in rejections. Many of those variables were not examined in analyses of data released for the first time last year under the Home Mortgage Disclosure Act.
Lenders had hoped they would be exonerated by a more sophisticated analysis. But while the Boston Fed's basic conclusion is more favorable -- black and Hispanic applicants were 1.6 times as likely to be turned down than whites, compared with the HMDA survey's 2.7 times -- the new findings reinforce the notion that discrimination is entrenched.
"The results indicate that a significant problem exists in the market for mortgage loans," said Richard F. Syron, president of the Boston Fed. "Unfortunately, race plays a role, perhaps an unconscious and unintentional role, but a role nonetheless in mortgage lending decisions."
The Boston Fed study sought to correlate important financial, employment, and property characteristics with loan decisions. It thus digs deeper than the raw HMDA data to present a more thorough picture of why an application might be rejected.
Even though it has fewer statistical disparities than the national survey, the Boston study's strong evidence of bias should refute many bankers' criticisms of the HMDA analysis.
"The debate concerning whether disparate treatment of minorities is occurring in credit markets should be over," Federal Reserve Board Governor John P. LaWare told participants Thursday at a conference on credit and the economically disadvantaged in Denver.
"This may be a bitter pill, especially for those who believe that their institutions treat all applicants for credit equally, regardless of race," he added. "But frankly, it would be too much to assume that attitudes about race held by some in our society do not seep into the lending process."
Credit History, Risk Profile
In the Boston study, researchers asked 131 financial institutions for data on 38 variables not included in the HMDA surveys covering 1990 mortgages. These included debt and credit histories, loan-to-value ratios, risks of default, employment histories, and neighborhoods.
Of all the variables examined, only history of bankruptcy and denial of private mortgage insurance turned out to be more important predictors of loan denial than race, said Alicia Munnell, principal author of the study.
"We're not surprised," said Ronald A. Homer, chairman of Boston Bank of Commerce, a minority-owned bank. "This really just helps to uncover how our system works a little better. Unfortunately, we don't live in a color-blind society."
Factors in Loan Denial
The study found that minority applicants, on average, have greater debt burdens, higher loan-to-value ratios, and weaker credit histories. They are also less likely to buy single-family homes. These factors account for about two-thirds of the difference in denial rates.
But no factor other than race could count for the remaining gap, researchers found.
"This means that 17% of black or Hispanic applicants instead of 11% would be denied loans, even if they had the same obligation ratios, credit history, loan-to-value, and property characteristics as white applicants," the study said.
Most Borrowers Fall Short
The researchers also found that minorities with unblemished credentials are 97% certain of being approved. But the majority of borrowers - both white and black - have some problems with their applications, and here race can have a big impact.
"The results of this study suggest that for the same imperfections, whites seem to enjoy a general presumption of credit-worthiness that black and Hispanic applicants do not, and that lenders seem to be more willing to overlook flaws for white applicants than for minority applicants," Ms. Munnell said.
The model created for the Boston study could have many additional applications, Ms. Munnell said. Bank examiners and even banks themselves could use the model to better detect discrimination, she said.
It has already been put to broader use. After analyzing data on loans, the Boston Fed provided regulatory agencies with lists of individual cases that their model showed could have involved discrimination.