Statistic Static: Bias Studies 'All Over the Board'

WASHINGTON - For Cathy Bessant, the fact that a study released last week cast doubt on an earlier report of redlining in the nation's capital proves one point: Statistics can be misleading.

"It is clear that we are seeing studies all over the board," said Ms. Bessant, NationsBank's Community Reinvestment Act guru. "It is like we are seeing the study of the week. I think this proves the theory that statistics can show what we want them to show."

Ms. Bessant said its time to move away from studies and back to loan file and bank policy reviews.

'All that these statistical studies do is neutralize each other," she said

The study in question, conducted by Office of Thrift Supervision staff economists Clifford Rossi and Fred Phillips Patrick, re-examined a study The Washington Post ran in June 1993.

The newspaper's study, which Attorney General Janet Reno said provided the impetus for the now famous fair-lending prosecution of Chevy Chase Federal Savings Bank, found that banks and thrifts make twice as many loans in white parts of the city as they do in minority areas.

The OTS researchers, when they accounted for other variables such as the percentage of boarded-up properties in the neighborhood and the unemployment rate in the area, found that the disparities disappeared.

"There was no evidence of pervasive redlining," Mr. Rossi said as he gave the Society of Government Economists a preview of the study last Tuesday.

"I think what we found is that there are economic factors that lead to the observed findings," he added.

Reporter Liz Spayd, who contributed to the Post articles, said the OTS study "flies in the face" of other research in the field.

John Taylor, president of the National Community Reinvestment Coalition, also criticized the report, saying lending decisions shouldn't depend upon unemployment rates or boarded-up properties.

For reprint and licensing requests for this article, click here.
MORE FROM AMERICAN BANKER