The study, which controlled for differences in income and loan amounts, found that African-American borrowers living in neighborhoods with the lowest share of non-white residents had a 24 percent chance of receiving a subprime loan. That number increased to 38 percent if the borrower lived in a neighborhood with the highest share of non-white residents.

For white borrowers in neighborhoods with the lowest share of non-white residents, the chance of receiving a subprime loan was 5 percent. Living in a neighborhood with the highest share of non-white residents increased the white borrower's chance of receiving a subprime loan to 18 percent.

The figures for Hispanic borrowers in neighborhoods with lowest and highest share of non-white residents were 14 and 31 percent, respectively.

The Furman study, called "The High Cost of Segregation: Exploring the Relationship Between Racial Segregation and Subprime Lending," also linked the rates of subprime lending to racial segregation at the national level.

In the national data, researchers found that an African American borrower was 11 percent more likely to receive a subprime loan if he or she lived in a highly segregated city. A black borrower in one of the least segregated cities had a 45 percent chance of receiving a subprime loan, compared to a 56 percent chance for black borrowers in highly segregated cities.

White borrowers were found to have the same chance of receiving a subprime loan, 16 percent, whether they lived in one of the most segregated cities or one of the least.

The national results are based on 2006 data reported by lenders under the Home Mortgage Disclosure Act, while the New York City conclusions come from HMDA data over a four-year period, from 2004-2007. 

The High Cost of Segregation Furman Center Policy Brief November 2009