Mortgage seekers from minority groups may pay more in fees than similarly qualified white borrowers, according to a team of researchers.
In a study, the researchers found that when minorities seek mortgages they pay about 8 percent more -- or $400 more -- than white borrowers when they seek loans from white mortgage agents. Mortgage agents can assess fees, such as the broker origination fee, which are negotiable, or can even be waived.
Earlier studies had tended to focus on the race of the borrower, not the broker, according to Brent Ambrose, Smeal Professor of Real Estate, professor of risk management and an associate of Penn State's Institute for CyberScience (ICS). However, because the dataset, which contained information from more than 25,000 different mortgage brokers, included broker last names, the researchers were able to use a statistical method to infer the broker's race based on their last names and give them a better understanding of the role race played on both sides of the mortgage-buying process. By matching the brokers' surnames with a list of last names from the U.S. Census Bureau, the researchers were able to infer race and ethnicity. The method is similar to the ways that regulators and courts infer race and ethnicity.
"All the previous studies that have looked at race and mortgage lending have only been able to observe the race of the borrower, so it's never really clear what's driving the results," said Ambrose, who worked with James K. Conklin, assistant professor of real estate, University of Georgia, and Luis A. Lopez, doctoral student in business administration, Penn State.
The researchers, who presented their findings at Singapore Management University's Conference on Urban and Regional Economics in December, also found that minority borrowers do not receive preferential treatment when they seek loans from minority brokers. Whites working with minority brokers pay about the same fee as their minority counterparts.
To determine whether race was correlated with the fees paid by the borrowers, the researchers controlled for several other factors using data provided in the mortgage data, including financial literacy, fee transparency and the selection process of both borrowers and brokers. None of these factors had a significant impact on pricing the fees.
The researchers also investigated what was causing the discrimination. They tested whether the white brokers demonstrated overt racism -- or animus -- toward minority groups, or whether it is the result of statistical discrimination, a form of racial profiling.
In a first test, the researchers examined whether fees paid by minorities varied across credit scores, which would be a sign of statistical discrimination. In the second test, they investigated whether the use of Google search terms with racially charged language in specific areas correlated with preferential loan treatment in those areas. The results from both tests were inconsistent with the hypothesis that the observed differences in fees arose from animus on part of the white brokers.
According to Ambrose, the results suggest that statistical discrimination is behind the higher mortgage fees, rather than actual animosity toward minorities.
Brokers have a high degree of discretion on the fees they assess, which can cause borrowers to pay more, or less for their mortgages, said Ambrose, who also serves as the director of the Institute for Real Estate Studies. For example, they can assess front-end fees, such as application fee, underwriter fee, mortgage brokerage firm fee and points. These fees are usually paid at closing. There are also back-end fees, including yield spread premium and correspondence premium.
The researchers also examined how recent regulations of the mortgage industry may affect the ability of brokers to charge different fees. The results suggest that regulations put in place following the Great Financial Crisis did reduce the ability of brokers to vary fees across borrowers. However, the researchers also found that this outcome comes at a cost in the form of reduced access to credit.
"There's a cost and benefit to regulation, which is the point we're trying to make," said Ambrose. "The good side is you reduce disparities, but the bad side is you generate potential credit rationing."
The researchers used data from applications for brokered, first-lien, residential loans approved by New Century, a subprime lender, between January 2003 and March 2007. Computations for this research were performed on the Penn State Institute for CyberScience Advanced CyberInfrastructure (ICS-ACI).
The work was also supported by Penn State's Institute for Real Estate Studies.