Investigation Finds Freddie and Fannie Algorithm Negatively Impacts Borrowers of Color
An investigation by The Markup found that lenders in 2019 were more likely to reject loan requests from people of color than white people, even if their finances looked much the same.
The Markup examined more than 2 million conventional mortgage applications using statistical analysis that held 17 factors steady to account for explanations traditionally used to explain racial disparities in lending: the borrower’s credit history, debt-to-income (DTI) ratio and loan-to-value (LTV) ratio.
According to The Markup, the bias was evident in the credit scoring algorithm “Classic FICO,” which was developed in the 1990s. Fannie and Freddie require lenders to use Classic FICO to determine whether an applicant meets their minimum threshold. Classic FICO focuses on traditional credit but does not take into account things like on-time payments. It also penalizes people for past medical debts, even if they’ve since been paid.
As a result, the model is considered beneficial for white people who have more access to traditional credit while negatively impacting people of color.
“The mortgage approval process is rooted in a racially unjust history that persists to make homeownership a far more difficult dream to achieve for many Black Americans. Owning a home is a major way to generate, keep and pass down wealth, and unequal access to mortgages only serves to further entrench inequality,” said Zillow economist Joshua Clark. A Zillow analysis of lending came to a similar conclusion earlier this month.
The report showed that compared to white applicants in similar financial circumstances, lenders were far more likely to deny people of color. The national breakdown is as follows:
- 80% more likely to reject Black applicants
- 70% more likely to reject Native American applicants
- 50% more likely to reject Asian/Pacific Islander applicants
- 40% more likely to reject Latino applicants
The issue of racial disparity in mortgage lending has been the subject of similar reports in the past. In 2018 the MBA responded critically to a Reveal News report with similar findings to The Markup investigation, saying the report was flawed due to the fact that it only took conventional mortgage applications into account and did not make allowances for some of the factors included in the Markup’s analysis.
“This is how structural racism works,” National Consumer Law Center attorney Chi Chi Wu said about the algorithm. “This is how racism gets embedded into institutions and policies and practices with absolutely no animus at all.”