A recent study by the National Bureau of Economic Research (NBER) found that discrimination in mortgage lending declined between 2009 and 2015, and automated or algorithmic loan offerings from financial technology firms may play an important role in that trend.

The NBER found that face-to-face lenders charge Latinx and African-American borrowers 7.9 basis points more for purchase mortgages and 3.6 basis points more for refinance mortgages than comparable borrowers—costing minority borrowers $765 million per year in extra interest. Online algorithmic lenders also charge minority borrowers more than others—5.3 basis points extra for purchase mortgages—but the amount of discrimination is reduced by 40% when compared to face-to-face lenders, according to the study. 

In-person lenders reject minority applicants about 6% more often than comparable non-minority applicants, while algorithmic lenders show no discrimination in loan rejection decisions, according to the study. The NBER calculated that between 740,000 and 1.3 million minority loan applicants were rejected by in-person lenders from 2009 to 2015 who would have otherwise been accepted, also resulting in reduced profits for lenders.

The premium paid by minority buyers is consistent with in-person lenders and algorithmic loan offerings charging borrowers more in markets with less competition or identifying borrowers who are less likely to shop around, according to the study. However, competition from financial technology firms and the increased ease of shopping for mortgages online may help reduce bias in the lending market, the study found. 

Graph: The Racial Discrimination Premium in US Home Loans, 2009-2015 from the National Bureau of Economic Research