The Effect of the First Wave of Automated Mortgage Underwriting Engines on Borrowers in Marginalized Groups

Abstract

Fannie Mae, a U.S. government-sponsored enterprise, is the nation’s largest source of conventional mortgage funds that was created for the purpose of providing liquidity to the mortgage market. In June of 1995, Fannie Mae released its first Desktop Underwriter, which was the first automated mortgage underwriting engine to exist in the mortgage industry. Studies have shown that automating processes have generally led to increased productivity, but I will use this 1995 automation event to specifically explore how the automation of the mortgage lending process has negatively impacted marginalized groups involved in the mortgage lending market, specifically marginalizations categorized by race, sex and socioeconomic status. Utilizing a differences-in-differences experimental model, I find that the majority of the marginalized groups were more likely to have their loan applications rejected after the lending process became automated, although there were varied results within the individual categories of race, sex, and socioeconomic status.