How Automated Home Appraisal Mistakes Can Affect Black Homeowners
Accurate home appraisals are essential not only for obtaining mortgage financing, but also for establishing a homeowner’s level of home equity.
Having previously shown that in-person assessments are subject to bias and discrimination, Urban Institute research associates recently conducted a study to understand how automated home ratings – which have been used with the COVID-19 pandemic – work for various racial groups.
They say the data, when analyzed, shows that automated home appraisal technology errors disproportionately affect residents of predominantly black neighborhoods.
“Automated valuation models (AVMs) are a type of financial technology that can be an alternative or supplement to in-person valuations,” said Michael Neal, Linna Zhu and Caitlin Young, who wrote the report. “Many lenders use MAVs as part of their loan decision process. MAVs apply mathematical algorithms to a database of housing activity, including sale transactions, to calculate a home’s value. In theory, by limiting the element of human valuation and potential racial bias, AVMs could provide a more accurate picture of a home’s value in predominantly black neighborhoods, which could hold promise for closing the gap. of racial wealth. “
Digging into data from Atlanta, Memphis, and Washington, DC, the authors set out to determine if and how AVMs contribute to racial disparities in home value estimates. Researchers investigated the accuracy of MAVs in predominantly black neighborhoods compared to predominantly white neighborhoods by comparing the MAV estimate with the actual selling price.
The researchers first found the magnitude AVM error was greater in predominantly white neighborhoods, however, after adjusting for the selling price, which is lower in predominantly black neighborhoods, they found that the magnitude in percentage error was higher in predominantly black neighborhoods.
“This means that the absolute dollar value error may be smaller in these neighborhoods, but the proportional size of the error is larger because the average price of homes in predominantly black neighborhoods is much lower,” noted authors. For example, the percentage difference between an AVM estimate of $ 100,000 and a sale price of $ 115,000 (15%) is greater than the percentage difference between an AVM estimate of $ 400,000 and a sale price of. $ 415,000 (4%), even though the absolute difference ($ 15,000) is the same. “
The full report is available at urban.org.