Brief Revisiting Automated Valuation Model Disparities in Majority-Black Neighborhoods
New Evidence Using Property Condition and Artificial Intelligence
Linna Zhu, Michael Neal, Caitlin Young
Display Date
Download brief
(521.68 KB)

Automated valuation models (AVMs) represent the promise of greater efficiency and lower costs for the mortgage industry. However, research has suggested that AVMs can produce racially disparate outcomes—namely, higher error as a percentage of value in majority-Black neighborhoods—that highlight the importance of technological equity. Potential inequities produced by AVMs may reflect data omission. They may also result from racial disparities in model inputs or from the modeling techniques AVMs use. In this brief, we build on our previous study by testing each of these possibilities. We find that gathering additional data on property condition and employing more sophisticated artificial intelligence techniques can help us more accurately assess the percentage magnitude of AVM error and its underlying contributors. But even with data improvement and artificial intelligence, we still find evidence that the percentage magnitude of AVM error is greater in majority-Black neighborhoods. This finding indicates that we cannot reject the role historic discrimination has played in the evaluation of home values. We also suggest more research exploring the dimensions of data and modeling to ensure the home-buying process benefits everyone seeking to achieve or maintain the American dream.

Research Areas Housing finance Housing Wealth and financial well-being Economic mobility and inequality Neighborhoods, cities, and metros Greater DC Artificial intelligence
Tags Homeownership Housing and the economy Fair housing and housing discrimination Housing finance data and tools
Policy Centers Housing Finance Policy Center
Cities Atlanta-Sandy Springs-Alpharetta, GA Washington-Arlington-Alexandria, DC-VA-MD-WV Memphis, TN-MS-AR