Research Question: Prior research suggests that white families with children tend to avoid moving to homes in (otherwise similar) school attendance areas when those schools enroll a higher proportion of minority students. Does the incorporation of algorithmic school ratings into online housing market search tools weaken or maintain the effect of aversion to racial minorities in white families’ neighborhood demand?

Data: Analytic data for the study are based on a unique linkage of housing transactions from Zillow Inc.’s Ztrax database, HMDA mortgage origination data, and school assessment data from the U.S. Department of Education.

Methods: Matched difference-in-difference regression design.

Challenges: Small differences in the reporting of residential property transactions across states, localities, and time complicates the process of linking Ztrax and HMDA data at the property level.

Findings: The analysis is still in progress. Based on theories of algorithmic reactivity, we expect that local residential purchase rates will become increasingly responsive to the school ratings following their release in Zillow’s web portal. We expect that this will result in increased racial sorting across school attendance areas overall, but slightly diminished racial place discrimination by white buyer between school attendance areas that are otherwise similar except for their ethnographies-racial composition.