Under-valuations that more accurately reflect the homes’ “true” value as opposed to the contract price will also alert the buyer, not just the lender, that he or she may be over-paying, which often triggers a renegotiation… when the seller and buyer settle on a new price after the appraisal, the new lower price reduces credit risk, costs to the borrower, and ultimately results in greater wealth for the buyer.
The AEI Housing Center recently released an analysis revealing that reports by the Federal Housing Finance Agency (FHFA) and by Brookings, attributing the greater prevalence of under-valuations in home purchase appraisals to appraiser racial bias, suffer from the same pitfalls as Freddie Mac’s studies. All these reports arrive at the same premature and potentially flawed conclusion that suggests appraiser racial bias as the sole explanation for differences in the share of under-valuations across census tracts.
Exploring Alternative Explanations for Appraisal Under-Valuation
A critique of FHFA, Brookings, and Freddie Mac reports on Racial Bias
The newly-released Uniform Appraisal Data (UAD) Aggregate Statistics Data File and Dashboards, which are derived from more than 47 million appraisals conducted between January 1, 2013, and June 30, 2022, provide the number and share of homes appraised below the contract price at the census tract level. These new data were the impetus for a blog post by the Federal Housing Finance Agency (FHFA) and a report by Brookings seemingly connecting the greater prevalence of under-valuations in home purchase appraisals to appraiser racial bias.
However, their methodology suffers from the same pitfalls as Freddie Mac’s 2021 research note “Racial and Ethnic Valuation Gaps in Home Purchase Appraisals” as well as its follow-up study “Racial & Ethnic Valuation Gaps in Home Purchase Appraisals – A Modeling Approach” which used Freddie Mac’s proprietary data. The former was later cited by the Interagency Task Force on Property Appraisal and Valuation Equity (PAVE) task force and others to justify far-reaching policy actions regarding the appraisal process. All four reports arrive at the same premature and potentially flawed conclusion that suggests appraiser racial bias as the sole explanation for differences in the share of under-valuations across census tracts.
Our critique points out some significant shortcomings of these studies. Notably, they focus on entire neighborhoods, when they should study the actions of individual appraisers. They use the greater share of under-valuations as evidence of racial bias, when they should consider explanations unrelated to bias that might account for under-valuations, including for example the greater share of first-time homebuyers, who tend to overbid, or the greater presence of seller concessions, which will reduce the appraisal amount when they are properly accounted for. They also fail to note the size of these under-valuations. Using the Aggregate Statistics Data File and Dashboards, our analysis indicates they are relatively small, averaging about $1,100 to $1,900. These levels are too high if they are in fact due to racial bias. However, our analysis finds multiple other plausible explanations for under valuations of this magnitude such as the greater presence of first-time homebuyers or seller concessions. We also note that under valuations of this magnitude of are unlikely to depress entire neighborhoods and, they may in fact provide a disproportionate consumer benefit to minority homebuyers.
We conclude with a renewed call on Fannie Mae, Freddie Mac, and regulators such as the FHFA to mass screen individual appraisers for racial bias and inaccuracies. We have already laid the groundwork for this research with a published working methodology.