How Racial Bias Can Materialize in Residential Appraisal
As part of a webinar that we hosted along with NAMMBA, our CEO Brian Zitin had the opportunity to interview Jillian White, head of collateral at better.com about how bias can manifest in the residential appraisal world. The conversation was wide-ranging and included how the estimate of value provided by the appraiser and the very definition of an appraisal is an opinion.
But when can that professional opinion turn into a personal bias? Brian and Jillian broke it down three ways.
The first is how an appraiser selects their comparables, otherwise known as “account selection.” Here’s how that happens. An appraiser will look at comparable sales to determine the value of a home. Let’s say there are 15 sales that happened recently that the appraiser might take a look at and consider. In this scenario the appraiser can choose to throw out five because they’re not truly comparable — either because the sales are too dissimilar from the subject property or anybody who’s interested in buying the subject property might not be interested in buying one of those other houses. It is ultimately at the appraiser’s discretion to make the determination about which comparable sales to include to enable the report. That means we’re dealing with opinions — so one appraiser’s opinion might differ from another appraiser’s opinion.
The second area where opinion determines the outcome is when it comes time to make the adjustments to the comparables. In residential real estate appraising, an appraiser will look for homes that have sold that are as comparable to the property being appraised as possible. Ideally, it is desirable to use sales that are so similar to the property being appraised, that no adjustments are needed. However, in many areas, it is difficult (if not impossible) to find three or more sales that are nearly identical to the property being appraised. Therefore, adjustments are made. When the appraiser is doing their analysis to find out how much they should make the adjustment, it is not a pinpoint precise number. Typically it’s a range — this market seems to pay anywhere from $5000 to $15,000 more for an additional bedroom. So again, it’s at the appraiser’s discretion to determine where within that range they’re going to make their adjustments. There are huge implications if they use the lower end of the adjustments versus the higher end.
The third area where an appraiser’s opinion creeps in is in the reconciliation of the value. Real estate valuation reconciliation is the process through which appraisers take values derived through different techniques and/or different people, and review the different figures to arrive at a final estimate of value. The appraiser will have all of these adjusted values, and sometimes it can be a large range — in some instances anywhere from $350,000 to $400k to $450k. Again, it’s up to the appraiser to reconcile that and decide which comparables to use.
According to Jillian, between the comp selection, the adjustments, and the reconciliation of value “there’s a lot of room for artistry and not just a strict science.” That’s precisely why an individual’s potential bias can show up in the appraisal process, according to White.
Still, there is another layer of bias that could exist in the appraisal that is less talked about and less affected by the individual appraiser’s opinion, and that is structural bias. Whereas the individual bias of an appraiser is an unconscious bias and will vary on a person-by-person basis; structural bias has to do with the system that the individual operates within.
In appraising, an example of that is redlining. Jillian explains that redlining was the practice of literally drawing a red line around black neighborhoods and saying do not lend here and white neighborhoods have green lines indicating that these borrowers could get mortgage financing.
The practice was outlawed in 1968 but the housing market continues to be plagued by the remnants of redlining. “If you live in an area that at any point in its history was redlined, or another way of putting it, if you live in an area that at any point in time, was a predominantly black neighborhood, then your property values would not have appreciated at the same rate as those of white neighborhoods,” Jillian said.
That structural bias means that an appraiser, even the most unbiased of them, would still have to pick comparable sales that sold recently in the neighborhood. If those property values have been suppressed as a result of redlining, that is a bias that can’t be resolved by the individual because it has nothing to do with the individual, it has to do with the system. In such cases, the sales comparison approach “propagates what was done in the past and the gap can never be bridged while using the sales comp record.”
To that point, a 2018 study from Brookings looking at the devaluation of assets in black neighborhoods concluded that homes in the majority-black neighborhoods were undervalued by 20% compared to identical or similar neighborhoods that are predominantly white.
Other research such as this AEI study looks to use big data to understand how common appraisal racial bias really is. They concluded, based on the data available, that racial bias by appraisers on refinance loans is uncommon and not systemic.
The bottom line is that tackling individual bias won’t move the needle on the structural bias, and vice versa. There are also many other factors that influence valuations that should not be overlooked. As an industry it will be vital to continue conversations like this one, and to openly share opinions, perspectives, data, and research to help educate and level the field for the future.