The appraisal gap - by Albert Franke

The appraisal gap. We’ve been hearing a lot about this lately. What is it? The appraisal gap is the difference between a property’s appraised value and its contracted sale price. The term is commonly used when the appraised value falls short of the sale price. It is now being linked to allegations of racial bias in residential appraisal.

The Federal Home Loan Mortgage Corp., better known as Freddie Mac, recently completed a study and concluded that minorities are more likely to receive an appraised value that is lower than the contract price in a purchase transaction than non-minorities or whites. This article aims to provide the reader with an interpretation of Freddie’s study, rather than to take a position on the controversial topic of racial bias and whether it exists systemically in appraisal.

Freddie Mac, in its analysis, defines appraisal gap as “the percent difference in the share of properties or applicants receiving appraisal value lower than contract price between minority and White groups.” Freddie examined more than 12 million single-family property appraisals submitted to it between January 1, 2015 and December 31, 2020. The neighborhoods, or tracts, were delineated as census tracts using population, race and ethnicity data from the 2010 census. The study concluded that overall, 8.3% of appraised values fell below the contract price during this period. In white tracts, 7.4% of appraised values in a transaction came in below the contract price. For latinos, the percentage increased to 15.4% and for blacks, the percentage of appraised values falling short of the contract price was 12.5%. The 8% difference between whites and latinos and the 5.1% difference between whites and blacks are Freddie’s defined appraisal gaps and contribute to the racial bias allegations. The study also found that as the concentration of black or latino persons in a tract increased, so did the appraisal gap.

Freddie used a number of “robustness checks” to test the outcomes in the study. These checks included whether an interior inspection was performed versus an exterior inspection or desktop appraisal, the occupancy type, property condition, value trends and level of urbanization. It concluded that the research was solid and that “the appraisal gaps seem pervasive.”

Freddie then took the 12 million appraisals and merged them with loan application data, wherein the borrower provided information on race and ethnicity. Freddie looked at data from January 1, 2016 to December 31, 2020 and found that 6.5% of appraised values came in lower than the contract price for white applicants as compared to 9.5% for latinos and 8.6% for blacks. This suggests that minority applicants experienced an appraisal gap of 2% to 3%. It begs the question of how an appraiser would know the race or ethnicity of the loan applicant.

Freddie went on to examine whether these gaps were caused by a small number of appraisers or by a larger group. It concluded that the majority of appraisers exhibited appraisal gaps in their reports when comparing black and white tracts. One takeaway was that 6% of its appraiser sample was responsible for a 10% appraisal gap. The percentage of appraisers responsible for the latino gap was similar.

In looking for explanations, the GSE examined four factors independently: the geographic distance of comparable sales, the reconciliation of comparable sales, variance in the sales price of comparable sales and the possibility of purchaser overpayment. It found that the average distance between the subject property and the comparable sales included in the appraisal reports was considerably less for black (0.82 miles) and latino (0.73 miles) tracts as compared with white (1.4 miles) tracts. Freddie found these differences to be consistent in both urban and suburban areas. When examining how the adjusted sales prices were reconciled, and specifically how far away the appraised value was from the lowest adjusted sale price (expressed as a percentage of the range), it found similar results: Overall (37.6%), White (37.7%), Latino (36.6%) and Black (37.5%).

Freddie then examined the highest and lowest sales price of the comparable sales as a percentage of the contract price. The higher the percentage, the larger the variance. Starting with the assumption that minority neighborhoods are heterogeneous, Freddie sought to find out whether there were greater variances in minority tracts versus white tracts. It concluded that black tracts (24.1%) have a greater variance than white tracts (21.7%), and that latino tracts had a lower variance (20.1%) than white tracts.

Additionally, Freddie explored the possibility that Black and Latino purchasers may pay more for a similar property than White purchasers. To do this, it used mortgages that it funded, with the final data set including nearly 3 million observations. While controlling for all variables is challenging, Freddie found statistically insignificant results. It estimates that between 2008 and 2021, Black buyers paid 0.05% more than White buyers and Latino buyers paid 0.39% more than White buyers.

In other words, it ruled out these four factors as explanations for the appraisal gaps.

Finally, using additional modeling, Freddie found that a property was more likely to receive an appraised value lower than the contract price in minority neighborhoods, when taking into account structural and neighborhood characteristics.

Freddie admits that its analysis “has not yet determined the full root cause of the gap”. The implications for appraisers, however, could be significant. Freddie is looking to create new valuation approaches and is “testing whether alternatives to traditional appraisals offer a more objective analysis of property value”. I am certain that we will be hearing more on this topic going forward.

Albert Franke III, MAI, SRA is president of Advisra Consulting, LLC, New Haven, CT.