Another "paired" appraisal allegation, Seattle

[sputnam]

The conclusion isn't supported by the methodology or by the data. At best, if there is a large variance, 'paired appraisals' is evidence that at least one of the appraisers didn't do a good job. There are many possible reasons that one (or both) of the appraisers got it wrong including, but not limited to, incompetence, inexperience, and, yes.... bias. To assume that it is bias without a deeper investigation is politics... not rational thought.

 

[djd09]

A Blueprint for Mass Screening Appraisers for Racial Bias and Inaccuracy Based on an Atlanta, GA Study​

By Edward J. Pinto | Tobias Peter

A comprehensive statistical evaluation can rank appraisers based on their performance in each category. The worst performing appraisers would then be subject to a desk review by a panel of expert appraisers, which would render a final verdict on the appraiser’s performance. Where evidence of bias and/or inaccuracy is found, such appraisers should be subjected to enforcement or disciplinary action.


Using a limited set of individual appraisers in the Atlanta metro, we demonstrate that this approach can be applied nationally to conduct a mass appraiser screening. All the required data already exist today to screen individual appraisers for racial bias and inaccuracy. Due to our lack of access to complete sets of data, such a screening needs to be undertaken by entities with such access, namely by Fannie Mae, Freddie Mac, or the Federal Housing Finance Agency (FHFA).[3] After its completion, the entity should release the methodology and anonymized data to independent researchers to replicate and confirm the findings.



November 18, 2022

A Blueprint for Mass Screening Appraisers for Racial Bias and Inaccuracy Based on an Atlanta, GA Study

The AEI Housing Center finds that contrary to recent analysis by The markup/Associated Press, HMDA data on purchase loans do not show evidence of systemic discrimination by the mortgage lending industry when accounting for risk-adjusted loans.

www.aei.org

 

[djd09]

Edward Pinto, Author at Appraisal Buzz​

https://appraisalbuzz.com › author › edward-pinto

 

[CGinMN]

A Blueprint for Mass Screening Appraisers for Racial Bias and Inaccuracy Based on an Atlanta, GA Study​

By Edward J. Pinto | Tobias Peter

A comprehensive statistical evaluation can rank appraisers based on their performance in each category. The worst performing appraisers would then be subject to a desk review by a panel of expert appraisers, which would render a final verdict on the appraiser’s performance. Where evidence of bias and/or inaccuracy is found, such appraisers should be subjected to enforcement or disciplinary action.


Using a limited set of individual appraisers in the Atlanta metro, we demonstrate that this approach can be applied nationally to conduct a mass appraiser screening. All the required data already exist today to screen individual appraisers for racial bias and inaccuracy. Due to our lack of access to complete sets of data, such a screening needs to be undertaken by entities with such access, namely by Fannie Mae, Freddie Mac, or the Federal Housing Finance Agency (FHFA).[3] After its completion, the entity should release the methodology and anonymized data to independent researchers to replicate and confirm the findings.



November 18, 2022

A Blueprint for Mass Screening Appraisers for Racial Bias and Inaccuracy Based on an Atlanta, GA Study

The AEI Housing Center finds that contrary to recent analysis by The markup/Associated Press, HMDA data on purchase loans do not show evidence of systemic discrimination by the mortgage lending industry when accounting for risk-adjusted loans.

www.aei.org

I know appraisers here might think this is Big Brother, but I think furnishing appraisers with targeted insights based on big data analysis could be one of the ways the GSEs improve appraisal quality and combat whatever implicit bias exists within the industry. I support these type of measures, so long as the desk reviews are performed by local experts to validate the findings, and appraisers are given an opportunity to respond and improve their work before they are subjected to discipline or blacklisting.

 

[Elliott]

A Blueprint for Mass Screening Appraisers for Racial Bias and Inaccuracy Based on an Atlanta, GA Study​

By Edward J. Pinto | Tobias Peter

Peter's methodology is a start, but I don't know what happened to having a good, full review appraiser look at two appraisals and decide what was going on, rather than a sensational 'bias' story done by the media.

I wonder why Freddie doesn't do a study about 'number hitters' which would seem to be more important to the integrity of appraisers and their risk.

 

[George Hatch]

Almost all of these cases involve refi transactions where there has been no recent MLS exposure for us to look at, there have been no offers to buy for a seller and their broker to accept/reject, or for an appraiser to analyze or consider, or for an appraisal to come in "low" in relation to the market activity for that property. If you want to say that "a tie goes to the runner" in an appraisal assignment there was no "runner" in any refi to justify an appraiser sticking their neck out. No benefit of the doubt for the appraiser to consider.

I can only think of one media example (Stockton ex-mayor's home) where there was a contract price that didn't appraise. And in that example the seller ended up entering (and closing) a second contract at a lower price that actually did appraise out in a subsequent appraisal. Which in turn indicates that the first appraisal was within reason. In that case there ended up being additional sales data in that neighborhood afterwards which also demonstrated that the first appraisal was reasonable.

 

[CGinMN]

I also noticed that the borrower in the Seattle case referenced a conversation with their agent who told them that $800k-$900k was low. Agents and LO selling high numbers and misinforming borrowers also seems to be a common thread, as that happened with LO in the Baltimore refi.

 

[djd09]

I know appraisers here might think this is Big Brother, but I think furnishing appraisers with targeted insights based on big data analysis could be one of the ways the GSEs improve appraisal quality and combat whatever implicit bias exists within the industry. I support these type of measures, so long as the desk reviews are performed by local experts to validate the findings, and appraisers are given an opportunity to respond and improve their work before they are subjected to discipline or blacklisting.

could you imagine if the loan officers were complaining that the number was low. somewhere there is a petition about that

 

[George Hatch]

According to the complainants comments this WAS the loan broker making those comments to their borrower. Their problem was they apparently didn't control the appraiser. Leastwise, not on the first assignment.

20 years ago there's a high likelihood their pet appraiser would have given them a comp check and been engaged and paid contingent on returning the predetermined value. That didn't happen in the first appraisal because they apparently didn't have access to the appraiser or the appraiser's fee.

 

[Joe Flacco]

You can't ignore these properties. You can add other comps if you want but can't ignore these.