Working Paper 24-06: Home Purchase Appraisals in Minority Neighborhoods

[Mejappz]

Of course no mention of quality, qualifications, or experience. Geo competency, AMC fee, and turntimes, etc. If you don't hire qualified professionals you get what you pay for.


We study the impact of neighborhoods’ race composition on appraisers’ valuation decisions in home purchase appraisals. Controlling for many appraisal inputs, including the appraiser themselves, we find that low appraisals (below the contract price) are at least 23 percent more likely in majority African American neighborhoods relative to neighborhoods with no African American residents. Instrumental variable estimates, based on historical race shares, indicate an impact of at least 13 percent. However, this effect dissipates when appraisers work in neighborhoods in which they have appraised before or in which many appraisals were recently completed, facts consistent with information based models of discrimination.

Working Paper 24-06: Home Purchase Appraisals in Minority Neighborhoods | FEDERAL HOUSING FINANCE AGENCY

Author: Daniel Grodzicki, Sean Cannon, Christopher W. Davis, Ken Lam

www.fhfa.gov

 

[George Hatch]

------------------------

 

[George Hatch]

Neighborhoods of more heterogenous property attributes are more difficult for both the appraisers AND the market participants to value. Neighborhoods that are "thinner" in recent sales activity are more difficult for both the appraisers and the market participants to work in. When compared to a 400-unit residential subdivision that was built with 3-4 floorplans with perhaps 25% difference in the sizes; in that neighborhood a 5yr old with a dartboard can't get too far away from the real value. Especially not a professional appraiser but that also includes the market participants.

Whatever else might be going on, it's simply harder for the buyer/seller to get to "well informed or well-advised" in a neighborhood where the market is thinly traded and/or consists of a wider variety of age, size, quality and condition.

One difference between appraisers and market participants is that the appraisers are supposed to be looking for and seeing everything that's relevant, not just a smaller subset of everything a broker would present to a buyer or seller. That becomes more of a problem on our end when some appraisers do skip that intermediary step of looking at everything in the neighborhood prior to looking for just the sales they're going to present as "most similar".

I hate to say it, but MAYBE the GSEs conflation of neighborhood vs market segment on the 1004 is prompting some (most, really) appraisers to skip that intermediary step of looking at the macro before attempting to interpret the fewer data points of the micro in their SC analysis. Because we don't skip that step on the CG side. We work from the macro to the micro, even if the level of detail of that macro analysis is far more limited in some assignments than others. One reason we do that is to develop some perspective on the wider trends in order to provide more context for the much fewer number of datapoints in our rental analysis or sales comparables analysis.

I have thought for many years that neighborhood composition and pricing trends should be analyzed separately from the smaller subset that comprises the subject's market segment. Many times in a conforming subdivision ("thicker" in both quantity and homogeneity of property attributes) the two concepts are exactly the same. Any home in that subdivision can be compared for value with any other home because the range of variables is so much tighter. But that isn't the case in all neighborhoods, particularly the rural areas, custom home neighborhoods or those more heterogenous composition neighborhoods where the individual properties vary so much from each other.

 

[George Hatch]

Edit to add, but it looks to me like the new URAR is skipping the "neighborhood" section altogether. If I was designing that form I definitely wouldn't do that under any circumstances. I mean, how can you even do an HBU analysis or get to an opinion of REL without seeing and considering the composition and prevailing pricing and remodeling/development trends of the immediate neighborhood?

Everyone knows that the REL of a beater in a neighborhood of similar might be indefinite, whereas the REL might already be zeroed out if that same beater is located in a neighborhood with a heavy trend of remodeling or redevelopment.

Appraisers only see what they look for. They CAN'T see what they aren't looking for. All anyone who is looking at the subject in isolation of it's immediate environment is going to see is what the subject itself is like. Even though the buyers and sellers really are taking the wider perspective.

Lastly, if the GSEs aren't even collecting those neighborhood data from the appraisers then they can't analyze what the appraisers THINK they are seeing or how they're using their observations in their analyses. Even if/when the appraisers are wrong about their neighborhoods their observations are still affecting their analyses. Sure, there are other ways for the GSEs to gather that neighborhood info, but what they can't gather is what the appraisers actually think they're seeing in those neighborhoods and how it's affecting their work.

-------------------
You cannot answer the question :

Does the property generally conform to the neighborhood (functional utility, style, condition, use, construction, etc") ​

without knowing what that neighborhood is like. And that question very much speaks to which (substitutes) really are the most similar in the eyes of the market participants.

A 1500sf home in avg condition will almost always have a different value when located in a neighborhood full of the same/similar ages, design and other attributes than when it's an orphan located in a neighborhood comprised of much superior or much inferior physical attributes or some widely varying combination thereof.

 

[djd09]

no waivers in east cleveland...

 

[Mejappz]

no waivers in east cleveland...

The motivations behind hiring decisions made by (AMCs) are just as influential as any other factor. However, apart from Philly's report, this crucial aspect often does not receive the attention it deserves. I've been discussing the issue between market area and neighborhood for years, but it always seems to fall on deaf ears. I guess I was right all along.

 

[George Hatch]

And BTW, the above comments about the neighborhood analysis actually support the advantages and the utility of the usage of the conventional 1004 w/appraiser inspection wherein the appraiser is analyzing more than just the subject's property physical attributes. The PDCs can't do it even if they were tasked to do it, and a remote appraiser cannot see anything outside of whatever their various data sources are reporting.

You cannot see what you don't look for. Not just in the appraisals they're complaining about in this controversy but across the entirety of appraisal practice.

 

[djd09]

I guess I was right all along.

no more guessing...of course you are right...look at all the conspiracies that have come true....there is no party like the 15th st party

 

[Terrel L. Shields]

Wouldn't the exact same thing occur in rural markets with scarce data regardless race? The range of values is higher, the margin of error is higher, and geocompetence is an issue where long-time practitioners of the arts would have been insights upon local nuances.

 

[George Hatch]

Wouldn't the exact same thing occur in rural markets with scarce data regardless race? The range of values is higher, the margin of error is higher, and geocompetence is an issue where long-time practitioners of the arts would have been insights upon local nuances.

Of course. Someone coming in from the outside CAN develop an informed perspective on the values of a property but that takes a lot more time/effort than running a single 3-min MLS query and picking the top 3 sales on the list that prints out. Such as the appraiser might be able to get away with in their own back yard where they've already been watching those trends in real time and going back several years. In that scenario the appraiser has already performed that analysis, even if not all in one sitting.