Skip to main content

“Disparate” means that you are treating one person differently than another.  Now many of you will jump to the defense and swear that you don’t treat any client differently than the other.  For some of you that may be true, but based on mystery shopping results the examiners don’t see it that way.

You may have heard the terminology “Disparate Treatment” and “Disparate Impact”.  They start with the same word, but have decidedly different meanings.

Disparate treatment occurs when a lender treats applicants differently.  This may occur based on our unconscious biases about people or situations, or done as a pattern of practice.  For instance, you may get two emails from two separate prospects and decide to contact one of them that night versus both of them.  For one person you were quicker to respond and offer assistance.  Was it because they were in an area of higher loan amounts or less difficult loans?  Or was it because one had a last name of Smith and another Rodriquez?  Did one come from an easy real estate agent to deal with and one from a high maintenance real estate agent?  These are just a few examples that illustrate how clients may receive disparate treatment from your staff.  Random treatment like this can only be helped with consistent training.

However, Disparate Treatment that is done on a prohibited basis is another story and may occur in patterns of treatment to certain types of customers.  This is where mystery shoppers determine if this disparate treatment is a pattern of practice at your company, by following scientific methodology.  In one of the recent settlements, BancorpSouth was mystery shopped to determine if a lender violated ECOA, by treating customers differently based on race.  White shoppers and Black/African American shoppers were given similar credentials (the black/African American shopper actually had a better financial profile).  They also were sent to the same originator each time.  If you read the settlement, the white shopper was given more favorable feedback than the black/African American shopper, despite the fact that the black/African American shopper had a stronger financial profile.  When this testing was repeated in multiple branches, a pattern of practice soon developed.  This pattern of practice led to Disparate Treatment on a prohibited basis; which in this example was Race.

On the other hand, Disparate Impact occurs when a policy or practice is applied equally to all applicants but has a disproportionate, adverse impact on applicants from a protected group.  It is a lot subtler, but just as dangerous in an exam. Disparate impact occurs when you apply your policy “neutrally” (everyone gets treated the same); yet by virtue of your policy you cause a Disparate impact on a prohibited basis in your market.

Example:  Your investors allow for a 620 credit score.  Your communication to your origination staff is that your company requires a 680 credit score.  After all, this is what risk management said you should require, so that your company gets loans that perform well.  What risk management forgot to think about are the fair lending implications, and the business justification for this overlay!  You see, risk management is about protecting the company bottom line AND regulatory compliance; not one over the other.  This is a delicate balance.

So, the examiners pay a visit and see that you don’t get much business from certain areas in your market.  Those areas happen to be the higher minority or underserved areas.  However, your peers seem to be doing business in those areas.  Since your evaluation is generally done as a comparison to your peers, they ask “why do they do more business in the higher minority areas than your company”?

Your first instinct is to say “We market everywhere and treat everyone the same.  I should not be held responsible just because someone did not apply here”!

So the examiners begin to look at your policies and practices and notice the credit score overlay.  They ask, “if your investor allows a 620 credit score, then why do you require a 680 credit score”?  Risk management looks at each other and says that they have a business justification for that.  They proceed by explaining that they believe loans with a 680 credit score perform better.  To which the examiner asks “what data do you have to substantiate that theory”?  You see, the company who institutes the overlay is the company responsible for empirical data, which proves out their business justification.  This is your overlay and not your investor overlay, so your company is responsible for the empirical evidence to back up that business justification.  Otherwise, the examiner may see that because of your policy, people from those areas are not applying for financing.  They may also notice that you did not test out your theory first by doing proper due diligence.

Without a business justification backed by empirical data, it is likely that you are causing Disparate Impact on an area of your market, which may be in a high minority market (prohibited basis = race) and this is prohibited under ECOA.  Your policy is facially neutral (we treat everyone the same) yet by virtue of that policy you cause Disparate Impact on a prohibited basis!

What is the Solution?  Well by now, I hope you know that I do not like giving you a problem without a solution.

If you are going to add an overlay to pricing or underwriting perform proper due diligence:

  1. Look at the average buyer per census tract in your assessment area. For mortgage bankers, that means the MSAs you serve.
  2. Will the overlay cause disparate impact? e The average homebuyer in census tract 8500.01 has a credit score of 660.  Our peers require 640 and we require 680.  This means your company requires 20 points higher than the average homebuyer in that area, and 40 points higher than your peers.  You will need a strong justification for your reasoning, because it is likely to cause disparate impact in that area.

Moral to the Story:  If you have any overlay, you must back it up with a business justification and reams of empirical data.  And even this may not help if your peers are able to do business in underserved areas without that overlay!

If you need assistance in determining overlay criteria, or Fair Lending Training to battle the mystery shoppers, I’m just a phone call or email away!

Tammy Butler, Master CMB

Author Tammy Butler, Master CMB

More posts by Tammy Butler, Master CMB

Leave a Reply