Skip to content

DOJ's Latest Fair Lending Enforcement Action

Just when you thought it was safe to go in the water, the Department of Justice ("DOJ") brings another fair lending action. This latest case was resolved with a consent order filed on September 5, 1996. The case is brought against Long Beach Mortgage Company for loans made between 1991 and 1994. Long Beach is a former savings and loan that has become a mortgage bank instead of a thrift in order to specialize exclusively in mortgage lending. The investigation by DOJ was begun based on a referral from OTS before the institution dropped its thrift charter.

OTS originally referred the case based on an early fair lending examination in which questions arose concerning loans purchased from a variety of loan originators. Like many purchasers of loans, Long Beach purchases from a large number of originators that may have differing underwriting standards.

The Department of Justice allegations bring several new aspects to the existing fair lending enforcement cases. First, this case involves the lending criteria used by the defendant, Long Beach Mortgage Company in originating loans and in purchasing loans brought to the company by loan brokers. The case thus touches on indirect lending practices, an area that DOJ has been watching closely. DOJ applied a statistical model to the loans and based its allegations on the pricing discrepancies identified by that model. For example, the model showed that the price paid by elderly African-American women as a group was greater than the price paid by white men.

Second, this case is the first to involve a lender that deals in "B" and "C" credits. It thus involves situations that are less than clear-cut lending decisions and that often involve adjustments and compromises in underwriting. Making loans to less than perfect applicants necessarily involves work and flexibility. Use of hindsight to analyze this type of decision, particularly the "C" credits, is troublesome because the factors taken into consideration may differ from the thoughts and focus of the lender while processing the application.

Third, it deals directly with the pricing issues of setting rates and fees for "B" and "C" credits. Consumer advocates have expressed concerns that setting higher rates for customers that the lender perceives as riskier may be discriminatory. The industry's defense of the practice is based on the fact that higher risk customers cost more. The theory is that among "C" borrowers, some will pay satisfactorily but there will be a higher slow pay and default rate than in the "A" portfolio. Setting higher rates for the "C" portfolio is designed to cover this risk. However, when analyzed by demographics, the "C" portfolio may contain a larger number of minorities, providing a basis to the charge of discrimination. This case may direct attention to this concern.

Finally, the case involves more than only the race of the applicants. The case is based on the combination of the applicant's race, age, and gender. Generally, the government has investigated discrimination by controlling for all but one prohibited factor. Cases have compared treatment of Blacks to Whites, or Hispanics to Whites. Alternatively, a case may be based on treatment of elderly applicants as compared to youthful applicants. The reason for this focus on a single prohibited basis is that when several factors are used in conjunction, the result may be misleading.

For example, a comparison of elderly women to youthful men may show differences that appear to be gender discrimination. However, a comparison of elderly men to youthful women could also show the same difference. Then, the investigator must still determine what sort of discrimination, if any, is involved. What may in fact be happening in this hypothetical case is that elderly borrowers are treated differently than youthful borrowers. In that case, the discrimination issue would be one of age discrimination, but not gender discrimination. Alternatively, there could be differences in the qualifications of the elderly borrowers, differences in the products they sought, or differences in the market that caused the differences.

In your bank's fair lending program, it is critical to understand precisely what is behind any differences in borrower groups. This case makes clear that a fair lending self assessment must consider and compare treatment of customers on a variety of prohibited bases. It must also compare lending decisions by product and by source, such as loan broker or loan originator.


  • Whether your bank originates, sells, or purchases, review the underwriting criteria for each type of loan the bank makes or purchases.
  • Compare loans purchased by or sold to different entities and look for consistencies or differences in underwriting standards. If you find differences, analyze the loans by different categories of race, gender, age, and marital status.
  • Review your fair lending audit findings. If you found any inconsistencies in how the bank's underwriting criteria were applied, take steps now to implement uniform standards. This may mean revising policies and procedures and careful training.
  • If your bank is doing business with a loan purchaser whose standards may be discriminatory, discuss this concern and the Long Beach case with your management. Your bank may want to bring any such concerns to the attention of your regulator for guidance.
  • If you make loans to "B" and "C" customers and set rates based on their credit risk, review your risk rating method carefully to be sure that it is consistently applied and that each factor can be defended.

Copyright © 1996 Compliance Action. Originally appeared in Compliance Action, Vol. 1, No. 14, 9/96

First published on 09/01/1996

Banker Store View All

From training, policies, forms, and publications, to office products and occasional gifts, it’s available here:

Banker Store

hot right now

image description

Looking for effective, convenient training on a particular subject?

BOL Learning Connect offers more than 200 courses ON-DEMAND or on CD ROM from AML to Reg Z and every topic in between.

Search Topics