WASHINGTON – A new comprehensive study of more than 8.2 million loan records by Charles River Associates concludes that the method used by the Consumer Financial Protection Bureau (CFPB) to measure for discrimination in an auto lender’s portfolio is “conceptually flawed in its application and subject to significant bias and estimation error,” reported the NADA.

The peer-review study calls into question the reliability of a testing methodology that the CFPB has used to level allegations of unintended discrimination against—and extract settlements from—auto lenders and to pressure auto lenders to change the way they compensate dealers for originating finance contracts.

The study, released today, reviewed more than 8.2 million new- and used-vehicle finance contracts issued during 2012 and 2013, and measured differences in dealer reserve paid by minorities and non-minorities using the CFPB methodology. Dealer reserve represents the compensation that dealers receive from lenders for their role in the auto-financing process.

The study concluded that the CFPB’s methodology frequently misidentifies the background of consumers and dramatically overestimates differences in dealer reserve paid by different groups of consumers. The methodology also fails to account for numerous factors unrelated to the consumer’s background that affect the amount consumers paid for dealer reserve. The study further explains that the CFPB’s examination of differences in dealer reserve at the portfolio level is meaningless because it completely fails to account for legitimate reasons for pricing differences at the retail level. These collective flaws result in a testing methodology that is inherently unreliable.

“This study shows that the CFPB’s attempt to upend the auto lending process is insufficiently informed and the victim of flawed assumptions and inadequate peer review,” said Peter Welch, president of the National Automobile Dealers Association (NADA). “Allegations of potential discrimination are explosive and certainly should not be made without a reliable foundation in data.”

Currently, 136 members of Congress from both parties – 86 Republicans and 50 Democrats – have cosponsored legislation in the U.S. House of Representatives to rescind the CFPB’s 2013 guidance that serves as the centerpiece of the bureau’s attempt to change the highly efficient and pro-competitive dealer-assisted financing model. The bill, H.R. 5403, co-sponsored by Reps. Marlin Stutzman (R-Ind.) and Ed Perlmutter (D-Colo.), would also require transparency and public input prior to the issuance of future CFPB guidance in auto lending.

The legislation came after dozens of letters—from Congressional Democrats and Republicans to the CFPB—urged the disclosure of the CFPB’s testing methodology, which is lacking in the bureau’s guidance. The CFPB repeatedly failed to fully respond to the questions it was asked, leading to the Stutzman-Perlmutter legislation.

Dealers have also offered up an optional program that addresses fair credit risks. Based on a fair credit risk mitigation model developed by the U.S. Department of Justice in 2007 to resolve fair credit investigations of two dealers, NADA in January 2014 released its comprehensive Fair Credit Compliance Policy & Program. When implemented, the NADA program documents those instances when dealers discount interest rates and ensures the discounts are for legitimate business reasons, like meeting a competitive finance offer. Rather than require costly and inaccurate statistical testing, the program controls for risk on the front end of the transaction. Many dealers, including several large dealer groups, have implemented the program. NADA has called on the CFPB to urge finance companies to incorporate the program into their compliance management system.

“Had the CFPB followed the process set forth in the legislation before it issued its guidance to indirect auto lenders, it could have avoided the flawed assumptions and lack of clarity that have come to characterize this guidance,” said NADA’s Welch. “The way forward is for the government to promote broad industry adoption of NADA’s fair credit program, which would address fair credit risks where they matter—at the retail level.”