ABA and several other trade groups shared with the CFPB an independent study that calls into question the bureau’s methodology for finding fair lending violations by indirect auto lenders.
Because car dealers are not required to report demographic data about purchasers, the CFPB uses a proxy methodology to identify racial and ethnic information using geographic locations and surnames. This method “creates significant measurement error,” the groups said — it overestimates minorities in the population by as much as 41%. The CFPB’s own materials say the method overestimates minorities by 21% but do not explain how the bureau corrects for the discrepancy.
The study, conducted by Boston-based Charles River Associates, shows that the auto finance market is vastly complex. Any apparent discrepancies are explained by correcting for bias in the proxy methodology and taking into account the complexity and diversity of car sales, including new versus used, seasonality, trade-ins, add-ons and other factors.
Read the letter.
Read the study.