Studies conducted in the 1990s of racial differences in mortgage rejection rates suffered from a number of problems, including underadjustment for relevant factors (as discussed on the Underadjustment Issues sub-page), the general failure to recognize the pattern whereby the rarer an outcome the greater tends to be the relative difference in experiencing it and the smaller tends to be the relative differences in failing to experience it, and misattribution of significance to the comparative size of relative differences in high-income groups (as discussed on Disparities – High Income sub-page). But, at least such studies met the minimal criterion of a plausible effort to determine whether persons seeking some outcome are treated differently on the basis of their demographic group membership in that they examined the entire universe of applicants seeking some desired outcome and the entire universe of applicants offered that desired outcome.
The same cannot be said of studies of patterns of assignment to subprime loan and differences in costs of loans underlying the complaints underlying the $335 million and $175 million settlements in United States v. Countrywide and United States v. Wells Fargo. The claim of discriminatory assignment to subprime status is simply a distortion of a claim of discriminatory denial of a prime loan. One might reasonably endeavor to determine whether a lender has discriminatorily denied prime loan by examining the universe of applicants seeking such loans and the universe of applicants offered such loans. The analysis underlying the Wells Fargo and Countrywide complaints, however, ignores (a) applicants offered no loan at all; (b) applicants offered prime who loans who refused them (presumably because they did not find the terms favorable enough or as favorable as those they could secure elsewhere); (c) applicants offered subprime loan who refused them (presumably because they did not find the terms favorable enough or as favorable as those they could secure elsewhere, which might involve prime loans). Thus, analyses of assignment to subprime status would be fundamentally unsound even if they were able to fully adjust for group differences in factors related to securing a prime loan.
Similarly, while the most reasonable way to analyze claims of discrimination in loan pricing is not completely clear, it is clear that one cannot analyze such claims while ignoring the applicants who declined a loan because the offered terms were unsatisfactory.