Editor: David Reiss
Brooklyn Law School

August 14, 2013

Regression Aggression: Statistics and Disparate Impact

By David Reiss

The Third Circuit affirmed, in Rodriguez et al. v. National City Bank et al., No. 11-8079 (Aug. 12, 2013), the denial of final approval “of the parties’ proposed settlement and certification of the settlement class” in a mortgage loan discrimination case brought by minority borrowers who claimed a disparate impact resulting from how the defendants charged borrowers. (4)

The part of the opinion that I found most interesting (but not compelling) was where it discussed the statistical work that the plaintiffs had done to support their case.  The Court states that

Even if Plaintiffs had succeeded in controlling for every  objective  credit-related variable – something no court could have reviewed because the analyses are not of record – the regression analyses  do not even purport to control for individual, subjective considerations.  A loan officer may have set an individual borrower’s interest rate and fees based on any number of non-discriminatory reasons, such as whether the mortgage loans were intended to benefit other family members who were not borrowers, whether borrowers misrepresented their income or assets, whether borrowers were seeking or had previously been given  favorable loan-to value terms not warranted by their credit status, whether the loans were part of a beneficial debt consolidation, or even concerns the loan officer  may have  had  at the time  for the financial institution irrespective of the borrower. While those possibilities do not necessarily rebut the argument that the Discretionary Pricing Policy opened the door to biases that individual loan officers could have harbored,  they  do undermine the assertion that there was a common and unlawful mode by which the officers exercised their discretion. (26-27)

I have not read the plaintiffs’ study, but the Court’s logic seems suspect to me. I can’t imagine how a statistician would “control for individual, subjective considerations,” particularly as that appears to be an infinite set of variables. Indeed, the Court gives little meaningful guidance as to what a comprehensive regression analysis would look like. What “individual, subjective considerations” would they include?  Where would that data exist to be studied?

The court does note some serious problems with the plaintiffs’ case, including the fact that they did not introduce their data or regression analyses into the record.  But those failings are not sufficient to explain the Court’s reasoning in the selection quoted above.

This case might be ripe for reconsideration or an en banc review, even if just to clarify what the Court wants from plaintiffs in future disparate impact cases.

| Permalink