Absolute Differences - Lending
(Feb. 18, 2013; rev. May 10, 2013)
Prefatory note: The subject of this subpage is addressed more fully in Section A of the Appendix to the March 4, 2013 letter to the Federal Reserve Board. That letter and the response to it are discussed on the Federal Reserve Letter sub-page.
In quite of few articles, most of which are listed on the main Lending Disparities page, I discuss the fact that federal regulators encourage lenders to relax lending criteria or otherwise reduce adverse lending outcomes, but, unaware that reducing adverse lending outcomes tends to increase relative differences in adverse outcome rates, continue to monitor fair lending compliance on the basis of relative differences in adverse outcomes. Thus, lenders most responsive to federal encouragements make themselves the most likely targets for litigation. Three recent articles in some manner discussing this issue are: Misunderstanding of Statistics Leads to Misguided Law Enforcement Policies” (Amstat News, Dec. 2012); “’Disparate Impact’: Regulators Need a Lesson in Statistics” (American Banker, June 5, 2012); “The Lending Industry’s Conundrum,” (National Law Journal, Apr. 2, 2012). The National Law Journal item also discusses the fact that the practices that generally would be perceived as less discriminatory alternatives typically would increase relative differences in adverse outcome rates. See also Less Discriminatory Alternative - Substantive sub-page of the Disparate Impact page.
But some lending disparity studies by the Federal Reserve System measure disparities in terms of absolute differences between rates. See Marvin M. Smith and Christy Chung Hevener, “Subprime Lending Over Time: The Role of Race” (Federal Reserve Bank of Philadelphia, October 2010) and Robert B. Avery, Kenneth P. Brevoort, Glenn B. Canner, “The 2006 HMDA Data” (Federal Reserve Bulletin, December 2007) .
There are a number of implications of measuring lending disparities in terms of absolute differences. First, most adverse lending outcomes are in rate ranges where reducing the frequency of adverse outcomes will tend to reduce absolute differences between rates. See introductory material to the Scanlan’s Rule and pages 18-21 of the Harvard University Measurement Letter. Thus, reliance on absolute differences will not yield the anomaly whereby lenders who reduce their adverse outcome rates will be perceived as the most discriminatory lenders.
Second, the Disparities – High Income of the Lending Disparities page discusses the points observers make on the basis of the fact that relative differences in adverse lending outcomes tend to be larger among higher-income groups than lower-income groups. As shown in the tables on that subpage, measured in terms of absolute differences between rates, disparities will tend to be larger among lower-income groups than higher-income groups.