A Review of Statistical Problems in the Measurement of Mortgage Market Discrimination and Credit Risk

Abstract

This paper examines the fundamental assumptions within the statistical analysis of discrimination and credit risk and the impact these flawed models had on the mortgage market ranging from false findings of discrimination to incorrectly detecting the future rise of default rates. Serving as a valuable resource to identify the flaws that continue to be overlooked in today's analyses, the paper promotes the improvement of the measurement of mortgage market discrimination and credit risk for a more accurate assessment.

Keywords: Fair lending, Mortgage credit risk, Simultaneous equations bias, Omitted variables