Premium
A General Test for Compliance with Equal Credit Opportunity Standards
Author(s) -
INGRAM F. JERRY,
WARNER ARTHUR E.
Publication year - 1981
Publication title -
journal of consumer affairs
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.582
H-Index - 62
eISSN - 1745-6606
pISSN - 0022-0078
DOI - 10.1111/j.1745-6606.1981.tb00711.x
Subject(s) - discriminant , actuarial science , test (biology) , linear discriminant analysis , loan , discriminant function analysis , compliance (psychology) , multivariate statistics , set (abstract data type) , credit risk , econometrics , statistics , psychology , business , economics , mathematics , computer science , finance , social psychology , artificial intelligence , paleontology , biology , programming language
The basic objective of this study was to justify, develop, describe, apply and evaluate a statistical test for assessing the lending patterns of financial institutions for evidence of discrimination. Reduced to its essentials, the proposed test consists of a comparison of the discriminating ability of two alternative multivariate discriminant functions. One of these discriminant functions contains only generally accepted risk and return factors as discriminant variables, while the other is composed of the same risk and return factors plus a set of variables representing protected classes of loan applicants. An empirical application of the methodology in a mortgage lending setting led to the conclusion that protected classes of applicants such as minorities, women, and the elderly were afforded the same access to credit as their unprotected counterparts in the same risk categories.