The political desire for further integration within the European Union will have an increasing effect on the financial services industry, including banking and credit lending. Harmonisation can potentially have great benefit for consumers as the European banks compete for their business across the Union. Harmonisation, however, brings with it a set of decisions for both credit grantors and regulators. There are issues associated with information that can be used in risk assessment to ensure fair decisions in granting credit. In the past the rules have been developed under national legislation. EU directives seek to harmonise the national rules and provide a new guidance on variables that may be held. The major concern relates to the issue of discrimination. The aim of the law is to promote the principle of equal treatment, which can be interpreted in a number of ways. The legal interpretation of discrimination does not necessarily coincide with the economic standpoint. Furthermore, previous empirical research suggests that prohibition of variables may not only affect the ability to distinguish between good and bad risk, but may also be disadvantageous to the groups the legislation is supposed to protect. This paper explores these issues from both theoretical and practical points of view.
[1]
G. Yule.
NOTES ON THE THEORY OF ASSOCIATION OF ATTRIBUTES IN STATISTICS
,
1903
.
[2]
L. Browne,et al.
Mortgage Lending in Boston: Interpreting HMDA Data
,
1992
.
[3]
Dennis L. Hoffman,et al.
Lender Reactions to Information Restrictions: The Case of Banks and ECOA
,
1986
.
[4]
E. Phelps.
The Statistical Theory of Racism and Sexism
,
1972
.
[5]
S. Julious,et al.
Confounding and Simpson's paradox
,
1994,
BMJ.
[6]
D. Hand.
Psychiatric Examples of Simpson's Paradox
,
1979,
British Journal of Psychiatry.
[7]
D. Durand.
Risk elements in consumer instalment financing
,
1942
.
[8]
E. H. Simpson,et al.
The Interpretation of Interaction in Contingency Tables
,
1951
.
[9]
The Equal Credit Opportunity Act: An Evaluation
,
1979
.
[10]
Robert E. Martin,et al.
Loan Performance and Race
,
2000
.
[11]
J. Hanley,et al.
Simpson's paradox in meta-analysis.
,
2000,
Epidemiology.