Journal of Real Estate Finance and Economics, 21:1, 23±44, 2000
# 2000 Kluwer Academic Publishers. Manufactured in The Netherlands.
Is Lending Discrimination Always Costly?
MICHAEL F. FERGUSON
Department of Finance, College of Business Administration, University of Cincinnati, Cincinnati,
STEPHEN R. PETERS
Department of Finance, University of Illinois, Champaign, IL 61820
How can economically costly discrimination persist in a competitive market? Previous research into this question
has focused on market imperfections which prevent competitive forces from eliminating the economically costly
behavior. In this paper we show that lending discrimination is not always costly (to the lender). This has two
important implications. First, lending discrimination may persist inde®nitely, even in a competitive market.
Second, tests for lending discrimination based on pro®ts (or default rates) may be unable to detect discrimination
when it exists.
Key Words: credit rationing, discrimination, mortgage
Is there racial discrimination in mortgage lending? This has been intensely debated in a
recent series of studies. The ®rst, and most in¯uential, of these studies was conducted by
the Boston Fed (Munnell, Browne, McEneaney, and Tootel, 1992). In the version
published in The American Economic Review, Munnell, Browne, McEneaney, and Tootell
(1996) report that minorities were 2.8 times (28 versus 10 percent) as likely to be denied
mortgage loans as whites.
Even after controlling for relevant default-risk characteristics,
minority applicants were a statistically signi®cant 1.8 times as likely to be denied credit.
These relative denial rates suggest racial discrimination in lending.
On the other hand,
Berkovec, Canner, Gabriel, and Hannan (1994) report that white borrowers default 2 to 4
percent less frequently than minority borrowers. Using traditional models of
discrimination, such as Becker (1957), some would interpret these relative default rates
as evidence against the existence of lending discrimination.
Whether or not these
apparently con¯icting statistics suggest that lending discrimination exists depends in part
on the underlying model of discrimination.
Therefore, we are primarily interested in a
related issue: is lending discrimination costly? The answer to this question is central to the
interpretation of models of discrimination.
In his seminal work The Economics of Discrimination, Gary Becker (1957)
characterizes racial discrimination as a ``taste.'' In the context of mortgage lending,
discrimination would consist of turning down pro®table minority loan applicants. In his