Economics 286W
Honors Seminar
Spring
2003
Tuesdays 3:30-5
HRM 303
R. N. Langlois
322
Monteith X63472
Office
hours MT 1:30-3 or by appointment
Assignment 6
Our next speaker will be Steve Ross, who will talk about racial discrimination in mortgage lending — and what economists have been able to say about it. There are three background readings.
· Stephen L. Ross and John Yinger. In Press. “Looking the Other Way: A Critique of the Fair Lending Enforcement System and a Plan to Fix It.” Center for Policy Analysis Policy Brief. Syracuse University.
· Margery Austin Turner, et al., “Other Things Being Equal: A Paired Testing Study of Discrimination in Mortgage Lending,” Working Paper 2003-09, Department of Economics, University of Connecticut, March 2003.
· Stephen L Ross, “Paired Testing and the 2000 Housing Discrimination Study,” Working Paper 2002-15, Department of Economics, University of Connecticut, 2002.
The Ross and Yinger policy
brief is essentially an advocacy piece that takes a particular position
concerning the appropriate enforcement of fair-lending laws. The major emphasis behind the brief is the
failure of current enforcement approaches to address so-called disparate-impact
discrimination. In more recent work, Steve
Ross and his coauthors have revisited the issue using a “paired testing”
approach, and have found (in Working Paper 2003-09)
that there is substantial discrimination against minority buyers who make preliminary
inquiries about mortgages (the pre-application stage). The third paper by Ross (Working Paper 2002-15)
discusses the conceptual issues surrounding the use of paired testing to study
discrimination.
Your assignment is this. Evaluate the arguments of Ross and Yinger in light of the new evidence of discrimination at the pre-application stage. Do their recommendations still make sense? What role do you see for testing as an enforcement mechanism? A major criticism of paired testing is that behavioral differences between the white and minority testers may lead to the observed differences in treatment. If tester attributes actually influence treatment, how does that influence your previous conclusions?
Due: April 29.
