This paper considers the use of instrumental variables to identify a correlated random coefficients model in which coefficients are correlated with (or stochastically dependent on) the regressors. A correlated random coefficients model is central to the human capital earnings model. Conditions are given under which instrumental variables identify the average rate of return. These conditions are applied to David Card's version of Gary Becker's Woytinsky lecture.

James J. Heckman is a professor of economics at the University of Chicago and a Senior Fellow at the American Bar Foundation. Edward Vytlacil is a Sloan Fellow at the University of Chicago. The essential identification results in this paper are presented in Heckman (1982, 1995), and have been widely circulated. This research is supported by NSF-SBR-97-09-873, NIH:R01-HD32058-01A1, and the American Bar Foundation. The authors thank Richard Blundell and Chris Taber for comments on a 1995 draft of this paper, and Kevin Murphy for insightful comments on the 1982 version. Data used in this article can be obtained from the authors from March 1999 through February 2002.