DOES THE PROGRESSIVITY OF INCOME T AXES MATTER FOR HUMAN CAPITAL AND GROWTH? ELIZABETH M. CAUCUTT University of Western Ontario SELAHATTIN ˙ IMROHORO ˘ GLU University of Southern California KRISHNA B. KUMAR University of Southern California Abstract We develop a model with heterogeneity in skills to study the effect of tax progressivity on economic growth. The probability of becom- ing skilled depends positively on expenses on teacher time. We con- sider growth resulting from an externality due to skilled workers and from their employment in research and development. We show changes in the progressivity of taxes can have growth effects even when changes in flat rate taxes have none. The response is stronger with externality-driven growth. Progressive taxation, often suggested to reduce inequality, can increase the long-run skill premium and decrease the upward mobility of the poor. 1. Introduction The growth effects of tax reform have typically been studied in representative agent models by considering changes in flat rate taxes. The initial influence for this literature is the study by Lucas (1990), who uses an endogenous Elizabeth M. Caucutt, Department of Economics, University of Western Ontario, So- cial Science Centre, London, Ontario, Canada N6A 5C2 (ecaucutt@uwo.ca). Selahattin ˙ Imrohoro˘ glu, Department of Finance and Business Economics, Marshall School of Busi- ness, University of Southern California, 701 Exposition Boulevard #701, Los Angeles, CA 90089-1427 (selo@marshall.usc.edu). Krishna B. Kumar, Fuqua School of Business, Duke University & RAND, 1776 Main Street, P.O. Box 2138, Santa Monica, CA 90407-2138 (kkumar@duke.edu, kumar@rand.org). We are grateful to John Conley (the editor) and two anonymous referees for their sug- gestions. We also received helpful comments from V.V. Chari, Harald Uhlig, Steve Yamarik, and participants at several seminars. Received December 1, 2003; Accepted October 27, 2004. © 2006 Blackwell Publishing, Inc. Journal of Public Economic Theory, 8 (1), 2006, pp. 95–118. 95