This paper considers to what extent dynamic incentive models such as Green (1987), Phelan and Townsend (1991), and Atkeson and Lucas (1992) can quantitatively explain why some individuals consume (or work) more than others and why a typical individual's consumption (or leisure) varies over time. A simple repeated agency model is shown to be able to better match several population moments concerning life-cycle labor supply and consumption variation than the same model without incentive constraints. On the other hand, I show for the standard stylized models of repeated agency, incentive-induced variability is an insufficient explanation of the variability in the data.
Bibliographical noteFunding Information:
*This work is drawn from my University of Chicago Ph.D. dissertation of the same title. I would like to thank Andrew Atkeson, William Carrington, John Heaton, Robert Lucas, John Matsusaka, Karl Snow, the members of the Money and Banking Workshop at the University of Chicago, and seminar participants at Cornell University, Princeton University, the University of Pennsylvania, Northwestern University, the University of Wisconsin, the University of Illinois, the University of Rochester, Boston University, the University of Virginia, the 1990 meetings of the Society for Economic Dynamics and Control, and the associate editor for helpful comments. My committee members John Cochrane and Lars Hansen were especially helpful. Finally I want to thank my committee chairman Robert Townsend for his comments and constant support throughout my graduate education. Computer support from the National Center for Supercomputing Applications, Champaign, Illinois, and financial support from the Earhart Foundation, Bradley Foundation, National Science Foundation grant SES-911 l-926, and my wife, Elizabeth M. Bloomer, is appreciated.
- Consumption variability