Abstract
Although the neoclassical labor economics literature assumes that hours of work are determined solely on the supply side as a result of individual demand for leisure, an abundance of evidence points to the importance of employer demand factors in the market for hours of work. Despite the appeal of models allowing for simultaneity in the market for hours, the scarcity of appropriate data has made their estimation difficult. In this paper I attempt to incorporate labor demand into the problem of hours determination in an empirically tractable manner by exploiting the theoretically distinct roles played by commuting time at the individual and aggregate levels. Applying instrumental variables techniques to data from the 1990 U.S. Census yields larger cross-sectional wage elasticities of labor supply for both men and women than are generally found using conventional estimation methods.