Microfoundations and Macro Implications of Indivisible Labor


Labor economists have noted that participation is an important margin of labor supply and have emphasized the micro-econometric implications of discrete choice in the labor market. But what are the implications of indivisible labor for "macroeconomic" data - measures of economic activity that are aggregated over time and across people? I build a micro model of "indivisible labor"- the bunching of work in time - as the consequence of fixed costs of beginning a work session. The model emphasizes the possibilities for substitution over time and implies that "participation" is an important margin of aggregate labor supply, although it also predicts that aggregate employment and hours move together. I show that, in a special case, the micro model has as its reduced form the labor supply side of the indivisible labor models of Diamond and Mirrlees (1978, 1986), Hansen (1985), and many others. I then build a "macro" model by aggregating this special case over time and show that it is equivalent in form to the divisible labor model of Lucas and Rapping (1969). This equivalence means that the indivisibility of labor per se need not have implications for macro data on earnings and hours, although I show that the responses of macro data to nonlinear tax rules do depend on the existence and importance of indivisibilities.

Click here to send email to me requesting an Adobe Acrobat copy of the paper. The request should include your email address and institutional affiliation.

© copyright 1998 by Casey B. Mulligan.