Date of Award

Winter 1998

Project Type


Program or Major


Degree Name

Doctor of Philosophy

First Advisor

Michael D Goldberg


Representative agent models do not match up well with three stylized facts of the business cycle: a money-output connection, countercyclical markups, and acyclical real wages. This thesis investigates whether a New Keynesian model which departs from the representative agent assumptions and models heterogeneity and imperfect competition in the labor market is more consistent with these stylized facts.

One possible explanation of countercyclical markups and acyclical real wages is that labor markets are monopsonistic and monopsony power is weaker during expansions than in recessions, This would require that the elasticity of labor supply be procyclical. This is not possible if worker preferences are homothetic.

An aggregate labor supply function for heterogeneous labor is constructed. Labor is indivisible, and workers are heterogeneous with respect to their nonlabor income endowments and preferences for risk. Nonlabor income is assumed to be distributed Lognormally. Workers' optimizing choices in a take-it-or-leave it job market are determined by a reservation wage function which relates reservation wages to nonlabor income. Aggregate labor supply is a composite function of the Lognormal distribution of nonlabor income and the reservation wage function.

The parameters of the aggregate labor supply function are affected by changes in the aggregate price level and the interest rate. An increase in the money supply increases aggregate labor supply. If workers have increasing relative risk aversion, an increase in money also increases the elasticity of labor supply. The magnitude of this increase depends upon the magnitudes of the interest- and wealth-elasticities of the aggregate money demand function.

If firms are monopsonistic, the elasticity of the aggregate labor supply function will be procyclical with respect to monetary policy, and markups will be countercyclical. A calibrated version of the model indicates that the real wage would be weakly countercyclical, acyclical, or weakly procyclical, depending on the short-term elasticities of the price level and the interest rate with respect to changes in M2. The model implies a wealth-effects transmission channel from monetary policy to aggregate labor supply, employment and output, restoring a traditional Keynesian theme of a monetary theory of production.