Labor Supply Elasticities: Can Micro Be Misleading for Macro?

In this paper we compare in a consistent way micro and macro labor supply elasticities. The individual elasticity is obtained from the Panel Study of Income Dynamics (PSID). The aggregate, time-series, elasticity is estimated from the exact aggregation of the individual units in the PSID, each year. Our aggregation procedure is legitimate since it relies on exact aggregation of first-order conditions in a simple life-cycle labor supply model with home production. We find that the individual elasticity is about 0.1, a low value that agrees with standard micro estimates, but that the aggregate elasticity is 1.5, a much larger value that incidentally agrees with the pioneering estimate of Lucas and Rapping (1969). This result derives from a pure aggregation effect: not surpisingly, most of the difference is due to the extensive margin, i.e. participation/employment decisions. An implication of our result is that micro evidence is not always a reliable guidance for calibrating aggregate macroeconomic parameters.