The aggregate Frisch elasticity of labor supply has played a key role in business cycle analysis. This paper develops a statistical aggregation procedure which allows for worker heterogeneity in observables and unobservables and is applicable to an individual labor supply function with non-employment as a possible outcome. Performing a thought experiment in which all offered or paid wages are subject to an unanticipated temporary change, we can derive an analytical expression for the aggregate Frisch elasticity and illustrate its main components: (i) the intensive and extensive adjustment of hours worked, (ii) the extensive adjustment of wages, and (iii) the aggregate employment rate. We use individual-specific data from the German Socio-Economic Panel (SOEP) for males at working-age in order to quantify each component. This data base provides indirect evidence on non-employed workers' reservation wages. We use this variable in conjunction with a two-step conditional density estimator to retrieve the extensive adjustment of hours worked and wages paid. The intensive hours' adjustment follows from estimating a conventional panel data model of individual hours worked. Our estimated aggregate Frisch elasticity varies between .63 and .70. These results are sensitive to the assumed nature of wage changes.
