Human Capital and Worker Productivity: Direct Evidence from Linked Employer-Employee Data

Abstract

The long literatures on the determinants of wage rates at the individual levels and on the empirical relation between productivity and wage rates intersect when attention is focused on longitudinally linked employer-employee data. We estimate separate statistical components of wage rates associated with the observable individual characteristics, unobservable individual heterogeneity and unobservable employer heterogeneity. We define general human capital as the portable components of the full-time, full-year wage rate. Within each employer in the linked sample, we create employer-aggregates of the general human capital. We then estimate the relation between sales per employee, general human capital, and employer wage heterogeneity using micro data for the employing firms. The results reveal direct statistical links between the productivity outcome (sales/worker) and general human capital, controlling for firm-specific wage rate heterogeneity, which can be interpreted as specific human capital or as part of a firm-specific compensation strategy.

Extracted Key Phrases

8 Figures and Tables

Cite this paper

@inproceedings{Abowd2005HumanCA, title={Human Capital and Worker Productivity: Direct Evidence from Linked Employer-Employee Data}, author={John M. Abowd and Francis Kramarz}, year={2005} }