Piece-Rates, Principal-Agent Models, and Productivity Profiles: Parametric and Semi-Parametric Evidence from Payroll Records

This paper uses data on the wages received by piece-rate workers to estimate worker productivity profiles. The data were collected from the payroll records of a British Columbia copper mine. The advantage of these data is the close link between observed wages and worker productivity. An explicit model is used to control for worker effort as a function of observable worker characteristics and the parameters of the compensation system. The model implies a censored wage distribution, the parameters of which can be estimated using well-known econometric techniques. Semi-parametric estimation allows for the relaxation of the distributional assumptions of the model. Results suggest that while productivity profiles were increasing concave functions of tenure, they were also very flat. I relate these results to historical arguments on the skill-saving nature of technological change in the mining industry at the end of the nineteenth century.