Date of this Version

September 1992

Document Type

Discussion Paper

Publication Details

A. D. Hall and Francis Vella (1992) Estimating Australian Labour Supply Curves From Micro Data: Is It Really Meaningful?

This paper was partially written while Vella was a visitor in the Department of Statistics and the Research School of Social Sciences at the Australian National University. Hall acknowledges the financial support of the Australia Research Committee.

School of Business Discussion Paper ; No. 37, Sep. 1992, Revised

© Copyright A. D. Hall, Francis Vella and the School of Business, Bond University


This paper argues that due to the payment of fringe benefits and the effect of labour legislation entitling workers to non-wage benefits the observed hourly wage is an underestimate of the actual return to labour. Furthermore, as the proportion of total wage comprising fringe benefits increases with hours worked there is a spurious negative relationship between weekly hours and the average hourly wage rate. This is contaminated by the selection bias from individuals optimizing over hours and wage rates. We derive the conditional empirical wage\hours profile while adjusting for the possible endogeneity of hours worked to the wage rate. Our results indicate that the hourly gross wage rate is inversely related to weekly hours worked making the estimation of an upward sloping labour supply curve improbable.