Estimation of a Labor Supply Curve for Self-Employed Business Proprietors

Abstract

There have been many studies in the literature that have investigated empirically labor supply curves for individuals. The most common method has been to estimate a reduced form regression with hours worked as the dependent variable, and the wage rate, non-labor income and possibly some sociodemographic factors as the explanatory variables [2], [5], [6], [8], [9], [10], [14]. Another method has been to estimate a simple utility function with hours worked (or leisure) and income as the arguments [4], [11]. The latter method has been used primarily in studies of low income individuals in attempts to -determine the effects of income maintenance schemes. Finally some studies have relied on interviews with individuals, particularly in attempts to determine the effects of high taxes or the impact of income maintenance proposals on the work-leisure choice [1], [3], [12], [16]. Generally such studies have faced serious drawbacks-accurate data are not available, and the individuals studied often do not have much freedom in their work-leisure choice, thus leaving the observed variables somewhat spurious. Further since these studies generally rely on aggregated data, they have not treated satisfactorily the problems arising from the fact that the net wage facing an individual differs from the gross wage, and the former cannot be taken to be *exogenous. In this paper we attempt to overcome these difficulties by basing -our analysis on a sample of self-employed business proprietors, for whom the work-leisure choice is clearly not restricted by institutional rules.2 Information is available for these individuals on earnings, non-labor income, wife's earnings, weekly hours worked, weeks worked each year, and various sociodemographic characteristics such as age, marital status, and education. We use these data to estimate utility functions with after-tax income and leisure as arguments, attempting to take into account the endogenous nature of the net wage.3 We obtain the associated demand (supply) curves, and determine whether these individuals are on the upward sloping or backward bending seg-