Abstract
The objectives of this dissertation are to examine (1) static and dynamic labor supply behavior of individuals and (2) the separability of commodities from leisure. While these topics are concerned with labor supply behavior they are even more closely related here since they share (1) a theoretical approach which treats labor supply behavior as an aspect of the theory of demand, (2) the use of individual rather than aggregate data, and (3) the application of experimental techniques to animal subjects to generate the data. Many expositions of labor supply theory point out that the labor supply curve may be backward bending. The question of which utility functions allows the labor supply curve to bend back is of interest because of the quantity of empirical evidence consistent with this shape of the labor supply curve, yet this question has received only limited attention. In this dissertation we examine some of the utility functions which appear in the labor supply literature to determine if they allow the labor supply function be bend back. We test between several of these models using data from an experiment in which pigeons are used as subjects. In addition, we analyze these subjects adjustment to changes in wage rates and non-wage income. The primary purpose of this analysis is to determine if any common pattern of adjustment exists, and to characterize the adjustment process. Some problems with existing models are pointed out. An obvious generalization of the two-good (consumption - leisure) model is to the case where than one consumption good is available. This leads to the question of how changes in labor supply affect the composition of consumption. One effect is through changes in the level of income. In general, changes in the quantity of labor supplied will affect the composition of consumption even if income is constant. This latter effect is absent if commodities are separable from leisure. In the dissertation, we derive three new tests for separability. These tests are independent of the specification of the utility function, that is, the tests do not require the algebraic specification of the utility function. These tests are then implemented using data from an experiment in which rats are used as subjects.
Thistle, Paul David (1983). Essays in labor supply. Texas A&M University. Texas A&M University. Libraries. Available electronically from
https : / /hdl .handle .net /1969 .1 /DISSERTATIONS -563671.