Content area
Abstract
This study uses an economic model to analyze the relations among child quality, family structure, and the mother's shadow wage. Child quality is defined as a composite commodity, the components of which are characteristics which generate parental utility. It is deduced from the model that, ceteris paribus, child quality is higher in families with fewer children and wider child spacing. Furthermore, discrimination against women theoretically results in lower quality female offspring. The relation between child quality and sibling order remains ambiguous since forces are exerted in opposing directions.
The determinants of child quality are examined empirically using data from the Malaysian Family Life Survey. Two methods of estimation are employed. For a single equation model with child quality as dependent variable, ordinary least squares is used. For a two equation simultaneous system with child quality and number of children as endogenous variables, two stage least squares is performed. In both procedures, child quality is measured by educational attainment of the child relative to the mean education of children the same age. Variables that are considered in the determination of child quality are family size, child spacing, the proportion of children that are male, parents' education, race, and urban residence. The results of the two methods of estimation are similar. Most of the family structure variables show little interaction with child quality; family size, however, is negatively related to educational attainment in the ordinary least squares estimation. Parents' education and urban residence are positive influences on child quality.
The value of the mother's time is estimated using a technique devised by James Heckman. The method, which is a variant of Tobit analysis, produces estimates of the woman's asking wage and market wage, and the effects of factors influencing these wages. This analysis generalizes the method by including another parameter which proves very significantly different from zero. Furthermore, this study is the first application of Heckman's method to a developing country; previous research employed only United States data. The results in this thesis show that increases in education, family income excluding the woman's own earnings, the number of young children in the household, and the number of hours worked per week all induce increases in the woman's asking wage. An increase in the number of household members over the age of six brings about a decrease in the reservation wage. The market wage rises in responses to increases in education and experience. While the coefficients of the variables always have the predicted signs, they are rarely significantly different from zero. The influence of education on the market and reservation wages, however, is significant in all specifications.
