Labor supply, demand for children and wage rates of paid employees in Thailand

Photocopy of typescript.Thesis (Ph. D.)--University of Hawaii at Manoa, 1979.Bibliography: leaves 145-149.Microfiche.ix, 149 leaves 29 cmThe purpose of this study is to analyze the labor supply behavior of paid employees in Thailand. The family model of labor supply suggests that family size, wage rates and the labor supply behavior of each family member are interrelated. Therefore, a simultaneous equations model is developed, consisting of the husband's and the wife's labor supply functions, their wage functions, their demand for children and an income identity. The model is estimated by a two-stage least squares method. The data utilized cover 1631 working couples selected from the total number of households in the Survey of Labor Force, Round 1 (January-March), 1977, the National Statistical Office, Thailand. This study has demonstrated that R2 can be used to compare the predictive power of two estimation procedures, namely Heckman's and Kosters' models. Although such comparison proves inconclusive, the former is chosen as our estimation method. The labor supply models of married male and married female are separately estimated because (a) the aggregation test shows that the labor supply behavior of married women in Bangkok is significantly different from that in non-Bangkok areas, while the behavior of married men in these two areas is not significantly different, (b) an analysis of variance shows no significant differences in the average weekly hours of men between these two areas, while the mean hours of work of women in Bangkok are shorter than those in non-Bangkok, and (c) the F-tests confirm that most of the total variation in the working hours of women is explained by variation within (an area), while variation between plays a larger role in explaining the total variation of men's working hours. Results of the estimated labor supply of husbands (for the whole nation) show that the substitution and income elasticities are very low. Moreover, the wife's wage rate, household income, the number of children, the family size, and occupational dummy variables are statistically significant. The Neoclassical implication of symmetrical cross-substitution effects and of the implicit assumption that the husband's and the wife's incomes have identical effects on their labor supply cannot be rejected. In the estimated labor supply of married women in both areas, the husband's wage rate, own wage rate, household income, family size, and occupational dummy variables are statistically significant. Moreover, the number of children is also significant in the Bangkok model and so are age and age squared in the non-Bangkok model. When the assumption of equal income effect is maintained, the hypothesis of identical cross-substitution effects is accepted only in the non-Bangkok model of married women's labor supply. However, when this assumption is relaxed, all of the hypotheses concerning equal income effects and the simultaneous test of equal income and symmetrical cross-substitution effects must be rejected in both areas. This implies that the bargaining theory is a more appropriate model to estimate the labor supply of married women in our data set than the Neoclassical model. When the model is specified by bargaining theory, the estimated substitution and (wife's) income elasticities of the wives' labor supply are greater than one (which is consistent with the secondary workers hypothesis), while the husband's income elasticity of wives' labor supply is much smaller than one. Except for the husband's wage rate in the Bangkok model, all of the coefficients in the demand for children equations for the whole country and for Bangkok are statistically significant. These include the couple's wage rate, household income, the couple's education, the wife's age and an age square, the dummy variable of government-private sector employee and the maid dummy variable. However, in the non-Bangkok model, the couple's education and the maid variable are not significant. Therefore, children are normal goods and the opportunity cost of the couple's time has a negative effect on their demand for children. The results of the estimated wage functions of husbands and wives show that the human capital variables, i.e. education and experience, are always significant in every model. College graduates are found to have higher wage rates than non-college graduates. While the wage rates of married women in the government sector in Bangkok are lower than their counterparts in the private sector, the reverse is true in non-Bangkok areas. Finally, the number of children and the hours of work are not found to have any significant effect on wage rates. In conclusion, since paid employees are the only group of workers who fully pay the income tax, knowledge of the income and substitution elasticities of labor supply can be used to determine the effects of a change in the income tax structure or the effects of an introduction of a negative income tax program on the labor supply effort of paid employees