UC Riverside

This dissertation is composed of three essays that study gender inequality in Thailand and the effect of job training on immigrant workers in the United States. Essay 1 identifies the factors that account for mean earning differences in men and women, such as occupational sorting, demographic differences, human capital differences and the unexplained.

The outcome from OB decomposition indicates that the reduction over time in the mean wage gap is mostly due to an increase in female human capital accumulation and the improvement in female occupation outlook relative to men. One of the reasons that a sizable mean wage gap still exists in Thailand, despite the increase in education level of women, is because the increase in female human capital accumulation over the past decade is overshadowed by an increase in the return to observables characteristics of men

Essay 2 finds that the result from DFL decomposition is consistent with the OB decomposition. We find that if Thai women possess similar observable characteristics as men, the gender wage inequality will be greater for the majority of the wage distribution, particularly, for middle to high income workers.

Chapter 3 studies the effects of job training on immigrant workers found in the U.S., using the data from the Survey of Income and Program Participation (SIPP). This essay is one of the first empirical papers that look into the effects of job training on immigrants in the United States, using Random Effect (RE) model, a Propensity Score Matching, a Quantile regression (QREQ) model and a semi-parametric reweighting method. The Random Effects model indicates that the conditional effect of job training on the average earnings of immigrants (at 3.9 percent) is less than that of natives (at 7.6 percent).

From our distribution study, we found that job training had a positive effect on the wages of immigrant workers over most of the wage distribution. The results from QREQ also show that immigrants enjoy the largest conditional job training premium at the lower and middle part of worker earning quantiles. Examining counterfactual study, DFL reweighting technique shows that similar to Abadie, Angrist and Imbens (AAI) (2002) we found that the largest proportional impact of job training is at the upper part of the wage distribution for both natives and immigrants. Nevertheless, we still found that job training increases the wage premium of lower and middle income workers.