Reappraising the effects of job training on the labor market outcomes of women in Korea: Essays on theoretical and empirical analysis

This dissertation examines the effect of job training on the labor market outcomes of women in Korea. Theoretical and politic considerations and of job training have been suggesting the sufficient reasons why it is important in the labor market. Through job training, women might improve their economic well-being and empowerment in a market that might favor men. However, cultural and social aspects might impact the extent to which women benefit, if at all, from job training. Access to job training might be restricted due to social norms. Traditionally, in Korea, the child care and household work are emphasized as a woman's duty. By contrast, the role women play in economic activities---including job training---is less developed or even under appreciated. Through the three essays I evaluate the effect of job training for women and reappraise its implication considering the structure of the labor market and society in Korea.;In the first essay, I investigate the effect of job training on the wages and employment propensities of female workers in Korea. Using a panel data set from 1998 to 2005, I analyze the impact of job training using both a Random Effects (RE) model and a Fixed Effects (FE) model and compare the results. These results provide an upper and lower bound estimate of the job training effect. I find that job training has a statistically significant and positive effect on wages (5.8% per year in the RE model and 4.9% in the FE model). I also find that while on-the-job training increases wages, training conducted outside the workplace does not affect wages. Despite its insignificant impact on wages, training outside the workplace does appear to improve the employment propensity of women.;The second essay further focuses on the selection issues that arise in evaluating job training. Selection biases might take place, because individuals choose whether or not to participate in either the labor force or job training. Often we assume that unobserved factors in the selection process are correlated with individual characteristics. In that case, using the fixed effect model eliminates the unobserved individual characteristics, so that it can possibly solve selection problem. However, if selection process is correlated with time-variant variable, FE cannot correct the selection problem and is inconsistent. To solve this problem, I apply two econometric methods for correcting selection problems and endogeneity in a wage equation: Two-Stage Least Estimation (2SLS) using the inverse mill's ratio (IMR) and 2SLS using a generated instrument variable (IV) in the wage equation. The results from empirical analysis using these methods detect the presence of the selection bias in the wage equation. Using IMR or/and IV improves selection bias slightly: job training increases the wage rate of women in Korea by 6.1%, which is 1% lower than the estimates of job training without IV and without IMR.;In the third essay, I conduct a comparative study of effectiveness of job training between men and women in Korea. In fact, job training of Korean women has a reasonable effect on the labor market outcomes in Korea, However, it is disputable whether this effect alleviates the gender wage gap in Korea. The empirical analysis indicates that on-the-job training increases the wage rate of women by 8.1% in the random effect model (5.9% in fixed effect model). Moreover this effect is higher for female than males; for instance, it is, in the order of 4.6% in a random effect model (2.7% in a fixed effect model). Although there is no clear effect of off-the job training on wage rate for women, off-the job training increases slightly the wage rate for men. However, off-the job training improves to increase the probability of being employed for both women and men. In spite of a positive effect of job training for women, it seems that job training does not reduce the gender wage gap in Korea. Through the method of decomposition on the wage gap between men and women, I found that the gap of job training taken by men and women does explain the gender wage gap, but it does explain the gap of returns of job training between them. Restricted types of job training available for women and less amount time for job training devoted by women might be contribute to this result.