Panel Paper:
The Effect of Human Capital on the Economic Wellbeing of North Korean Defectors
*Names in bold indicate Presenter
A sample of this study is 152 individual NK defectors from the 2017 National Survey on NK defectors, which adopts the self-administered questionnaire and snowball sampling method. Economic wellbeing is captured by employment and welfare dependency levels*. Human capital is measured by 1) education levels and 2) work experiences in North and South Korea, 3) receipts of education benefits and 4) employment support in South Korea. Ordinary Least Square (OLS) analysis is carried out.
Preliminary outcomes are as follows: At the first stage with only independent and control variables, it is reported that job experiences in South Korea increase the employment by 67% (p<0.001) and that participating in the job fair and job-related programs decreases employment by 47% (p<0.1). At the second stage with moderating variables, education received in North Korea turns out to decrease the employment by 15% (p<0.1), and job experiences in South Korea increase employment by 100% (p<0.01). Female defectors are employed 26% lower than male defectors (p<0.1). However, there were no gender-differentiated effects in the relationship between human capital and employment.
In sum, among the above factors of human capital, only job experiences in South Korea are shown to promote the employment of NK defectors statistically significantly. This finding suggests that the settlement support program needs to help NK defectors to enter a job market as soon as possible and acquire necessary skills in the field rather than in the job programs of settlement support programs. This study has limitations from using data having a small sample size and collected by non-random sampling method. A gender-differentiated impact could have been detected with bigger sample size if limiting the sample by NK defectors in their prime ages. However, considering that most of the NK defector data is not opened, findings of this study are still valuable.
* I found out that a variable related to welfare dependency levels is omitted in the data. I requested that the data be sent again; I will be able to add findings related to welfare dependency levels later.