Inequality of Opportunity Analysis in Korea using Machine Learning
조규방, 박예지, 전훈철
재정학연구 · 2024
Inequality of opportunity arises when individual outcomes are determined by circumstances beyond their control. In previous research on this topic, a common issue was the arbitrary selection of variables representing individual circumstances, which leads to subjective classification of individual types based on these variables. To address this problem, our study uses machine learning techniques to identify significant circumstances, categorize individual types, and analyze income inequality of opportunity in Korea.
Specifically, we employ conditional inference trees and conditional random forest models. The key findings of our study are as follows. First, when comprehensively considering various circumstances, income inequality of opportunity in Korea shows a decreasing trend.
Second, parental circumstances, related to parental education statuses, industries in which parents were engaged, and occupational statuses of parents, have substantial impacts on income inequality of opportunity. This supports the validity of existing studies that used parental education as a proxy for circumstances that individuals may experience. Third, the impact of regional factors such as birthplace and the growing place on income inequality of opportunity has increased over time.
Last, the structure of income inequality of opportunity in Korea has been more complex over time. Compared to the previous studies, this study has the following advantages. We use machine learning techniques to classify individual types based on data without subjective variable selection.
Also, our findings have significant implications for future research and policy development related to inequality of opportunity in Korea, particularly in that it promotes our understanding of the relative importance of variables that influence income inequality of opportunity in Korea.