Academic journal article Seoul Journal of Economics

Wage Structure Determinants and Gender Pay Gap among Wage Earners: From Mean to Overall Log Wage Distributional Decomposition

Academic journal article Seoul Journal of Economics

Wage Structure Determinants and Gender Pay Gap among Wage Earners: From Mean to Overall Log Wage Distributional Decomposition

Article excerpt

(ProQuest: ... denotes formulae omitted.)

I. Introduction

Considerable research on gender pay inequality shows that gender pay inequality is relatively higher in Korea than in other industrialized countries. As citizens of a developed and high-income OECD member country, women in Korea face more severe employment stresses than women in similar countries. The average gender wage gap for OECD countries decreased from 23.1 in 2000 to 17.5 in 2014.1 However, the gender wage differentials in Korea topped this list in 2000 (41.7) and did not decrease to the same extent by 2014 (36.7, remained the first) as in other advanced OECD countries. With respect to other labor market characteristics, the labor force participant rate of 74% and 47% for men and women in 2000 changed to 74% and 49.2% in 2014. This change trend is comparable to the change from 73.7% and 50% for men and women in 2000 to 69.2% and 55.4% in 2014 for the OECD average. For educational attainment, the average educational level for Koreans significantly exceeds the OECD average, with 48% of men and 41% of women attaining tertiary degrees in 2014. Around 73.9% of the total working population in the country was working in 2015.

A large gender pay gap in Korea is well documented using various analytical frames and control variables. The present study investigates the gender pay gap for Korean wage earners. Gender pay gap is attributable mainly to observable productivity differential and unobservable discrimination. According to Becker (1971), discrimination unrelated to an individual's productivity is harmful to workforce optimization and social productive efficiency. Korea's significantly wide gender earnings gap is attributed to the stiff job market for women and poor social services for working mothers.2 Korea women typically have to quit their jobs when they give birth3, and their rate of return to full time work is low.4

Considering the large gender pay gap in Korea, we precisely investigate the gender imparity. The classical human capital theory explores the productivity difference by gender on wage structure. The dominant human capital components in wage structure are education, experience, and job training. Educational attainment significantly influences individuals' incomes in Korea and other countries. However, given the high average educational level, the returns to education may be lower in Korea than in other countries. Meanwhile, the importance of experience may obviously increase in magnitude.

The unobservable ability related to education in the error term may result in incorrect regression returns to education, that is, the endogenous estimation bias. To overcome the endogenous problem, economists proposed three major solutions for eliminating the variable bias. These solutions are using a proxy variable, using a twin sample in terms of fixed effect model, and using an instrumental variable. We adopt the proxy variable method to solve the ability bias problem in this study because of the sample and variable limitation. The common proxy variables are intelligence quotient (IQ), work assessment score, and family background. Following earlier literature, we adopt father's education as the proxy of ability. In general, ability is categorized into cognitive ability and non-cognitive ability. Estimating human capital factors of wage structure without considering the effect of non-cognitive ability may cause serious estimation bias (Heckman, and Rubinstein 2001). In addition, non-cognitive abilities significantly influence pooled sample earnings structure and lead to different estimated results by gender.5 Therefore, we adopt life confidence as a non-cognitive ability proxy that reflects individual's endeavor motivation.

Most earlier studies on gender pay differentials focused only on the analysis at the mean of the distribution. However, failing to consider the pay gap across the overall distribution may hinder the thorough investigation of the comprehensive reality. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.