Effects of Educational Mistmatch on Wages in the Korean Labor Market*

• Author: Hanol Lee,Jong‐Wha Lee,Eunbi Song
• Date: 01 December 2016
• DOI: http://doi.org/10.1111/asej.12105
• Published date: 01 December 2016

Effects of Educational Mistmatch on Wages in the

Korean Labor Market*

Hanol Lee, Jong-Wha Lee and Eunbi Song

Received 24 September 2015; Accepted 23 July 2016

Using data on Korean workers from the Programme for the International Assess-

ment of Adult Competencies, the present study empirically investigates the inci-

dence and wage effects of educational mismatch. Among full-time workers aged

25–54, approximately 27 percent are overeducated and 15 percent are underedu-

cated. Our results reveal that, after controlling for omitted variable bias and mea-

surement errors, return to an additional year of overeducation is significantly less

than that to a year of required education, whereas undereducatedworkers do not ap-

pear to suffer wage penalties associated with their deficit schooling. The findings

also show that returns to a year of overeducation vary across fields of study. The

returns to overeducation for college graduates from health and welfare, engineering

and manufacturing, and social sciences, business and law are relatively high com-

pared with those in agriculture, services, and humanities and arts.

Keywords: cognitive skills, education, job mismatch, wage inequality.

JEL classification codes: I26, J24, J31.

doi: 10.1111/asej.12105

I. Introduction

To what extent does investment in education affect individuals’academic achieve-

ment and labor market outcomes? There is a broad consensus that educational

investment has a positive effect on labor market outcomes such as employment

and wages. Rapid expansion of higher level education in many countries during

the past decades has contributed to improvements in workers’productivity and

income growth. However, at the same time, there has been rising concern about

educational mismatch between the supply and demand for educated workers in

the labor markets. This education mismatch may lead to a misallocation of

human resources in the labor market, causing workers to suffer wage penalties,

discouraging labor market participation and even wasting resources at the national

level.

Education–job mismatch is also a pressing issue in Korea and many other Asian

emerging economies. The abundance of well-educated individuals in the

*Jong-Wha Lee (corresponding author): Department of Economics, Korea University, Sungbuk-Ku,

Anam-dong 5-1, Seoul 02842, Korea. We thank an anonymous referee and participants at the

workshop at Korea University for valuable comments. Email: jongwha@korea.ac.kr.

© 2016 East Asian Economic Association and John Wiley & Sons Australia, Ltd

Asian Economic Journal 2016, Vol.30 No. 4, 375–400 375

bs_bs_banner

workforce as a result of educational expansion may have helped Korea strengthen

its technological capacity and overall competitiveness over the past few decades.

However, at the same time, the issues of the quality of education and education

mismatch have gained attention among academics and policy-makers in Korea.

In recent surveys, employers have often suggested that many graduates with

tertiary-level education lack the technical skills and competence to fulfill job

requirements in industries, while college and university graduates often find that

their education and skills are underutilized in the workplace (Kim et al., 2011).

The purpose of the present paper is to examine the incidence and wage effects

of educational mismatch in the Korean labor market using the data from the

Programme for the International Assessment of Adult Competencies (PIAAC).

Despite growing concerns on education mismatch, there are relatively few studies

on Asian emerging economies, let alone on Korea.

1

Thus, to fill the gap in the

literature, the present study investigates the effects of educational mismatch on

wages in Korea, which could also help policy-makers to form policy on

educational quality and human resource utilization. This paper also adopts an

instrument variable (IV) estimation technique, with the introduction of new IV

for the purpose of explaining educational mismatch in Korean society, in order

to address potential bias for observable educational variables included in the

empirical specification. Furthermore, the present study investigates whether

sample selection introduces significant bias in estimated wage effects. The paper

also contributes to the existing literature by investigating the wage effects of

educational mismatch by field of study in Korea.

A large supply of college graduates in the US labor market in the 1970s

triggered serious concerns regarding overeducation, and motivated a large body

of literature on this subject, starting with Freeman (1976), who points out a sharp

decline in the college premium as a consequence of the oversupply of college

graduates. In an influential paper, Duncan and Hoffman (1981) decompose actual

years of schooling in the Mincerian wage equation into three schooling

components (required, surplus and deficit years of education) to estimate returns

to overeducation using the 1976 wave of the Panel Study of Income Dynamics

(PSID). Their results show that return to an extra year of overeducation is

substantially lower than that to an extra year of required education.

A large body of subsequent studies provides results confirming Duncan and

Hoffman’sfindings. According to Leuven and Oosterbeek (2011), existing studies

for individual countries report that 30 and 26 percent of workers are classified as

overeducated and undereducated, respectively, at the overall mean from the 1970s

1 Previous studies (Kim et al., 2011; Lim et al., 2012; Ryu, 2013) exploring overeducation in Korea

show the existence of education mismatch in Koreaand confirm the findings of the literature that over-

educated workers suffer wage penalties, compared to well-matched workers. However, these studies

use dummy variables to indicate required education, overeducationand undereducation and hardly at-

tempt to address the potential biases arising from endogeneity, measurement errors and sample selec-

tion identified in the literature.

ASIAN ECONOMIC JOURNAL 376

© 2016 East Asian Economic Association and John Wiley & Sons Australia, Ltd