Comment on “Returns to Education and Skills in the Labor Market: Evidence from Japan and Korea”

• Date: 01 January 2017
• Published date: 01 January 2017
• Author: Daiji Kawaguchi
• DOI: http://doi.org/10.1111/aepr.12170

Comment on “Returns to Education and Skills

in the Labor Market: Evidence from Japan and

Korea”

Daiji KAWAGUCHI†

The Universityof Tokyo

JEL codes: I26, J24, J31

Lee and Wie (2017) investigate how adult skill proficiency is determined and how it is

rewarded in the labor market. They examine micro data from the Programme for the

International Assessment of Adult Competencies (PIACC), compiled by the Organisation

for Economic Co-operation and Development (OECD) toaddress their research questions.

The PIACC targets adults and contains three measures of skill proficiency, literacy,

numeracy, and problem solving ability using a computer, along with demographic and

educational background variables and information on family background in childhood. The

skill proficiencyscores are constructed basedon Item Response Theory, using the responses

to a common test battery; thus the proficiency scores are internationally comparable.

While Koreans generally outperform Japanese on the Programme for International

Student Assessment (PISA) which targets 15year olds, Koreans do significantly worse than

Japanese on the PIACC, on average. The first half of Lee and Wie’s paper analyzes why

Koreans perform worse than Japanese by regressing the proficiency scores on a variety

of demographic, educational, and family environmental explanatory variables. An Oaxaca

decompositionreveals that the less significant impact of high school andtertiary education

on proficiency scores is the main reason for the significant gap in the mean skill

proficiency scores. Lee and Wie do not provide reasons for the apparently better

performance of Japanese high school and tertiary education over its Korean counterpart.

A Juhn-Murphy-Pierce decomposition looking at the percentiles of the whole distribution

confirms that the difference in the returns toexplanatory variables explains themajority of

the distributional difference between the two countries. Lee and Wie alternatively could

have relied on Firpo et al.’s (2009) re-centered influence function method, but I believe

the results would not have been very different.

All threeskill competencies are wellrewarded in the Korean andJapanese labor markets.

In both countries, the estimated coefficients on test scores, which are around 0.002, imply

that a one-standard-deviation increase in the skill competency score (about 0.4) increases

wages by about 0.08 log points, which is larger than the return to an additional year of

education. In both countries, the inclusion of a test score reduces the estimated return to

education by about 20%, implying that the return to education is partially attributable to

†Correspondence: Daiji Kawaguchi, GraduateSchool of Economics, TheUniversity of Tokyo, Hongo

7-3-1, Bunkyo-ku, Tokyo, 113-0033, Japan. Email:kawaguchi@e.u-tokyo.ac.jp

doi: 10.1111/aepr.12170 Asian EconomicPolicy Review (2017) 12, 161–162

©2017 JapanCenter for EconomicResearch 161

bs_bs_banner