Gender, Sibling Order, and Differences in the Quantity and Quality of Education: Evidence from Japanese Twins*

• Date: 01 June 2016
• Author: Hisakazu Matsushige,Tien Manh Vu
• Published date: 01 June 2016
• DOI: http://doi.org/10.1111/asej.12088

Gender, Sibling Order, and Differences in the

Quantity and Quality of Education: Evidence from

Japanese Twins*

Tien Manh Vu and Hisakazu Matsushige

Received 17 February 2015; Accepted 16 February 2016

Using information on 1045 pairs of Japanese monozygotic twins, we examined

differences in education by considering both the years of schooling (quantity) and

the reputation of the last attended school (quality). We found that a difference in

learning performance at 15 years of age is one of the key factors determining the

differences. We also found that a female eldest child in the family from the 1950s

and 1960s birth cohorts averaged 0.54years less schooling than did her ‘younger’

twin. However, for the same birth cohorts, a male eldest child in the family gener-

ally had access to higher-quality education than his ‘younger’twin. Nonetheless, as

the Japanese economy matured in the 1970s and thereafter, educational differences

between twins disappeared, regardless of gender and sibling order.

Keywords: equality of education, gender, identical twins, sibling order.

JEL classification codes: I24, I25, J13, J16.

doi: 10.1111/asej.12088

I. Introduction

Whether there is a connection among gender, sibling birth order, education and

changes in these factors over time has been subject to some debate. De Haan et al.

(2014) obtained contradictive evidence in both developed and developing coun-

tries that birth order affected sibling differences in human capital development.

However, it is unknown whether any confounding problems with birth timing,

* Vu (corresponding author): International Research Fellowof the Japan Society for thePromotion of

Science, Osaka School of International Public Policy, Osaka University, 1–31 Machikaneyama,

Toyonaka, Osaka 560–0043, Japan. Email: t-vu@osipp.osaka-u.ac.jp. Matsushige: Osaka School of

International Public Policy, Osaka University, 1–31 Machikaneyama, Toyonaka, Osaka 560–0043,

Japan. Email: matushige@osipp.osaka-u.ac.jp. Theauthors gratefully acknowledge the valued support

of members of ‘The Assessment of the Quality and Productivity of Nonmarketable Services’research

project (Grant-in-Aid for Scientific Research No. A3243044, Research Representative: Takeshi

Hiromatsu, and No. A26043110). The authors would also like to thank an anonymous referee and

the editors of this journal, Charles Yuji Horioka, Midori Wakabayashi, Wataru Kureishi, Yoko Niimi,

participants of the Singapore Economic Review Conference held in Singapore on 5–7 August 2015,

and delegates at the joint conference of the International Association for Research in Economic

Psychology and the Society for the Advancement of Behavioral Economics held in Sibiu, Romania

on 3–6 September 2015 for their helpful comments and suggestions.

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

Asian Economic Journal 2016, Vol.30 No. 2, 147–170 147

bs_bs_banner

genetic difference and unobservable characteristics affect the estimation of the

differences in education by sibling order and gender (Rosenzweig and Wolpin,

2000). Kureishi and Wakabayashi (2011) report a son preference in Japan among

the 1920–1939 birth cohorts. In addition, Wakabayashi and Horioka (2009) found

that Japanese parents were more likely to reside with the eldest son in the family.

However, it is unknown whether the eldest son received some advantage in terms

of education.

We argue that monozygotic twins are better for comparing the educational

differences among siblings than singleton births, especially given certain prefer-

ences of parents for children. This is because genetics, birth timing and prenatal

investment are exactly the same for monozygotic twins. These precious and

unique features of monozygotic twins combined with first-differences techniques

help to remove any common unobservable genetic endowments (Rosenzweig and

Wolpin, 2000). Meanwhile, using siblings of singleton birth, all estimations are

always biased because proxies for genetic endowment, family condition (such as

wealth) and prenatal investment are difficult to obtain and measure precisely.

The purpose of this study was to examine educational differences in Japanese

monozygotic twins using a survey detailing many life events. More specifically,

we investigated whether the preferences of parents on sibling order could be

connected to educational differences. Compared with previous studies, our

research used monozygotic twins to completely control for unobservable genetic

endowments, birth timing, prenatal investment and even variation in children’s

abilities/performance during development. In addition, extant studies often

examine only the quantity of education using either the years of schooling or

the highest grade obtained. Our analysis considers not only the difference in

the years of schooling, including additional years for incomplete qualifications,

but also the quality of education using the reputation of the last attended

school/university.

We first investigated whether a difference in the years of schooling existed. We

measured the determinants of the difference using ordinary least squares (OLS)

models including sibling effects to exclude the common time-invariant factors.

Then, using probit models, we examined the probability of the ‘elder’twin,

who was often born just a few minutes earlier, receiving either more years of

schooling or attending a school with a better reputation than the younger twin.

We separated the data into birth cohort decades and gender to attain additional

insight.

Weobtained three main findings. First, the educational difference between twins

was significant and substantially contributed to by the difference in learning

performance when they were 15 years of age. Second, on average, the eldest female

twin in the 1950s and 1960s birth cohorts completed 0.54 years less schooling than

did her twin sibling. We found that the ‘elder’male twin, who was also the eldest

child in the family, had access to higher-quality education than his ‘younger’twin.

Finally, the apparent inequality in education by gender and sibling order in both

quantity and quality disappeared completely in subsequent decades.

ASIAN ECONOMIC JOURNAL 148

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