Age of Menarche, Adolescent Sexual Intercourse and Schooling Attainment of Women
| Published date | 01 August 2019 |
| Date | 01 August 2019 |
| DOI | http://doi.org/10.1111/obes.12284 |
717
©2019 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd.
OXFORD BULLETIN OF ECONOMICSAND STATISTICS, 81, 4 (2019) 0305–9049
doi: 10.1111/obes.12284
Age of Menarche, Adolescent Sexual Intercourse and
Schooling Attainment of Women*
Jian Huang,†Wim Groot,‡John G. Sessions§ and Yinyen Tseng†
†School of Public Finance and Taxation, Southwestern University of Economics and
Finance, Chengdu, China, (e-mail: jhuang1@swufe.edu.cn, ytseng@swufe.edu.cns)
‡Department of Health Organization, Policy and Economics, Faculty of Health, Medicine
and Life Sciences, Maastricht University, Maastricht, The Netherlands, (e-mail:
w.groot@maastrichtuniversity.nl)
§Department of Economics and IZA, Newcastle University Business School, Newcastle
Upon Tyne, NE1 4SE, England, (e-mail: j.g.sessions@newcastle.ac.uk)
Abstract
We use data from the 1970 British Cohort Study to measure the effect of adolescent
sexual intercourse on female schooling attainment. We emphasize the appropriate use
of menarcheal age as an instrumental variable (IV) for early intercourse. Our analysis
suggests that developmental trajectories varywith menarcheal age and, therefore, capturing
variations in individual cognitive capacities induced by pubertal timing is crucial for the
validity of the IV identification strategy. Our empirical results indicate that adolescent
sexuality reduces full-time education by approximately one year. Given that 37 percent of
females in our data exited virginity in adolescence, the aggregate loss of human capital as
measured by average years of female schooling could be up to one-third of a year.
I. Introduction
Sexual intercourse is common among adolescents. Almost 50% of US high school stu-
dents report having experienced sexual intercourse while the median age for first sexual
intercourse in the United Kingdom is approximately 16 years old (Tripp andViner, 2005).
Early experience of sexual intercourse has been linked to educational underachievement,
particularly for females (Donovan, 1996; Silverman et al., 2004). Studies in psychology,
sociology and economics have found evidence of a negative correlation between adoles-
cent sexual activity and human capital accumulation (see, for example, Mott and Marsiglio,
1985; Upchurch and McCarthy,1990; Kaltiala-Heino et al., 2003; Sabia, 2007; Blakemore
et al., 2010).
It is well recognized that the statistical association between early intercourse and edu-
cational attainment may be due, at least in part, to unmeasured characteristics. Researchers
JEL Classification numbers: I10, I18, I21.
*We are grateful to three anonymous referees and the Editor of this journal for helpful comments. The normal
disclaimer applies.
718 Bulletin
increasingly resort to the Instrumental Variable (IV) approach to identify the causal effect
of adolescent intercourse (or fertility) on female educational attainment, utilizing the age
of menarche as an instrument (see, for example, Klepinger et al., 1999; Sabia and Rees,
2009). These studies assume that the onset of menarche is a natural process that is re-
lated to educational attainment only through sexual behaviour and thus consider the age
of menarche as an exogenous source of variations in the risk of early intercourse.The core
assumption that characterizes the IV approach is, however, subject to debate. The effect of
pubertal timing on education development may operate through channels other than sexual
behaviour. Particularly, individual variation in the onset of menarche may cause individual
variations in the way that the brain is influenced by pubertal hormones (Sisk and Zehr,
2005; Blakemore et al., 2010). This raises concern as to whether the age of menarche is
capable of generating exogenous variations in virginity status.
Webuild upon this line of literature to investigate the impact of adolescent intercourse on
female educational achievement, using data from a British longitudinal study. Our study has
twomajor contributions. First, we offer a detailed analysis as to whether the age of menarche
can be used as an instrument for the risk of early intercourse. Our study suggests that
developmental trajectories vary with menarcheal age and, therefore, capturing variations
in individual cognitivecapacities induced by pubertal timing is crucial for the validity of the
IV identification strategy.Second, we provide solid evidence for an adverseacademic effect
of losing virginity at an early age. Our empirical analysis identifies adolescent intercourse
as a major hindrance to educational attainment.
The paper proceeds as follows. Section Two describes the background and literature
related to our study. SectionThree provides a description of the data from the 1970 British
Cohort Study and discusses the estimation methods and identification strategies based on
the available data. Section Four presents our estimation results and Section Five presents
a series of robustness tests. The Final remarks are collected in Section Six.
II. Background and related literature
Sexual behaviour in adolescence is widely considered to be a risk factor for educational
under-achievement. Empirical evidence suggests that adolescents with earlyexperience of
sexual intercourse are more likely to drop out of school and are less likely to acquire fur-
ther education (Upchurch and McCarthy, 1990; Dorius et al., 1993; Rector and Johnson,
2005). While the causal relationship is not well understood, a negative association may
arise from health and psychological factors. Adolescents are at a higher risk of acquiring
sexually transmitted diseases (STD) due to their immature reproductive and immune sys-
tems (Morris et al., 1993; Meschke et al., 2002; Bearinger et al., 2007). It has been argued
that sexual activity during the nervous system’s development might be interpreted by the
body as a stressor and may lead to emotional instability and depression (Donovan et al.,
1988; Whitbeck et al., 1993; Allen et al., 1994; Costa et al., 1995; Harvey and Spigner,
1995; Capaldi et al., (1996); Schvaneveldt et al. 2001). From a psychological perspec-
tive, sexually active adolescents are more likely to engage in anti-social behaviour and
substance abuse, more likely to experience disrupted family relationships and more likely
to exhibit low self-esteem (Jessor et al., 1983; Mott and Marsiglio, 1985; Upchurch and
McCarthy, 1990; Dorius et al., 1993; Meilman, 1999; Rector and Johnson 2005; Schvan-
©2019 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd
To continue reading
Request your trial