Parental Divorce and Students’ Performance: Evidence from Longitudinal Data*
| Published date | 01 June 2007 |
| Date | 01 June 2007 |
| DOI | http://doi.org/10.1111/j.1468-0084.2006.00199.x |
| Author | Daniela Vuri,Anna Sanz‐de‐Galdeano |
Parental Divorce and Students’ Performance:
Evidence from Longitudinal Data*
Anna Sanz-de-Galdeanoand Daniela Vurià
Economics Department, Universitat de Girona, Girona, Spain and IZA, Bonn, Germany
(e-mail: anna.sanzdegaldeano@udg.es)
àFacolta` di Economia, Univers ita` di Roma ‘‘Tor Vergata’’, via Columbia 2, 00133 Roma,
Italy (e-mail: daniela.vuri@uniroma2.it)
Abstract
In this article, we analysed data from the National Education Longitudinal Study to
investigate whether experiencing parental divorce during adolescence had an adverse
impact on students’ performance on standardized tests. To account for the potential
endogeneity of parental divorce we employed double and triple difference models
that rely on observing teenagers from intact and divorced backgrounds before and
after the divorce occurs. We found that parental divorce does not negatively affect
teenagers’ cognitive skills. Our results also suggest that cross-sectional estimates
overstate the detrimental effect of parental divorce.
I. Introduction
Establishing whether parental divorce has a causal negative effect on children’s
outcomes is a crucial issue for the evaluation of divorce and family laws. Several
states in the USA have recently started tightening divorce requirements, reversing
*We thank the US Department of Education’s National Center for Education Statistics for access to the
public use NELS data and Jennifer Thompson for kindly answering our questions about the data files. We
would also like to thank Charles Grant, Andrea Ichino, Julia´ n Messina and Frank Vella for helpful comments
on earlier drafts, as well as seminar participants at the EUI, UNICEF, CSEF,the 2005 ESPE, EEA and ESSLE
meetings, University of Florence and Tor Vergata University (Rome). Anna Sanz-de-Galdeano acknowledges
financial support from a European Community Marie Curie Fellowship (http://www.cordis.lu/improving).
This paper was started while Daniela Vuri was at the University of Florence. The authors are responsible for
information communicated and the European Commission is not responsible for any view or results
expressed.
JEL Classification numbers: J12, C23.
OXFORD BULLETIN OF ECONOMICS AND STATISTICS, 69, 3 (2007) 0305-9049
doi: 10.1111/j.1468-0084.2006.00199.x
321
ÓBlackwell Publishing Ltd and the Department of Economics, University of Oxford, 2006. Published by Blackwell Publishing Ltd,
9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.
the liberalizing trend in divorce laws that began around 1970.
1
The proponents of
tightening the divorce regime often argue that making divorce easier has negative
consequences for children. However, as pointed out by Gruber (2004), this
argument relies on three implicit suppositions. First, that easier divorce regulations
cause an increase of divorce rates. Empirical work on this supposition has reached
mixed conclusions: while Friedberger (1998) finds that there is an impact of
unilateral divorce on divorce rates in the USA, the evidence presented by Wolfers
(2003) indicates that the increase in divorce rates is only transitional, disappearing
after a decade. Secondly, that changes in divorce regulation only have an impact on
families and children through their effect on the propensity to divorce. The third
supposition that drives criticism of easier divorce regulations, on which this article
focuses, is that divorce has an adverse impact on children.
There is an enormous literature that finds that experiencing parental divorce is
negatively related to a wide variety of children’s outcomes such as educational
attainment, fertility choices (specially non-marital birth during teenage years), future
earnings, employment status and welfare recipiency among others (many of these
studies are reviewed in Amato and Keith, 1991; Haveman and Wolfe, 1995).
However, this large literature can hardly be interpreted causally because divorce is
associated with socioeconomic characteristics that also determine children’s
attainments. For instance, there is a negative relationship between divorce and
men’s earning ability (Sander, 1986). In addition, even if socioeconomic information
is available, the question of causality is further complicated because it is unlikely that
these observable variables can fully capture the unobservable differences that may
exist between families that choose to divorce and intact families; for example, it may
be the conflict associated with divorce, rather than divorce per se, what leads to
children’s inferior outcomes. Therefore, it is easy to overstate the detrimental impact
of divorce.
Several studies have stressed the difficulties associated with the endogeneity of
parental divorce. Manski et al. (2002) present and interpret alternative estimates of
the effect of family structure on highschool graduation, obtained under differing
assumptions about the process generating family structure and highschool outcomes.
Sandefur and Wells (1997) and Bjorklund, Ginther and Sundstrom (2004) use sibling
data to control for unmeasured characteristics of families that are common to siblings.
Corak (2001) assumes that parental loss by death is exogenous and argues that
children with a bereaved background offer a benchmark to assess the endogeneity of
parental loss through divorce, considering that any difference between the outcomes
of individuals from bereaved and divorced backgrounds represents the consequences
of an endogeneity bias. In a related article, Lang and Zagorsky (2001) also consider
parental death as an exogenous source of parental absence. Gruber (2004) states that
‘what is required to appropriately identify the impact of divorce is an exogenous
1
Unilateral divorce, which requires the willingness of only one spouse to divorce, rather than the consent of
both spouses, was rare before the late 1960s but was in place in most states by the mid-1970s.
322 Bulletin
ÓBlackwell Publishing Ltd and the Department of Economics, University of Oxford, 2006
To continue reading
Request your trial