Domestic Violence and Child Mortality in the Developing World
| Published date | 01 August 2020 |
| Date | 01 August 2020 |
| Author | Zahra Siddique,Samantha Rawlings |
| DOI | http://doi.org/10.1111/obes.12357 |
723
©2020 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd.
OXFORD BULLETIN OF ECONOMICSAND STATISTICS, 82, 4 (2020) 0305–9049
doi: 10.1111/obes.12357
Domestic Violence and Child Mortality in the
Developing World*
Samantha Rawlings† and Zahra Siddique‡
†Department of Economics, University of Reading, Whiteknights, Reading RG6 6AA, UK.
(e-mail: s.b.rawlings@reading.ac.uk)
‡Department of Economics, University of Bristol, The Priory Road Complex, Priory Road,
Bristol BS8 1TU, UK. (e-mail: zahra.siddique@bristol.ac.uk)
Abstract
We examine the effect of domestic violence on child mortality using Demographic and
Health Surveys from thirty-two developing countries. We first examine conditional asso-
ciations between violence faced by the mother and child mortality after controlling for
observable confounders. Children of (ever) physically victimized mothers are 0.4, 0.7,
and 1.0 pp more likely to die within thirty days, a year, and five years of being born. We
find similar associations when examining violence experienced in the last twelve months,
although these are no longer statistically significant. The association is statistically signif-
icant, and larger, if the mother experiences violence in the last twelve months often, rather
than sometimes. Violence is significantly associated with pregnancy loss, suggesting the
true effect on mortality is larger than estimates based on live-births would suggest. We in-
vestigate robustness of associations to omitted variable bias, assessing the role of selection
on unobservables to estimate lower bounds on the true effect. These continue to indicate
economically meaningful positive effects, suggesting selection on unobservables would
need to be 2.4–3 times that of selection on observables to nullify the estimated effect. We
provide evidence that maternal smoking and breastfeeding practices are mediators in the
relationship between domestic violence and child mortality.
I. Introduction
Domestic violence, defined as physical and/or sexual violence by an intimate partner, has
important consequences for large numbers of female victims across the globe. Using data
JEL Classification numbers: I14, J12, J13.
*The authors have no conflicts of interest to declare. Earlier versions of this paper were circulated as ‘Domestic
violence and child mortality’and ‘Domestic abuse and child health’. Wehave benefited from comments by seminar and
conference participants at the University of Reading, CESIfo Munich, the Institute of Development and Economic
Alternatives in Lahore (Pakistan), the University of Cologne, the 2015 IZA/World Bank conference, the Royal
Economic Society Conference, the European Society of Population Economics conference as well as comments by
Sonia Bhalotra, Marina Della Giusta, Uma Kambhampati, Andrew Oswald,Christopher Taber and Charlotte Watts.
Our thanks also to the editor, Climent Quintana-Domeque, and twoanonymous referees for comments and suggestions
which greatly improved the paper.All errors are our own.
724 Bulletin
from 81 countries, Devries et al. (2013) note that ‘globally,in 2010, 30.0% [95% confidence
interval (CI) 27.8 to 32.2%] of women aged 15 and over have experienced, during their
lifetime, physical and/or sexual intimate partner violence’. These prevalence rates are
particularly high in developing regions such as Central sub-Saharan Africa (65.64%) and
South Asia (41.73%).
Domestic violence has direct costs borne by victims: it is one of the leading causes of
homicide deaths among women, as well as being associated with poor health and reduced
earnings. An empirical literature in public health and medicine has also examined the
relationship between domestic violence experienced by mothers and health outcomes of
their children, documenting a negative correlation between the two(see Yount, DiGirolamo
and Ramakrishnan, 2011). However, much of this literature uses small non-random samples
from developed countries and does not account for omitted variables whichare potentially
correlated with both domestic violence and child health.1We use nationally representative
data sets from thirty-two developing countries to investigate this relationship, and we
provide a careful examination of potential omitted variable bias.
Two relatively recent studies in health economics examine the relationship between
assaults during pregnancy and birth outcomes using large scale administrative data from
the US. Aizer (2011) uses data on female hospitalizations and birth outcomes for the state
of California between 1991 and 2002. She finds that serious incidents of domestic violence
(where serious is defined as resulting in hospitalization) cause a reduction in birth weight
of 163 g. Currie, Mueller-Smith and Rossin-Slater (2018) examine the impact of assaults
during pregnancy on infant health outcomes using linked administrative data from New
York City and find a robust negative effect of assaults on birth outcomes. A key distinction
of our paper from both Aizer (2011) and Currie, Mueller-Smith, and Rossin-Slatere (2018)
is that by the nature of their data they focus on serious cases of domestic violence, that
result in either hospitalization (in the case of Aizer, 2011), or a report of assault to the police
(in the case of Currie et al., 2018). Instead, we examine both severe and less severe forms
of violence that occur within the home but do not result in hospitalization or reporting to
the authorities. A further difference is the geographical focus of our study. While Aizer
(2011) and Currie et al. (2018) focus on the US (California and New York City), our study
is about developing countries. Rates of domestic violence in developing countries tend to
be higher, particularly violence experienced in the last 12 months. In the US, around 30%
of women experience physical domestic violence in their lifetime, and around 4% in the
last 12 months (Black, Basile and Breiding, 2011). In contrast, in our sample, rates are
higher in the majority of countries, with highs of almost 60% lifetime prevalence and 50%
in the last 12 months (Figures A2 and A3).
A key contribution of our paper is the use of data from the Demographic and Health
Surveys (DHS), containing comparable information on domestic violence and child mor-
1A large-scale study was carried out by Silverman et al. (2006), who use data on women giving birth in 26 US
states as part of the 2000–03 Pregnancy Risk Assessment Monitoring System. They find that women experiencing
intimate partner violence are at increased risk for poor maternal health (such as high blood pressure, vaginal bleeding,
severe nausea/vomiting/dehydration,kidney or urinary tract infections, frequent hospital visits) as well as poor infant
health (such as delivery preterm, low birth weight, more likely to require intensivecare).
©2020 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd
To continue reading
Request your trial