Hidden in Plain Sight: A Statistical Analysis of Violence Against Children (2014)

HIDDEN IN
PLAIN SIGHT
A statistical analysis of violence
against children
Cover photo: A 7-year-old girl who
was sexually abused by an 18-year-
old boy at the Mena police station in
Makeni, Bombali district, Sierra Leone.
The boy, who is a close neighbour,
forced her to have sex with him.
© UNICEF/SLRA2013-0006/Asselin
© United Nations Children’s Fund

(UNICEF), Division of Data, Research
and Policy, September 2014
ISBN: 978-92-806-4767-9
Permission is required to reproduce

any part of this publication.
Permission will be freely granted
to educational or non-profit
organizations.
To request permission and for any

other information on the publication,
please contact:
UNICEF
Data and Analytics Section
Division of Data, Research and Policy
3 United Nations Plaza
New York, NY 10017, USA
Tel: +1 212 326 7000
Email: [email protected]
All reasonable precautions have

been taken by UNICEF to verify
the information contained in this
publication. For any data updates
subsequent to printing, please visit
<data.unicef.org>.
Suggested citation: United Nations

Children’s Fund, Hidden in Plain
Sight: A statistical analysis of violence
against children, UNICEF, New York,
2014.
HIDDEN IN
PLAIN SIGHT
A statistical analysis of violence against children
© UNICEF/NYHQ2006-1487/Pirozzi
ACKNOWLEDGEMENTS
This is a report of UNICEF’s Division of Data, Research and Policy. It was produced under the direction of Tessa Wardlaw,
Chief, Data and Analytics Section.
The preparation of the report was initiated and coordinated by Claudia Cappa, who was also responsible for data analysis,
interpretation of the results and report writing. Nicole Petrowski facilitated the preparation of the statistical figures,
contributed to the writing of the report and participated in the document review. Laura Cordisco Tsai (independent
consultant) was involved in data compilation and fact-checking, and prepared summaries of relevant literature. Yadigar
Coskun provided support for data processing and tabulations. Colleen Murray assisted in the calculation of global
estimates and preparation of statistical tables. Janet Njelesani (independent consultant) participated in the document
review. The report was edited by Lois Jensen, copy-edited by Tina Johnson, fact-checked by Baishalee Nayak and
designed by Era Porth (independent consultants).
Valuable feedback and inputs were provided by Tatiana Aderikhina (UNICEF Kazakhstan), Nadi Albino (UNICEF South
Africa), Line Baago Rasmussen (UNICEF Zimbabwe), Caroline Bakker (UNICEF Sri Lanka), Ramiz Behbudov (UNICEF
Azerbaijan), Casimira Benge (UNICEF Brazil), Patrizia Benvenuti (UNICEF South Africa), Stephen Blight (UNICEF Regional
Office for East Asia and the Pacific), Andrew Brooks (UNICEF Regional Office for West and Central Africa), Clarice
Da Silva e Paula (Child Protection Section, UNICEF headquarters), Karin Heissler (Child Protection Section, UNICEF
headquarters), Keshani Hiranthika (UNICEF Sri Lanka), Noriko Izumi (UNICEF Zimbabwe), Severine Jacomy Vite (UNICEF
Turkey), Hemamal Jayawardena (UNICEF Sri Lanka), Shane M. Khan (independent consultant), Theresa Kilbane (Child
Protection Section, UNICEF headquarters), Nela Kmic Brkovic (UNICEF Montenegro), Joost Kooijmans (Child Protection
Section, UNICEF headquarters), Jean-Claude Legrand (UNICEF Regional Office for Central and Eastern Europe and the
Commonwealth of Independent States), Seamus Mac Roibin (UNICEF South Africa), Paolo Marchi (UNICEF Bosnia and
Herzegovina), Tiina Markkinen (UNICEF China), Mendy Marsh (Child Protection Section, UNICEF headquarters), Amalee
McCoy (independent consultant), Nadine Perrault (UNICEF Regional Office for Latin America and the Caribbean), Fiorella
Polo (Water and Sanitation Section, UNICEF headquarters), Rashad H. Abdel Rahman (independent consultant), Clara
Sommarin (Child Protection Section, UNICEF headquarters), Violet Speek-Warnery (UNICEF Eastern Caribbean Area),
Gary Stahl (UNICEF Brazil), Heather Stewart (UNICEF Eastern Caribbean Area), Rosana Vega (UNICEF Bolivia) and
Danzhen You (Data and Analytics Section, UNICEF headquarters).
Gratitude also goes to those who facilitated access to relevant datasets: Donneth Ann Marie Edmondson (UNICEF
Jamaica), Alma Jenkins Acosta (UNICEF Panama), Elena Laur (UNICEF Moldova), Oladimeji Olowu (UNICEF Eastern
Caribbean Area), Manar Soliman (UNICEF Egypt), Vicente Teran (UNICEF Regional Office for Latin America and the
Caribbean), Frederic Unterreiner (UNICEF Eastern Caribbean Area) and Sabastian Waisgrais (UNICEF Argentina).
UNICEF gratefully acknowledges the inputs provided by the Global Initiative to End All Corporal Punishment of Children
(Elinor Milne and Peter Newell), the Health Behaviour in School-aged Children Study team (Lise Augustson and Oddrun
Samdal), the Statistics and Surveys Section of the United Nations Office on Drugs and Crime (Jenna Dawson-Faber and
Felix Reiterer) and the Health Systems and Innovation Cluster of the World Health Organization (Colin Douglas Mathers
and Wahyu Retno Mahanani).
This report was made possible through core funding to UNICEF. It is based on research funded in part by the Bill & Melinda
Gates Foundation. The findings and conclusions contained within are those of UNICEF and do not necessarily reflect
positions or policies of the Bill & Melinda Gates Foundation.
Foreword
Violence against children occurs every day, everywhere: the slaps of an
upset parent to control an ‘unruly’ child, the sexual victimization of a teen-
ager by a peer or a neighbour, the bullying of one child by another in the
schoolyard, the emotional degradation of a child bride by her spouse. Too
many children worldwide are affected by such violence, yet it is rarely ac-
knowledged, in part because it is so commonplace. The repercussions are
not inconsequential, with ripple effects throughout society as well as fu-
ture generations.
Everyday violence may be pervasive, but it is not inevitable. The first step
in curbing all forms of violence against children is bringing the issue to
light – in all its complexity. Despite the difficulties in measuring violence
against children, and considerable gaps, an unprecedented volume of data
on the subject has become available over the last two decades that is pro-
viding the evidence countries need to develop effective policies, legislation
and programmes to address violence. Solid data and research are essen-
tial in bringing the issue out of the shadows. They are also important in
revealing hidden attitudes and social norms that may perpetuate violence
against children and factors that may place certain children at higher risk.
The Convention on the Rights of the Child guarantees that children every-
where should live free from all forms of violence. For this to happen, the
true nature and extent of the problem must be documented. It is to that
end that this report is dedicated.
Jeffrey O’Malley
Director, Division of Data, Research and Policy, UNICEF
contents
01 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Measuring violence against children
The scope and structure of this report
A word of caution
02 Using data to make violence visible . . . . . . 14

Data sources
Methodological challenges
Ethical considerations
A moral imperative and strategic investment
03 Physical violence: A leading cause of

injury and death among children . . . . . . . 30
Wide-ranging consequences
Data sources
Preventable deaths
Homicides
Impact of interpersonal violence on mortality and morbidity
Non-fatal physical violence among adolescent girls
Non-fatal physical violence among adolescent boys
04 Sexual violence: Not limited to girls . . . . 60

The particular vulnerability of children
The far-reaching consequences of sexual violence
What remains hidden in most data collection efforts
Data sources
Sexual violence against girls
Sexual violence against boys
Age at first experience of sexual violence among girls
Age at first experience of sexual violence among boys
Forced first sex among girls and women
Forced first sex among boys and men
Perpetrators of sexual violence against girls
Perpetrators of sexual violence against boys
Typical settings where sexual violence occurs
The sexual exploitation of children
05 The most common form of violence

against children. . . . . . . . . . . . . . . . . . . . . . . . 94
Use of disciplinary methods
Violent disciplinary practices and children’s characteristics
Violent discipline and the economic and social status of households
2
HIDDEN IN PLAIN SIGHT
06 When children hurt one another. . . . . . . . 112

Physical attacks
Fighting
Bullying
07 Violent unions among adolescents . . . . . . 130

Lifelong – sometimes intergenerational – consequences
Prevalence of partner violence against ever-married adolescent girls
Partner violence initiated by adolescent girls
Prevalence of partner violence against ever-married adolescent boys
Dating violence
08 Exploring attitudes and social norms . . . 146

Attitudes towards wife-beating
Attitudes towards corporal punishment of children
Comparison of attitudes towards different forms of violence
Attitudes towards child sexual abuse
09 Key findings and CONCLUSIONS. . . . . . . . . . . . 164

Key findings
In our hands
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Statistical table . . . . . . . . . . . . . . . . . . . . . . . . 196
Panels
Defining violence against children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Violence in the life cycle of children. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
A note on the data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Emotional violence. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Violence against children in institutional care. . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
When different forms of violence overlap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Disclosing experiences of violence and seeking help. . . . . . . . . . . . . . . . . . . . . . 88
Two billion children without full legal protection . . . . . . . . . . . . . . . . . . . . . . . . 110
Steps to ending a global problem. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
3
Defining violence
physical Violence against children includes all corporal punishment and all other forms
of torture, cruel, inhuman or degrading treatment or punishment as well as physical bullying and hazing
by adults or by other children. ‘Corporal’ (or ‘physical’) punishment is defined as any punishment in which
physical force is used and intended to cause some degree of pain or discomfort, however light. Most involves
hitting (‘smacking’, ‘slapping’, ‘spanking’) children with the hand or with an implement – a whip, stick,
belt, shoe, wooden spoon, etc. But it can also involve, for example, kicking, shaking or throwing children,
scratching, pinching, biting, pulling hair or boxing ears, caning, forcing children to stay in uncomfortable
positions, burning, scalding or forced ingestion.
sexual Violence comprises any sexual activities imposed by an adult on a child against which
the child is entitled to protection by criminal law. This includes: (a) The inducement or coercion of a child to
engage in any unlawful or psychologically harmful sexual activity; (b) The use of children in commercial sexual
exploitation; (c) The use of children in audio or visual images of child sexual abuse; and (d) Child prostitution,
sexual slavery, sexual exploitation in travel and tourism, trafficking for purposes of sexual exploitation (within
and between countries), sale of children for sexual purposes and forced marriage. Sexual activities are also
considered as abuse when committed against a child by another child if the offender is significantly older than
the victim or uses power, threat or other means of pressure. Consensual sexual activities between children are
not considered as sexual abuse if the children are older than the age limit defined by the State Party.
mental Violence is often described as psychological maltreatment, mental abuse, verbal

abuse and emotional abuse or neglect. This can include: (a) All forms of persistent harmful interactions with
a child; (b) Scaring, terrorizing and threatening; exploiting and corrupting; spurning and rejecting; isolating,
ignoring and favouritism; (c) Denying emotional responsiveness; neglecting mental health, medical and
educational needs; (d) Insults, name-calling, humiliation, belittling, ridiculing and hurting a child’s feelings;
(e) Exposure to domestic violence; (f) Placement in solitary confinement, isolation or humiliating or degrading
conditions of detention; and (g) Psychological bullying and hazing by adults or other children, including
via information and communication technologies (ICTs) such as mobile phones and the Internet (known
as ‘cyber-bullying’).
Neglect or negligent treatment means the failure to meet children’s

physical and psychological needs, protect them from danger or obtain medical, birth registration or other
services when those responsible for their care have the means, knowledge and access to services to do so. It
includes: (a) Physical neglect: failure to protect a child from harm, including through lack of supervision, or
to provide a child with basic necessities including adequate food, shelter, clothing and basic medical care; (b)
Psychological or emotional neglect, including lack of any emotional support and love, chronic inattention,
caregivers being ‘psychologically unavailable’ by overlooking young children’s cues and signals, and exposure
to intimate partner violence or drug or alcohol abuse; (c) Neglect of a child’s physical or mental health:
withholding essential medical care; (d) Educational neglect: failure to comply with laws requiring caregivers to
secure their children’s education through attendance at school or otherwise; and (e) Abandonment.
These definitions have been adapted from: United Nations Committee on the Rights of the Child, General Comment No. 13 (2011): The right
of the child to freedom from all forms of violence, UN document CRC/C/GC/13, Office of the High Commissioner for Human Rights, Geneva,
18 April 2011. In addition to the main definitions listed here, the comment also defines ‘torture and inhuman or degrading treatment or
punishment’, ‘violence among children’, ‘self-harm’, ‘harmful practices’, ‘violence in the mass media’, ‘violence through information and
communication technologies’, and ‘institutional and system violations of child rights’. The comment also specifies types of physical violence
4 to which children with disabilities may be subjected.
against children
“All forms of violence against children, however light,

are unacceptable. […] Frequency, severity of harm and
intent to harm are not prerequisites for the definitions
of violence. States parties may refer to such factors in
intervention strategies in order to allow proportional
responses in the best interests of the child, but
definitions must in no way erode the child’s absolute
right to human dignity and physical and psychological
integrity by describing some forms of violence as
legally and/or socially acceptable.”
United Nations Committee on the Rights of the Child, General

Comment No. 13 on the Convention on the Rights of the Child
© UNICEF/NYHQ2001-0594/DeCesare
5
Chapter 1 OVERVIEW
The protection of children from all forms of violence
is a fundamental right guaranteed by the Convention
on the Rights of the Child and other international
human rights treaties and standards. Yet violence
remains an all-too-real part of life for children around
the globe – regardless of their economic and social
circumstances, culture, religion or ethnicity – with
both immediate and long-term consequences.
Children who have been severely abused or
neglected are often hampered in their development,
experience learning difficulties and perform poorly at
school. They may have low self-esteem and suffer
from depression, which can lead, at worst, to risky
behaviours and self-harm. Witnessing violence can
cause similar distress. Children who grow up in a
violent household or community tend to internalize
that behaviour as a way of resolving disputes,
repeating the pattern of violence and abuse against
their own spouses and children. Beyond the tragic
effects on individuals and families, violence against
children carries serious economic and social costs
in both lost potential and reduced productivity
(see Box 1.1).
Over the last decade, recognition of the pervasive

nature and impact of violence against children
has grown. Still, the phenomenon remains largely
undocumented and underreported. This can be
attributed to a variety of reasons, including the
fact that some forms of violence against children
are socially accepted, tacitly condoned or not
perceived as being abusive. Many victims are too
young or too vulnerable to disclose their experience
or to protect themselves. And all too often when
A PASSPORT TO PROTECTION
victims do denounce an abuse, the legal system

© UNICEF/DRCA2010-00013/Connelly
fails to respond and child protection services are

unavailable. The lack of adequate data on the issue
is likely compounding the problem by fuelling the
misconception that violence remains a marginal
phenomenon, affecting only certain categories of
children and perpetrated solely by offenders with
biological predispositions to violent behaviour.
6
Measuring violence against authority figures, peers and strangers, both within
children and outside the home. The report does not cover
certain forms of violence that take place within
Measuring the breadth and depth of violence against
the context of shared community, cultural or social
children is complicated by the fact that violence can
norms and values, like female genital mutilation/
take many forms (physical, sexual and emotional),
cutting (FGM/C), as this harmful traditional practice
occur in any setting (including the home, school,
occurs under specific circumstances and has been
workplace and over the Internet) and be perpetrated
addressed in other publications.2 Two additional
by individuals (parents and other caregivers, peers,
categories of violence are also outside the scope
intimate partners, authority figures and strangers)
of this report: self-directed violence and collective
or groups. A thorough assessment of its nature
violence. The former has been defined as violence a
and prevalence requires reliable statistics on all
person inflicts upon himself or herself (for example,
types of violence to which children are exposed as
suicide or other forms of self-abuse), while the
well as the range of circumstances surrounding its
latter is inflicted by larger entities such as States,
occurrence. These data needs, however, remain
organized political parties, terrorist organizations
largely unmet. Certain forms of violence are better
and other armed groups.3
documented than others, depending on a country’s
capacity and investment in data collection. Yet even
Given the general lack of uniformity in the way
when data are available, the quality and scope of the
data on violence against children are collected, this
information are often limited. For instance, mortality
report relies mainly on information gathered through
data, including counts of homicides, are available
internationally comparable sources, including
for many countries, but errors and incomplete
the UNICEF-supported Multiple Indicator Cluster
information in birth and death registries, challenges
Surveys (MICS), the US Agency for International
related to correctly determining causes of death,
Development (USAID)-supported Demographic
and weaknesses in countries’ mechanisms for
and Health Surveys (DHS), the Global School-based
recording vital events can affect their accuracy.
Student Health Surveys (GSHS) and the Health
In addition, detailed information on the victims
Behaviour in School-aged Children Study (HBSC).4
themselves and the circumstances surrounding
These international survey programmes have been
these fatalities are rarely recorded. Representative
almost exclusively implemented in low- and middle-
data on particular forms of interpersonal violence,
income countries (with the exception of the HBSC).
including violent discipline and bullying, have
So while the focus of this report is largely on these
increased over the last two decades, mainly through
countries, this should in no way be interpreted to
large-scale population-based surveys. However,
suggest that violence against children is not found
data on other forms of abuse, including systematic
in high-income nations. To that end, the report also
statistics on sexual violence against boys, remain
uses country-specific facts or evidence derived from
woefully lacking.
small-scale studies and national surveys to shed light
on certain aspects or circumstances from a variety
The scope and structure of countries for which representative or comparable
of this report data are unavailable. Using these combined sources,
While acknowledging these limitations, this report the report draws on data from 190 countries and
makes use of available evidence to describe represents the largest compilation of statistics to
what is currently known about global patterns of date on violence against children.
violence against children, using data compiled
CHAPTER 1: OVERVIEW
from a selection of sources. The analyses focus The presentation of the data has been organized into
primarily on forms of interpersonal violence, two main sections. The first section covers three
defined as violent acts inflicted on children by main types of violence experienced by children
another individual or a small group.1 The types of and committed by anyone in all possible settings:
interpersonal violence covered include those mainly physical acts of violence (both fatal and non-fatal),
committed by caregivers and other family members, emotional violence and sexual violence. The second
7
half of the report explores in more detail specific
manifestations of these forms of violence: violent
discipline in the home, peer violence (including
involvement in physical fights and bullying) and
intimate partner violence among adolescents.
The report also explores attitudes towards some
forms of violence to provide insights into deeply
rooted cultural beliefs that may help to explain their
persistence.
Each chapter follows a similar structure, beginning

with a brief overview of the definitions for the specific
type of violence covered and current knowledge
about potential risk factors and consequences.
Regional and/or country-level data, depending on
availability, are then presented. When possible,
disaggregated data are also included to reveal
certain characteristics of children who experience
violence, including information on their families
as well as some contextual factors surrounding
their experiences of violence such as the identity
of the most common perpetrators. Boxes within
each chapter highlight specific issues relevant to
the type of violence covered that deserve special
consideration.
A word of caution
While specific forms of violence have a distinctive
nature and can occur in isolation, any attempt
to ‘categorize’ violence is a somewhat artificial
undertaking. For one thing, the boundaries
between acts of violence tend to become blurred.
Sexual violence is often inflicted through the use
of physical force and/or psychological intimidation.
Moreover, experiences of violence often overlap.
While some children may experience rare and
isolated incidents of aggression, others may find
themselves repeatedly exposed to multiple forms
of abuse.
In addition to the possible overlap of various types of

violence, children can be victims, perpetrators and
CHAPTER 1: OVERVIEW
witnesses to violence – all at the same time. Children

who grow up in societies characterized by the
systemic use of violence by terrorist organizations
or other armed groups are at heightened risk
of interpersonal victimization and often end up
becoming violent themselves.5 Moreover, those
8
who suffer violence are sometimes driven to hurt

themselves in response to their own victimization.
Certain research, for example, has demonstrated a
direct association between experiences of physical
or sexual abuse and suicidal thoughts and behaviours
among adolescents.6 Rather than turning aggression
on oneself, another possible manifestation of being
victimized or witnessing abuse is to become angry
or hostile towards others. To take one example,
observing violence between parents or being the
target of child abuse is closely associated with the
perpetration of dating violence.7 All of these points
are important to keep in mind when reading this
report, since the reality of violence against children
is far more confounding and multidimensional than
any attempts to categorize and quantify it.
“1. States Parties shall take

all appropriate legislative,
administrative, social and
educational measures to protect
the child from all forms of physical
or mental violence, injury or abuse,
neglect or negligent treatment,
maltreatment or exploitation,
including sexual abuse, while in the
care of parent(s), legal guardian(s) or
any other person who has the care
of the child.
2. Such protective measures should,
as appropriate, include effective
procedures for the establishment
of social programmes to provide
necessary support for the child and
for those who have the care of the
child, as well as for other forms of
prevention and for identification,
reporting, referral, investigation,
treatment and follow-up of instances

CHAPTER 1: OVERVIEW
of child maltreatment described

heretofore, and, as appropriate, for
judicial involvement.”
Article 19 of the Convention on
the Rights of the Child
9
Box 1.1
The social and economic costs of

violence against children
Numerous studies have analysed Children subjected to violence are encing multiple forms of violence
the detrimental effects that violence also more likely than other children were highly likely to become ‘bully-
may have on behavioural, emo- to exhibit behavioural problems in victims’ at school (those who expe-
tional and psychological function- school. Childhood experiences of rience bullying while also bullying
ing, which have been found to vary abuse have been found to be asso- other children), extending the con-
according to the nature, extent and ciated with poor classroom learning sequences of child abuse from the
severity of exposure.8 The effects behaviour, suspensions from school victims themselves to their school-
of abuse on future social and eco- and absenteeism.20 A prospective based peers.25
nomic well-being have also been longitudinal study of randomly se-
explored, as outlined in the key find- lected children in the United States
Economic impact on individuals
ings below. 9 found that children subjected to vio-
lence were absent from school al- Several recent studies in the United
most twice as many days and were States have confirmed the dam-
Impact on educational performance suspended from school nearly twice aging economic consequences of
and achievement as often as other children.21 child abuse.26 Currie and Widom
The experience of violence while (2010), for example, conducted a
Experiences of violence at school prospective cohort study in which
growing up has serious conse-
can also hinder children’s educa- they matched children who had ex-
quences for children’s academic
tional progress. Victimization by perienced court-substantiated cases
success,10 including a heightened
teachers or peers may make chil- of physical and sexual abuse before
risk for many forms of educational
dren afraid of going to school and age 11 with a non-abused control
underachievement.11 Children who
interfere with their ability to concen- group. They followed both groups
are victims of violence are at higher
trate and participate in school activi- of children into middle adulthood
risk of dropping out of both high
ties. In severe cases, experiences of (that is, into their thirties and for-
school12 and college13 than their non-
school violence can result in truan- ties), comparing their economic out-
abused peers. Moreover, research-
cy and/or school drop-out.22 For in- comes. This study found that adults
ers have found that children who
stance, a nationally representative with documented histories of abuse
have been subjected to violence
survey of primary school students were, on average, 14 per cent less
are less likely than other children
in Ethiopia found that exposure to likely than the control group to be
to anticipate attending college.14
school violence reduced girls’ class employed and significantly less like-
In fact, experiences of childhood
participation, lowered their school ly to own assets (such as a home,
abuse have been found to be a good
performance, increased their grade vehicle, stock or bank account). Fur-
predictor of lower grades,15 lower IQ
repetition and increased their drop- thermore, the study estimated that
test scores16 and lower levels of edu-
out rates.23 childhood experiences of abuse re-
cational self-efficacy.17 Child victims
duced a person’s earning potential
of violence are also more likely to Children who experience multiple by an average of about US$5,000
be referred for special education.18 forms of violence are at especially per year.27
For instance, one longitudinal study high risk of poor educational out-
found that children who had been comes. In particular, children who Using data from a nationally rep-
maltreated were twice as likely as are bullied by peers at school while resentative study – the National
their non-abused peers to enter into also experiencing abuse at home or Comorbidity Survey in the United
special education, after controlling in their community are at consider- States – and controlling for key so-
for other developmental risk factors ably higher risk of poor academic cial and economic variables, Ziel-
at the child, family and community performance.24 Additionally, one inski (2009) found that individuals
levels.19 study found that children experi- who experienced child abuse were
10
at increased risk of financial and a financially harmful job loss in their direct and indirect costs due to in-
employment-related difficulties in household in the past year. Child- creased social spending and lost
adulthood. Adults who had experi- hood experiences of each form of economic productivity. In 2012, Pre-
enced any form of abuse were twice violence had a detrimental effect on vent Child Abuse America released
as likely as their non-abused peers victims’ income in adulthood. This an economic analysis positing that
to be unemployed and 80 per cent was particularly the case for those the prevalence of child abuse and
more likely to have had someone in who experienced sexual abuse, neglect in the United States costs
their household lose a job in the last who were 80 per cent more likely over $80 billion annually.31 The
12 months. Individuals who had ex- than non-victims to live below the study’s calculations included direct
perienced multiple forms of abuse federal poverty line and 90 per cent costs of abuse (related to medical
were particularly affected. The more likely to have family earnings treatment, mental health services,
study also revealed that abuse vic- that fell within the lowest income the child welfare system and law
tims were more than twice as likely quartile.29 enforcement) as well as indirect
to fall below the federal poverty line costs (related to special education,
In economic terms at least, women early intervention services, emer-
and almost twice as likely to live in
appear to be more severely affect- gency/transitional housing, physi-
a household with a total income in
ed by childhood abuse than men. cal and mental health care, juvenile
the lowest 25 per cent (quartile) of
According to Currie and Widom’s delinquency, adult criminal justice
the population. While those who
study, women who experienced vi- costs and lost worker productivity).
had experienced abuse were more
olence as children exhibited signifi-
than twice as likely as those who
cantly lower economic outcomes Fang et al. (2012) estimated the life-
had not to depend on Medicaid and/
than the control group in both time economic costs of new cases
or other public funding for health young and middle adulthood. As of child abuse in the United States
care, they were less likely overall to young adults, women who experi- in 2008 at $124 billion (in 2010 dol-
have health insurance of some kind.28 enced violence had completed few- lars). The calculation included life-
er years of education and had lower time costs of $210,012 per abuse
The same study also demonstrated
IQ test scores. The difficulties per- victim who survived and $1.3 mil-
divergent outcomes according to
sisted, with women in their thirties lion per victim who died. The esti-
the type of maltreatment. Adults
and forties earning less and being mated lifetime cost was comprised
with a history of physical abuse
less likely to own assets than their of productivity losses as well as spe-
in childhood suffered the greatest
peers who had not experienced vio- cial education, medical and health
employment-related difficulties in
lence in childhood. Although male care, child welfare and criminal jus-
adulthood: They were 140 per cent
victims of violence were significant- tice costs emerging from children’s
more likely to be unemployed and
ly less likely than their non-abused experiences of abuse, with the
80 per cent more likely to have ex- peers to have a skilled job in young largest component stemming from
perienced a financially harmful job adulthood, this trend became non- productivity losses.32 In another
loss in their household in the past significant for men in their thirties study, Fang et al. (2013) estimated
year than those who had not been and forties. The study suggests the the economic cost of child abuse in
victims of physical abuse. While no disproportionate long-term burden East Asia and the Pacific to exceed
significant impact on employment of childhood abuse on women’s $160 billion (in 2004 dollars) based
status was found among those who economic well-being.30 Further re- on economic losses due to death,
experienced sexual abuse or severe search is needed, however, to un- disease and health risk behaviours
neglect as children, individuals who derstand the social and economic attributable to child abuse.33
were sexually violated were 50 per impact of child abuse by sex.
cent more likely than non-victims to The findings from these studies
live in a household in which some- confirm the considerable social and
CHAPTER 1: OVERVIEW
one had lost his or her job in the Economic impact on society
economic burden of child abuse.
last year. Those who suffered se- Beyond the toll on individuals, vio- From a cost-benefit perspective
vere neglect were 60 per cent more lence against children inflicts dam- alone, they highlight the critical im-
likely than non-victims to have had age at the societal level, including portance of investing in prevention.34
11
Violence in the life cycle of children
The passage from childhood to adulthood is marked by a series of life stages involving transitions in a child’s cognitive,
physical and moral development. These stages represent opportunities for building children’s capacities, life skills and
resilience.35 At the same time, they are associated with certain risk factors for various forms of violence.
As girls and boys develop, they become more independent and begin interacting with people other than their primary
caregivers. This increases the likelihood of abuse by both caregivers (often in response to a child’s increased independence)
and people outside the home. The types of violence that children are exposed to change as they age, as do the possible effects.
Prenatal period and birth early childhood
The experience of violence can affect the A growing body of evidence suggests that infants and
foetus in the womb. More specifically, when young children are particularly at risk of violence by
maternal levels of the stress hormone cortisol primary caregivers and other family members because
are raised during pregnancy, the result is poor of their dependence and limited social interactions
foetal growth (which is linked to subsequent outside the home.42 Children can be inadvertently
development of adult diseases) and negative hurt in incidents of domestic violence directed to their
effects on brain development.38 The stress of mothers, or they may be intentionally threatened or
violence and abuse on a pregnant woman hurt in retaliation against their mothers. “Even when
may lead to other consequences as well, they are apparently lying passively in their cots, infants
such as failure to obtain adequate nutrition are sensitive to their surroundings and especially to
and medical care in pregnancy, which can the emotional signals given out by their caregivers,
lead to low birth weight and other adverse including the caregiver’s depressed, anxious, fearful
outcomes for both mother and child.39 In or angry mood.”43 Research suggests that exposure to
certain instances, miscarriage and stillbirth extreme trauma and toxic stress during the early years
may be attributable to violence.40 And has serious effects on the organization of the brain.44
research confirms that pregnant women who Toxic stress occurs when children experience prolonged,
have been subjected to or witnessed abuse strong and/or frequent adversity, such as physical,
are more likely to have complications during emotional or sexual violence and/or chronic neglect,
childbirth.41 At its most extreme, sex-selective without adequate adult support. This can disrupt the
infanticide can be quite common in societies development of the brain, increasing children’s risk for
where girls and women are thought to be of developmental delays as well as stress-related health
lower status than boys and men. problems – such as heart disease, substance abuse
and depression – later in life.45 Exposure to violence
can also result in disrupted attachment to caregivers
as well as ‘regressive’ symptoms such as increased
bedwetting, delayed language development and
additional anxiety over separation from parents, which
in turn may affect children’s ability to learn and to get
along with other adults and with peers.
12
Early and prolonged exposure to violence has the potential for the most adverse and long-lasting impact since it affects the
subsequent chain of development.36 Younger children appear to exhibit higher levels of psychological and emotional distress
as a result of violence than older children. This is because older children possess more fully developed cognitive abilities that
may allow them to understand the violence and adopt coping strategies to alleviate its upsetting effects.37
The forms of violence that children are exposed to tend to increase with age. However, the
potential impact of violence is greater at younger ages, with early and prolonged exposure
having potentially more adverse effects.
middle childhood Late childhood and adolescence
As girls and boys move towards middle In most contexts, puberty marks an important
childhood, they are faced with increased transition in girls’ and boys’ lives. At this stage,
exposure to interpersonal violence. Research gender, sexuality and sexual identity begin to assume
has found that children aged 5 to 9 years are greater importance.48 At the same time, evidence
more likely to experience violent punishment also points to increased vulnerability to particular
than older children, often as a result of their forms of violence. Teenagers, especially boys, may
new-found independence (typically reinforced become the targets of violence based on their non-
by attendance in school) and the increased traditional sexual orientation or gender identity.49
influence of peers.46 Interactions with peers Girls are at greater risk of sexual violence and forced
and participation in school activities are also or early marriage than most boys, with implications
associated with new forms of victimization, for the transmission of HIV. For many girls, the first
including, for instance, sexual harassment experience of sexual intercourse in adolescence is
and violence on the way to and from school, unwanted or even coerced.50 Young brides are at
corporal punishment by teachers and bullying greater risk than older brides of physical, emotional
by schoolmates. Evidence of the impact and sexual violence and of curtailed personal
of school violence shows compromised freedom and decision-making power.51
attendance, lower academic results and higher Puberty is also a period in which girls and boys are
drop-out rates. Among girls, school violence more likely to engage in risky behaviours such as
can also have an impact across generations, drug and alcohol abuse and unprotected and unsafe
resulting in higher fertility rates, lower health sex that increase their vulnerability to victimization.52
status and a weakened household economy.47 In addition, adolescence is associated with increased
© UNICEF/NYHQ2004-0741/Holmes
self- and political awareness and with independent

interactions outside the household.53 In some
settings – particularly conflict and armed violence –
this may make involvement in political movements,
armed struggles and criminal activities compelling,
particularly for disadvantaged and marginalized
children.54
Impact of violence on child development Forms of violence to which children may be exposed
13
Chapter 2 Using data
to make
violence
visible
Evidence – in the form of reliable data – is key to
unveiling the hidden nature of violence against
children and beginning to erode its social acceptance.
Robust data are also crucial in mobilizing resources
and action, informing the development of policies,
legislation and programmes and instituting a sound
monitoring process to assess results and impact.1
A full understanding of violence against children

requires many types of information. Prevalence
estimates are needed to draw the ‘big picture’ –
the magnitude of the problem – while information
on the family and the social environment in which
children live can reveal factors that may increase or
mitigate risk. Insights into perceptions, attitudes,
beliefs and cultural practices pertaining to child
abuse, perpetrators and victims can uncover the
social norms that may explain the occurrence of
violence and how people justify it. Solid evidence
of what works in terms of prevention and response
is also necessary in developing and implementing
successful strategies to address the problem. Finally,
records from the police department, child welfare
agencies, the criminal justice system, doctors
and hospital emergency rooms provide important
accounts of the outreach of entities in charge of
service delivery as well as basic information on the
circumstances related to incidents of abuse.

© UNICEF/IDSA2010-00027/Purnomo
Despite the wide-ranging types of data needed,

information on violence against children remains
sketchy in both coverage and scope. Several
obstacles have affected the availability and quality of
violence-related data, including methodological and
ethical challenges as well as the lack of international
14
standards for data collection, limited country capacity holistic picture. However, when this is done,
for data gathering and unsystematic investments in special attention must be given to reconciling any
the production of comprehensive statistics on the differences in the definitions, dates, coverage,
subject. comparability of the samples and quality of data
collected across datasets.4
Data sources
For instance, medical records and emergency
Data related to violence against children can be
room reports can be useful in providing diagnostic
found through multiple sources. These include
information pertaining to the nature of violence-
administrative records, surveys and qualitative
related injuries in children – including how, when
studies – all of which have strengths and
and where the injuries took place. However, medical
weaknesses.
records may not include information surrounding
the circumstances that led to the injury and may
Administrative records not be available for research purposes due to
Administrative records summarize information from confidentiality issues.5
a variety of official and non-governmental sources,
including public and private social and medical Police records may provide information on the
services, agencies of the criminal and civil justice relationship between the perpetrator and the
systems, research and documentation centres, and victim, whether weapons were involved and the
services such as shelters and helplines designed circumstances surrounding the event. Such records,
specifically to respond to victims of violence. however, naturally only address violence that is
reported to law enforcement. Previous research
Administrative data offer many advantages when has found that many victims of violence do not seek
conducting research on child abuse, including the help from the police. A household survey conducted
fact that they are routinely collected as part of in South Africa, for example, revealed that between
the day-to-day operations of involved agencies 50 and 80 per cent of victims received medical
and organizations. Using this information to its attention for a violence-induced injury without filing
full advantage is a cost-effective approach to data a police report.6 Similarly, a survey conducted in the
collection that does not rely on the availability United States found that 46 per cent of victims who
of sizeable research grants to sustain it. It also sought emergency medical care did not report the
streamlines the data collection process and avoids incident to the police.7 The potential applicability of
encumbering research participants with the need to police data is, therefore, limited due to high levels
answer sensitive questions on multiple occasions. of underreporting and the possibility of particularly
However, since these records are designed and vulnerable populations not being included in the
CHAPTER 2: USING DATA TO MAKE VIOLENCE VISIBLE

instituted for administrative purposes (and not analysis.8
primarily for the sake of research), they cannot be
used to calculate prevalence estimates and their Similarly, the use of criminal justice system data
potential usefulness is limited by numerous factors.2 and child welfare records also involves challenges
related to the composition of their samples.
Key concerns include data coverage (that is, whose Prison records, for instance, often contain detailed
data are and are not included within records), information pertaining to perpetrators but tend to be
the definitions used and their compatibility with biased towards more severe crimes (that is, those
international standards. Another common problem that led the perpetrator to be incarcerated). And
is inconsistency in the implementation of data criminal justice system data only capture instances
collection and quality assurance processes and of violence in which the alleged perpetrators were
procedures, along with missing data due to worker prosecuted. An additional challenge associated
error and/or negligence.3 Given these challenges, with such data is the process of identifying the
it may be necessary at times to combine multiple most appropriate unit of analysis. With justice
sources of administrative data to generate a more system data, this can vary based on the stage in
15
the judicial process at which the data are collected. collection of detailed data on the characteristics
For instance, units of analysis can include offenders, of both victims and perpetrators, the frequency
charges, suspected crimes, substantiated and circumstances surrounding the abuse and
crimes, victims, individuals charged with a crime, other pertinent information. A second approach
convictions and sentences, among others. While involves the inclusion of specific questions relating
each of these can provide useful information about to children’s experiences in stand-alone surveys on
violence against children, it is important to have violence that cover all ages of the population and
clarity about which stage the data refer to and avoid thus provide data that can be disaggregated by age.
double counting.9 A similar concern emerges when Finally, questions on violence against children can be
using data provided by child welfare agencies. included within generic household surveys that also
Different agencies employ child-based approaches inquire about a wide range of social, demographic
for monitoring cases or family-based approaches and health topics. When resources are limited, such
that track all instances of child abuse within one an approach can be the instrument of choice since
family. Comparisons of data across child welfare it may allow for the collection of data on violence on
agencies require that the data be disaggregated, amore regular and systematic basis.
which may not be possible when agencies adopt a
family-based method.10 Surveys offer numerous advantages when
conducting research on child abuse. For one, they
Vital statistics – which include government capture data about incidents that have not been
statistics on live births, deaths and foetal deaths reported to the police or other authorities. Moreover,
– are another common type of administrative data they can include detailed and targeted information
pertaining to violence against children. Birth and about the identity of the victims and/or perpetrators
death registries are maintained in many countries – such as socio-demographic characteristics,
worldwide alongside government records pertaining attitudes, behaviours and potential involvement
to homicides and suicides. Mortality data may be in violence or abuse in the past. Surveys can also
particularly useful since they can help estimate, for capture details relevant to specific incidents that
example, the numbers of deaths resulting from child may not be covered in administrative records.14 The
abuse and neglect, of homicides among children, quality of survey data is, however, highly dependent
of accidental injury deaths, of injury fatalities with on the rigour of the sampling method and the size of
undetermined intentionality and of sudden infant the sample. Data must be collected with a sufficient
death syndrome cases.11 Although mortality records sample size in order to allow for disaggregation
are widely available, calculating prevalence rates according to key variables (including age, sex and
from them can be problematic, since datasets are level of education).15 Random sample approaches
frequently incomplete. In addition, vital statistics to data collection are particularly advantageous if
may be a poor source of data on violence-related conducted properly, since they enable researchers
deaths among children due to the limited information to generate data that are nationally representative
available on many death certificates, strict coding and/or representative of the population being
guidelines that may require a documented history studied. Representative data are critical in making
of child abuse and/or reluctance among individuals generalizations about the broader population of
completing the death certificate to register child interest and thereby informing policy decisions.
abuse as the cause of death.12
There are also some other challenges to the
implementation of survey research. Surveys can
Surveys
be an expensive approach to data collection when
Collecting data on violence against children compared to administrative records and vital
through surveys can take three main forms.13 statistics, and they are more vulnerable to budget
The first approach involves the development and cuts and instability in times of limited financial
implementation of a stand-alone survey with a resources and political changes.16 In addition, they
specific focus on the issue. This allows for the are often retrospective in nature and therefore
16
limited by the victims’ ability to recall details about time to develop trust with participants22 and allow
past violence. Given the sensitive nature of the children the time to tell their story, in their own
issue, some respondents may not feel comfortable words, as they become more comfortable.23
disclosing experiences of violence to a person However, in order for children to benefit from the
they do not know, leading to concerns about social open-ended and flexible nature of the qualitative
desirability bias (that is, people reporting what interviewing process, interviewers need to have
they believe will be viewed in a favourable light skills, knowledge and supervisory support to build
by others). Even though standardized interview rapport with child participants and interact with
protocols are often used in survey research, the children in a sensitive manner. Interviewers must
quality of the data collected still depends on how also be able to recognize and properly respond to
the interview process unfolds – including the level verbal and non-verbal demonstrations of distress
of training and skill of the interviewers, the ways in among child research participants.
which survey questions are worded and asked, and
the location of the interview, among other factors.17 Despite the advantages, qualitative research can
Differences among findings in various surveys may also be time-consuming and labour-intensive and
arise as a result of the data collection method used. involve complex data analysis processes.24 In
For example, computer-assisted self-interviews, addition, since qualitative researchers are heavily
which allow research participants to privately record involved in interpreting their findings, this method
their answers, have been found to produce higher has been criticized for researcher bias.25 The
rates of self-reporting in victimization surveys than use of open-ended interview protocols can also
face-to-face interviews.18 Implementation of survey make comparisons across different qualitative
research requires a highly skilled team that is able studies challenging. Moreover, qualitative studies
to properly manage the statistical, operational and are often small in scale and tend to rely on non-
ethical challenges associated with the research random samples. As a result, they are not useful in
process. Nevertheless, when managed properly, measuring the prevalence of violence or generating
surveys are thought to provide more accurate results that can be generalized to a larger population.
estimates of prevalence within a population than
most other data sources. Surveys can also be highly
beneficial in developing nationally representative
“ States [should] improve data
datasets incorporating internationally validated collection and information systems
instruments and standardized definitions.19 in order to identify vulnerable
sub-groups, inform policy and
Qualitative studies programming at all levels, and
To generate a holistic picture of violence against track progress towards the goal of
children, it is often useful to combine quantitative preventing violence against children.

sources of data with qualitative information. States should use national indicators
Surveys are useful in assessing how many children
based on internationally agreed
experience violence, what forms are most common
and which children are most at risk. Qualitative
standards, and ensure that data are
approaches, on the other hand, can generate an compiled, analysed and disseminated
in-depth description and understanding of human to monitor progress over time. States
experiences, behaviour and social dynamics.20 As a should develop a national research
result, qualitative studies can address how violence
agenda on violence against children
occurs and why it persists.
across settings where violence
Qualitative research involves in-depth interviews occurs, including through interview
that are often open-ended, which makes it studies with children and parents,
particularly useful when asking children about with particular attention to vulnerable
highly sensitive topics such as their experiences of
groups of girls and boys.”
violence.21 When conducted appropriately,
qualitative interviewing approaches give researchers UN Study on Violence against Children
17
Methodological challenges perpetrators – even in anonymous surveys. Children
Collecting reliable data on violence against children may feel pressed to conceal incidents of abuse,
is a complex undertaking that raises serious particularly when perpetrated by people they know
methodological challenges. One key consideration and trust, or may be unwilling to report them for fear
is identifying the most appropriate and reliable of retaliation or stigma. Depending on their age and
informants. Widely perceived as a social taboo, stage of development, children may also be unable
violence is seldom reported by either victims or to provide accurate accounts of their experiences.
Measuring child neglect: Many

Box 2.1
unanswered questions
‘Child neglect’ is a broad term used Similarly, in 2011, the prevalence tion or inaction is truly intentional
to describe a failure to provide for of neglect among children under or rather attributable to their social
the physical and emotional needs age 17 in the United States was or economic status, including poor
of a child, to protect a child from estimated to be 15 per cent.30 How- education and lack of awareness of
danger and/or to obtain services for ever, even in high-income countries a child’s needs? How can we quan-
a child when the person(s) respon- where child neglect is common (as tify parental support and love to-
sible for her or his care have the substantiated by child protection wards a child? In the absence of a
means, knowledge and access to do services),31 it remains the most uncommon operational definition and
so.26 General Comment No. 13 on derstudied type of child maltreat- comparable measures, identifying
the Convention on the Rights of the ment. And, as with other forms of child neglect may be subjective and
Child elaborates further: Child ne- maltreatment, inconsistency in the contextual.32
glect includes: (1) physical neglect, measurement of neglect, differen-
such as failure to provide basic neces in the age and other characteris- With the exception of extreme
cessities including adequate food, tics of children involved in research cases, as when a child is severely
shelter and clothing; (2) psychologi- and the use of non-representative injured or dies, cultural norms re-
cal or emotional neglect, such as samples have resulted in studies of garding neglectful behaviour vary
lack of love, support and attention varying scope and quality that make in different societies. In some coun-
to the child along with exposure comparability of research findings tries, for example, leaving an infant
to domestic violence or substance extremely challenging. in the care of an 8-year-old sibling
abuse; (3) neglect of a child’s health, may be widely accepted; in other
including the withholding of essen- The scarcity of data on neglect is countries, this would be considered
tial medical care; (4) educational ne- due in part to problems associated neglectful of both children.33 Mur-
glect, or failure to comply with laws with measuring it, some of which ray Straus proposed a definition of
requiring school attendance; and (5) are similar to those confronted by neglect that acknowledges the rec-
child abandonment.27 data collection on other forms of vi- ognition of certain extreme forms
olence against children. Historically, as harmful to children across socie-
Data on child neglect are scarce, defining and measuring neglect ties, but also allows room for differ-
with existing statistics limited most- has been a subject of debate, parent cultural interpretations, describ-
ly to high-income countries.28 For ticularly in cross-cultural contexts, ing neglect as the “failure to act in
example, a 2011 population-based with many questions hanging in ways that are presumed by the cul-
survey in the United Kingdom the balance. For instance, what are ture of a society to be necessary to
showed that neglect is a commonly the minimum requirements associ- meet the developmental needs of a
reported form of child maltreatment ated with caring for a child? What child”.34 While this approach con-
in the family, with 16 per cent of constitutes failure to provide ‘ad- fers numerous advantages, it also
young adults aged 18 to 24 indicat- equate’ food, shelter, clothing and poses challenges when comparing
ing that they experienced neglect at protection? How can we measure data across countries or regions.
home at some point in childhood.29 whether caregivers’ neglectful ac- Moreover, various cultural groups
18
Adults often fail to report incidents as well, even Other methodological challenges include the fact
when they are not directly responsible. When that certain types of violence, such as neglect, can
interviewed about their experiences as children, be particularly difficult to quantify due to their largely
adults may likewise be unwilling or unable to report intangible nature (see Box 2.1), while others, such
what happened to them. Caregivers of victimized as certain forms of sexual abuse, are particularly
children may be reluctant to disclose information in hard to measure due to the conditions of extreme
an attempt to protect them from further harm. secrecy and illegality that surround their occurrence.
within the same country may have result of circumstances outside the clinician-rated scales assessing car-
different definitions of neglectful direct control of the caregiver(s). In egivers’ ability to provide for chil-
behaviour, raising the question of contrast to other forms of maltreat- dren’s basic needs. Implementation
where the line is drawn when devel- ment, child neglect is committed of the CWBS involves an independ-
oping measures incorporating dif- by omission, making it difficult to ent rating of household issues (such
ferent cultural norms. Timing must observe in practice. In high-income as overcrowding, sanitation, safety
also be carefully considered. What countries, the most common sourc- and hygiene) and relational issues
constitutes neglect varies accord- es of data on child neglect are usual- between caregivers and their chil-
ing to the age and developmental ly administrative records from child dren (including acceptance, expec-
status of the child.35 For instance, protection services. The usefulness tations, discipline and stimulation,
in some countries, leaving a child of this data is, however, limited by among others).41 Scales relying on
unattended for an hour is deemed biases in the reporting and screen- parent-child observation, such as
neglect when the child is extremely ing process.38 The vast majority of the CWBS, have been widely used
young, but not when she or he is an neglect cases are never reported to by child welfare and social ser-
adolescent. Additionally, societal authorities, and caseworkers may
vice agencies in Canada and the
standards for risk and harm are con- not be consistent in their assess-
United States as a method for as-
stantly evolving over time, resulting ment of what constitutes neglect.39
sessing neglect during home visits.
in ongoing changes in legal defini-
In spite of these many obstacles, These approaches have, however,
tions and interpretations of what
some scales measuring neglect been criticized on methodological
constitutes neglect.36
do exist. In 2004, for instance, re- grounds for issues such as the lack
Extensive debate has also surround- searchers developed a child self-re- of standardized procedures for ad-
ed the issue of whether to adopt a port measure of neglect, the Multi- ministration, the absence of clinical
caregiver focus and/or a child focus dimensional Neglectful Behavior cut-off points and the inability to ad-
in defining neglect – that is, should Scale (MNBS), which includes four equately capture levels of extreme
the emphasis be on caregiver omis- primary domains of neglectful poverty, among others. Addition-

sions in care, on children’s needs behaviour (emotional, cognitive, ally, scales such as the CWBS were
not being met or on some combi- supervision and physical). Items only intended for families involved
nation of the two? In developing range from “My parents left me with child protection services, thus
measures of neglect that centre on somewhere and did not come back” limiting their applicability.42
caregiver actions and inactions, to “My parents do not tell me that
measurement is complicated by the they love me.”40 While the short Future efforts to measure child ne-
reality that caregivers may face conform of the MNBS has been vali- glect should continue to focus on
straints out of their control, such as dated in several cultural contexts, improving clarity and consistency.
insufficient income, limited access research has largely been limited At the same time, implementation
to medical care, unsafe housing and to high-income countries. Other ne- of further research on the subject
violence in their community.37 Even glect scales rely instead on observa- outside high-income countries is
when child-based measures are tion of parent-child interactions and warranted to deepen the under-
used, it is challenging to distinguish of the child’s home environment. standing of child care and neglect
whether the failure to meet a child’s For instance, the Child Well-Being across a variety of social, economic
need is voluntary rather than the Scales (CWBS) consist of a series of and cultural contexts.
19
Furthermore, some forms of violence that are necessary for statistical analysis, particularly on
socially condoned may not be perceived as variables that are not central to the implementation
deserving of public attention and intervention, of relevant administrative processes.44 In cases of
and therefore may not be considered a priority for incomplete data, information from multiple sources
data collection. Nationally representative datasets can theoretically be merged together. However, in
pertaining to violence against children are not practice, this may not be possible due to lack of
usually available due to the high costs associated comparable research measures or methods.45
with conducting large-scale surveys. While
administrative data can be used in the absence of The comparability of data across multiple sources
surveys to provide information on certain aspects and/or countries poses a formidable challenge
of the problem, the sample covered – as explained for data analysis pertaining to violence against
earlier – may not properly represent the population children. To facilitate comparisons across countries,
of interest and will only capture violence that has regions and globally, data collection processes and
been reported to the authorities.43 In addition, the measurement systems must share a standardized
data collected for administrative purposes may approach.46 However, measurement of violence
be of insufficient quality and/or level of detail against children often varies widely across datasets,
The potential impact of children’s

Box 2.2
participation in research on violence

A fundamental concern in research research may generate discomfort, towards (30 per cent) or disagreed (14
on violence against children is the fear or even re-traumatization in the per cent) with the statement that other
potential impact on children of form of memories or flashbacks. For surveys should ask similar questions
their participation in research. Key children who have not been victimized, about violence.
questions include: the concern is related to the potential
harm of being exposed unnecessarily Another US study showed similar
• Is it justifiable to include children in to distressing or uncomfortable issues results. This gathered information
abuse research classified as non- that are beyond their experience or from young people about the degree
therapeutic that has limited, indirect knowledge.54 These feelings may be to which they became upset while
or minor benefits for children? exacerbated in situations where the completing a self-reported survey
• What are the possible short- and protection of the research participants about sensitive events. The study
long-term consequences? cannot be guaranteed or where found that 30 per cent reported some
interviewers without appropriate level of upset, with victims of sexual
While there is general agreement on skills and training engage in the data violence most affected.56
the importance of obtaining reliable collection.
statistics on violence, the impact of Talking about traumatic experiences
data collection on individual children For example, a national survey may bring relief to some children who
remains largely unknown because conducted in the United States in are involved in therapeutic treatment.
minimal research has been done on 2009 on different forms of violence Even in a non-therapeutic research
the subject.52 What the research does (including sexual abuse) found that context, findings from a study with
suggest, however, is that children and nearly one quarter (23 per cent) of the Bosnian refugee children confirm that
youth are more likely to be emotionally participants reported becoming upset many children rated participation in
affected by survey questions than by the questions asked, with girls and interviews as a positive experience,
adults.53 younger participants most affected.55 indicating that they felt “relief” after
The survey involved 1,588 participants the interview. Children in this study,
In the case of children who have aged 10 to 15 years. An additional 44 however, were interviewed about
experienced violence, involvement in per cent of participants were neutral their experiences as refugees and not
20
studies, countries and regions.47 The types of violent national comparison, developing measures that
acts reported are often inconsistent, resulting in adequately capture nuances in different cultural
underestimates of some forms of child abuse.48 contexts and for different subpopulations is also a
Moreover, differences among national, regional key concern.50
and even provincial or territorial definitions and/
or theoretical understanding of violence against Ethical considerations
children can make the aggregation of statistics
Numerous ethical issues confront researchers when
impossible. Another problem is that reports
collecting data on violence against children and
addressing violence against children often do
not provide a detailed enough explanation of the there are potential safety risks for all involved. For
research methods to enable an outside reviewer to example, surveys have the potential to re-traumatize
properly assess issues in the data collection process research participants if questions are not asked in
and their possible effect on the measurement of key a sufficiently sensitive manner (see Box 2.2).51 It
constructs.49 While standardization of definitions, is therefore crucial that fundamental principles are
indicators and questionnaires is critical for cross- followed for the ethical collection of sound data.
directly questioned about previous another consideration in ethical as age, sex and other characteristics
history of interpersonal violence, decision-making. Parents usually on children’s willingness and ability
thereby limiting the potential play a key role in supporting and to participate in such research; the
applicability of their experiences to protecting their children, although relationship between the provision of
interviews on that topic.57 There is this is negated when parents are information to parents and children
also the risk that researchers who are the perpetrators of violence against and participation rates; and how
not trained may, in fact, exacerbate a them. Ethical guidelines are, at times, parents weigh up risks and benefits
child’s distress. Some studies indicate strongly influenced by the possibility of children’s participation in such
that it is the negative responses or assumption that the parent may be research. The review identified a
respondents receive to their disclosure the perpetrator, which has a negative significant gap in the ethical guidelines
that causes harm, rather than the impact on the potentially protective on what to do if no services are
disclosure itself, and that any benefits role parents can play. This affects available for the referral of children
of talking about the trauma can be ethical decision-making about the to psycho-social support or child
marred by those harmful effects.58 consent process, the information protection. Some documentation
provided and the strategies planned advises strongly against proceeding
The potential impact on children of
for child protection. Family violence is, with the research if no referral services
participation in such research needs

to be considered in light of their however, only one aspect of violence are available, while other experts
history of victimization. While the against children. Parental support can advise finding alternatives such as
literature tends to focus on children be a welcome consideration when services that have been developed in
who have experienced abuse, which is other aspects are researched. response to violence against women.
critically important for some research
A recent review of the available The findings of the literature review
activities, researchers are ethically
literature identified gaps and areas point to the need for further research,
responsible for all children. Ethical
where further research is needed.59 particularly in the case of sexual
considerations must therefore take
These include: the emotional impact violence and especially when data
into account the different experiences
children have had and consequent on all children of participating in collection is undertaken as part of
areas of sensitivity and concern – both research on violence; the nature, large-scale efforts, as opposed to
for those children who have been duration and magnitude of distress small-scale research studies carried
victimized and those who have not. caused by such participation, among out with adequately trained personnel.
both children who have a history of Given all these areas of ethical
The relationship between the child violent experiences and those who do uncertainty, caution is warranted in
and the perpetrator of violence is not; the role of different factors such undertaking data collection.
21
The core responsibility of a researcher working with must also be established to guide researchers in
children is to protect them from potential harm and deciding whether and when to intervene when a
ensure that they benefit from the results of the study. child participant is found to be in possible danger.64
Informed consent, for instance, must be granted In some contexts, sufficient services may not be
before carrying out any interviews. This involves an available for children who experience abuse.65 In
explicit agreement to participate that is predicated these cases, researchers must decide whether
on interviewees fully understanding the research or not to proceed with their work and, if so, how
process, not being coerced into participation, and to avoid inadvertently causing harm to research
having the freedom to withdraw from the study participants.
at any time.60 In research with children, a central
question is the extent to which they understand A moral imperative and
the nature of the process and the possible strategic investment
repercussions of their engagement in it. With
younger children particularly, there is the concern While collecting and analysing data on violence
that even though they may know conceptually that against children is full of potential hazards, such
they can withdraw from the study at any time, they data are nonetheless essential to document
may not be sure how to do so in practice and/or may the widespread nature of violence, to support
be hesitant to do so while viewing the researcher government planning and budgeting for universal
as an authority figure.61 Another concern is whether and effective child protection services, and to
and when parental consent is necessary and/or in inform the development of effective laws, policies
the child’s best interest. In research pertaining to and prevention programmes worldwide.66
violence against children, parental consent may in
fact put children in danger of further harm from Recent years have witnessed a surge in the number
parents and/or other caregivers. It may not be of activities aimed at filling existing data gaps,
possible to obtain parental consent for some groups primarily through population-based sample surveys.
of children for logistical reasons or because they International survey programmes are playing a
lack an active guardian (as is the case with children central role in strengthening the knowledge base
living on the street). Furthermore, legal guidelines through the collection of comparable data on
on and cultural attitudes towards parental consent specific forms of violence against children (see Box
vary substantially among countries and must be 2.3). The number of targeted surveys dedicated
carefully considered when conducting cross- exclusively to the collection of comprehensive
national research.62 information on children’s experience of violence has
also grown in recent years. Initiatives have been
Yet another ethical consideration is the need to undertaken by or with the support of international
maintain confidentiality, which involves careful agencies, international and local non-governmental
steps to ensure that data are collected while drawing organizations (NGOs), government institutions
minimal attention to subjects, thus safeguarding and researchers. The fact that governments and
their privacy. Data must also be stored in a secure others have expressed interest in advancing this
manner and research findings must be presented area and have invested in improved data collection
in a way that safeguards anonymity.63 When such efforts is a positive step forward. However,
protective measures cannot be adopted, collecting these activities have largely been carried out in
data on violence from young respondents may not isolation, and some of them remain unknown to
be desirable and should be avoided, since children the broader child protection community. Moreover,
may be exposed to the risk of retaliation from the use of different definitions and methodologies
perpetrators. At the same time, guidelines must has produced data of varying quality and made
be instituted to ensure that referral and support comparisons across countries problematic (see Box
systems are in place for children who report being 2.4). In addition, different standards and practices
at risk during the research process. Protocols have been followed for ensuring the protection of
22
respondents and interviewers and providing follow- to sex, age, residence, education, ethnicity and
up support to victims, which has raised important other household and family characteristics. New
questions as to the potential risks stemming from approaches to data-gathering and information-
the involvement of children in the data collection sharing, including ‘big data’, represent new methods
process.67 worth further exploration in the monitoring and
reporting of violence. Finally, broad dissemination
is needed of available data in formats that are
Moving forward, a solid framework is needed to
easily understood by a wide audience to raise
guide researchers on sound approaches to data
awareness of the pervasive nature of violence and
collection and on ethical practices – one that is
its detrimental consequences. All this points to the
sensitive to the multiple contexts in which research need to develop, adopt and implement international
is conducted.68 It is also recommended that States standards for the collection, analysis and
continue to build data collection mechanisms, dissemination of violence data as key steps
use indicators and tools that are rigorously tested towards improving the availability, quality and use of
and validated, and disaggregate data according resulting statistics.
23
International survey programmes with
Box 2.3
questions about violence against children

The comparability of statistics on vio- survey tools. The questionnaires are all and middle-income countries at regu-
lence against children is a key factor modular in nature and can be adapt- lar four- to five-year intervals with the
in accurately describing this phenom- ed or customized to the needs of the support of the US Agency for Interna-
enon across time and countries. The country. tional Development (USAID).71
availability of comparable data on
certain forms of violence, including The third round of MICS included for Several countries added the MICS
violent discipline and bullying, has the first time an optional module on child discipline module, or a modified
significantly increased in the last two child discipline adapted from the Par- version thereof, in DHS conducted
decades, mainly through the inclusion ent-Child version of the Conflict Tac- after 2005. The standard module was
of violence-related questions in inter- tics Scale (CTSPC), developed by so- used in the Albania DHS 2008-2009,
national survey programmes. ciologist Murray Straus in the 1970s.69 Armenia DHS 2010, Azerbaijan DHS
The MICS module inquires about the 2006, Jordan DHS 2012, Liberia DHS
use of eight violent (six physical and 2007 and Niger DHS 2012. The Egypt
Multiple Indicator Cluster
two psychological) and three non- DHS 2005 implemented a modified
Surveys (MICS)
violent disciplinary practices used at version of the module that included
UNICEF assists countries in collecting home. Some countries have custom- questions on only three violent and
and analysing data on the situation ized the module to include additional one non-violent disciplinary practice
of women and children through the forms of punishment such as isolating for children between the ages of 3
MICS programme. Since its incep- a child, withholding a meal or burning and 17. The module on child discipline
tion in the mid-1990s, this interna- a child with fire or a hot instrument, applied in the Congo DHS 2011-2012
tional household survey programme among others. The last item in the included two additional types of pun-
has enabled more than 100 low- and module probes the personal beliefs ishment (pulling a child’s ears and
middle-income countries to collect of the respondent about the neces- withholding a meal) while the Haiti
nationally representative and inter- sity of using physical punishment to DHS 2012 included three additional
nationally comparable data on more raise/educate children. In MICS3, the forms of punishment (pulling a child’s
than 100 key indicators in areas such mother or primary caregiver of one ears, withholding a meal and making
as nutrition, child health, mortality, randomly selected child was asked a child kneel).
education, water and sanitation, child whether any of the discipline meth-
protection and HIV and AIDS. To date, ods covered in the module had been A set of questions on child discipline
four rounds of MICS have been com- used by any member of the household (not the MICS module) was also in-
pleted (MICS1: 1995-1996, MICS2: in the past month. Beginning with cluded in the DHS conducted in the
2000-2001, MICS3: 2005-2006 and MICS4 (and all subsequent rounds), Plurinational State of Bolivia in 2003
MICS4: 2009-2012). The fifth round the methodology was changed so that and 2008. Men aged 15 to 64 years
of MICS (MICS5) is currently under any adult household member, not just and women aged 15 to 49 years were
way and is expected to be completed the mother or primary caregiver, can asked about their own behaviours
by 2015. act as the respondent for the child with regards to discipline in the home
discipline module. As of May 2014, and their agreement or disagreement
The MICS survey tools are developed data on child discipline had been col- with a number of justifications for
by UNICEF in consultation with rel- lected in 47 countries.70 Details on all hitting children. The Colombia DHS
evant experts from various UN or- the rounds of MICS can be found at in 2005 and 2010 included a similar
ganizations and interagency monitor- data.unicef.org. set of questions on child discipline,
ing groups. The core tools include a but these were posed only to women
household questionnaire, a question- aged 15 to 49 years residing in the
Demographic and Health
naire for individual girls and women household.
Surveys (DHS)
between the ages of 15 and 49, and
a questionnaire on children under The DHS collect nationally representa- In addition to collecting data on child
age 5 (administered to mothers or tive data on topics including popula- discipline, the DHS programme in-
primary caregivers). Beginning in tion, health, HIV and AIDS, nutrition cludes a standard module on violence
MICS4, an individual men’s question- and women’s status and empower- based on a modified version of the
naire has also been added to the core ment. Surveys are carried out in low- Conflict Tactics Scale. The first time
24
such data were collected as part of a Global School-based Student Health Health Behaviour in School-aged
DHS was in Colombia in 1990. In 1995, Surveys (GSHS) Children Study (HBSC)
questions were fielded in Egypt and
The GSHS are a collaborative sur- The development of the HBSC dates
again in Colombia. However, it is only
veillance project of the World Health back to 1982, when a group of re-
in 1998-1999 that the DHS programme
Organization (WHO) and the United searchers in England, Finland and Nor-
developed a standardized approach
States Centers for Disease Control way agreed to create and implement
to the measurement of violence and
and Prevention (CDC) to help coun- a common research methodology for
first implemented it as a part of the
tries measure and assess behavioural surveying school-aged children. The
1998 DHS in Nicaragua. The module
earliest HBSC survey was conducted
is addressed to girls and women aged risk and protective factors in 10 key
15 to 49 years and includes ques- in 1983-1984, when it was adopted by
areas among adolescents. The GSHS
tions on the experience of specific the WHO Regional Office for Europe
questionnaires are self-administered
acts of domestic and other forms of as a collaborative study and then re-
and are composed of 10 core mod-
interpersonal violence. In particular, peated seven more times (every four
ules, core expanded questions and
information is collected on any form years) until 2009-2010. Thus, trend
country-specific questions. One of the analysis is possible for countries with
of physical violence committed by
10 core modules is on violence and successive surveys.
anyone that has been experienced by
unintentional injury and contains two
girls and women since age 15, sexual
violence at any age (including wheth- questions about physical violence (ex- The HBSC study collects data on the
er first sexual intercourse was forced) perience of being physically attacked health behaviours and social envi-
and help-seeking behaviours, such as and involvement in physical fights in ronments of girls and boys ages 11,
if and from whom help was sought. the last year) and two about bullying 13 and 15 through self-administered
Ever-married girls and women are (frequency and type of bullying expe- questionnaires completed in the
asked about controlling behaviours rienced in the past 30 days). The GSHS classroom. Topics include, for exam-
of spouses or partners; experiences core expanded questionnaire also in- ple, body image, life satisfaction, oral
of emotional, physical or sexual vio- cludes questions on dating violence, health, relationships with family and
lence committed by their current or peers, sexual behaviour, substance
sexual violence, carrying of weapons,
most recent partner; frequency of use and physical activity. Questions
perception of safety at school, and
abuse; physical consequences of the on the experience of being bullied and
physical violence by teachers.
violence; and when the violence first bullying others have been included
began in the relationship. Information The GSHS are implemented upon re- since the first survey; information on
is also collected on women perpe- injuries and fighting has been col-
quest from countries. The first set of
trating spousal violence. In addition, lected since the 1993-1994 round. The
surveys was conducted in 2003 and
girls and women who have ever been standardized questionnaire enables
the latest surveys were completed in
pregnant are asked whether they ex- cross-national comparisons to be
2012; they have been implemented,
perienced any physical abuse during made across participating countries.
pregnancy and their relationship to or are currently under way, in 109
These findings have been summa-
the perpetrator. As of July 2014, data countries.72 Of these, 72 countries rized in five international reports (for
on violence against girls and women have collected information on all, or

the years 1993-1994, 1997-1998, 2001-
have been collected through DHS in some, of the questions pertaining to 2002, 2005-2006 and 2009-2010).
about 43 countries. Trend analysis is physical violence and bullying. For
possible for a number of countries some of these countries these data The HSBC has a regional focus on
that have collected these data more are available for more than one point Europe and North America and has
than once, including, for example, the in time, including Argentina, Chile, been implemented in 43 countries.73
Plurinational State of Bolivia, Cambo- Egypt, Ghana, Guyana, Jordan, Leba- Further details about the study can
dia, Cameroon, Colombia, Dominican be found on the HBSC website at
non, Mauritius, Morocco, Oman, the
Republic, Haiti, Kenya, Malawi, Peru, www.hbsc.org.
Philippines, Swaziland, Trinidad and
Rwanda, Uganda, Zambia and Zimba-
Tobago, United Arab Emirates, Uru-
bwe. A handful of countries have also
guay and the Bolivarian Republic of
included a version of the module to
collect information on the experienc- Venezuela. Further information about Excerpted from: Child Protection Monitor-
the GSHS can be found on the WHO ing and Evaluation Reference Group, Meas-
es of violence among boys and men.
uring Violence against Children: Inventory
Further information about the DHS website at www.who.int/chp/gshs/en/
and assessment of quantitative studies, Di-
can be found on the DHS website at and the CDC website at www.cdc.gov/ vision of Data, Research and Policy, UNICEF,
www.measuredhs.com. GSHS/. New York, 2014.
25
Recent data collection activities
Box 2.4
on violence against children

In 2010, Save the Children and or for which sufficient background training and use of a pilot test
UNICEF, together with other partners, information could not be found
established the global interagency through further research. A compara- • Quality control procedures, including
Child Protection Monitoring and Eval- tive analysis of the surveys was car- the use of call-back procedures and
uation Reference Group, known as the ried out across a selection of variables quality control checks in the field
CP MERG. A Technical Working Group within the following six dimensions:
• Ethical protocols, including in-
on Data Collection on Violence against
• Basic characteristics of the study, formed-consent procedures and
Children was created under the CP
including year and country of imple- follow-up procedures for abuse dis-
MERG to guide, support, facilitate and
mentation, coverage and commis- closures.
coordinate the establishment of guide-
sioning/implementing agencies
lines for data collection. As a first step, The review confirmed a growing in-
the Working Group identified the need terest in related research since the
• Definitions, indicators and content
to review recent quantitative studies.74 2006 publication of the UN Study on
of the questionnaires, including the
The review was conducted to iden- Violence against Children. Among the
definition of violence against chil-
tify major areas of progress, gaps and dren; target groups and gender fo- 38 studies, 28 were conducted in or af-
challenges in the development of data cus; types of violence, abuse or mal- ter 2006. Before 2006, 10 studies were
collection tools and methodologies. treatment researched; behaviours carried out (not considering the earlier
assessed; and research tool/instru- waves for repeated studies), with the
In total, 30 country studies, four multi- earliest individual country study in
ment used
country studies and four general sur- 2002-2003.
veys that included components on • Methodology and sample design, in-
violence against children were identi- cluding possible sources of errors or The majority of studies were commis-
fied. In addition, leads were found to bias and the use of sampling weights sioned by individual organizations or
25 other studies that were either not agencies, namely international organ-
published at the time of the review, or • Field implementation, including se- izations (most commonly UNICEF),
whose findings had been embargoed lection and profiles of field staff, international NGOs and government
26
© UNICEF/NYHQ2012-2288/Markisz
agencies. Several were commissioned Corporal punishment was the focus of ally, the assessment found that, in
under a partnership that usually in- fewer but still a significant number of many cases, information on basic pa-
cluded a national government and in- studies, possibly because of the defi- rameters (such as sampling methods,
ternational organization. The types of nitional difficulties in distinguishing it target groups, sample sizes and sam-
agencies and organizations responsi- from physical abuse or maltreatment. pling errors) was either not available
ble for implementation varied widely Bullying and neglect were less com- in the research reports or was buried
across studies, from national and local monly researched. However, while somewhere within them, which limits
NGOs to private sector agencies, aca- most surveys asked whether the child a reader’s ability to properly under-
demic institutions, research centres was subjected to different forms of stand the findings.
and individual consultants. The vast violence, the behaviours measured
majority of studies identified in the (that is, the ways in which violent Finally, most studies included no or
review (33) were conducted just once. acts were operationalized) in different very little information on the ethical
questionnaires were rarely the same. protocols followed to protect those
Overall, the scope and quality of the involved in their implementation.
studies identified and reviewed var- The assessment also found that most Many were found to have violated ba-
ied to a high degree. Most studies studies did not make it clear whether sic principles of research ethics: Only
used their own customized definitions extensive validity testing had been 25 study reports actually mentioned

of violence. Key terms (such as ‘vio- employed; such testing would deter- the use of informed consent. Among
lence’ and ‘abuse’) were often defined mine which approaches and sets of those that did, the reports did not al-
on an ad hoc basis that was unique questions yield the most reliable data ways detail how informed consent
to each study. This made comparison for measuring different dimensions was obtained, and only very few made
across studies difficult, despite their of violence. Only a few studies used the consent forms and procedures
use of similar labels, such as ‘physical sound research methodologies and available in the study materials.
abuse’. The tools used for data collec- approaches to, for example, sample
tion varied extensively also in terms of and questionnaire design or data en-
scope, target population and design. try and analysis. Indeed, some studies
All but four of the studies explored were found to violate basic quantita-
multiple types of violence. Most stud- tive research principles, including the Adapted from: Child Protection Monitoring and
ies were interested in the physical use of purposive samples in studies Evaluation Reference Group, Measuring Vio-
lence against Children: Inventory and assess-
dimension, followed closely by the aimed at obtaining representative ment of quantitative studies, Division of Data,
sexual and emotional dimensions. data at the population level. Addition- Research and Policy, UNICEF, New York, 2014.
27
a note on the data
Defining children Currency of data
Throughout this report, the term ‘children’ is used to The availability of data for different indicators of violence
refer to anyone under the age of 18, in line with the against children is highly uneven across countries. In
definition of childhood provided in the Convention on some cases, the latest source of comparable data dates
the Rights of the Child. Where some data represent back to 2000; in others, it is as recent as 2013. Data
respondents up to 19 years of age or older, this is noted indicate the status of the particular indicator at the time
in the text and different terms, such as adolescents or of collection and do not necessarily reflect the current
youth, are used. situation.
The analyses presented in the following chapters are
Definitions of violence based on data from the most recently available source
identified for each country. Year ranges provided in the
Given the limited availability of comparable data
sources for figures, tables and maps denote the period
pertaining to violence against children, multiple data in which data collection took place. For each country,
sources were combined to produce as comprehensive data refer to the most recent year available during the
a picture as possible. However, whenever multiple data specified range.
sources are used, concerns arise as to the comparability
of findings across sources. For instance, definitions of
partner or spousal violence vary considerably among
Data availability
data sources. Some surveys incorporate questions The ability to include data on specific indicators for any
pertaining to physical, sexual and emotional violence, given country is based on the availability of published
whereas others include only one or two of these forms. survey reports as well as access to the datasets. In some
Other surveys capture additional types of partner instances, only a selection of results were published in
violence, such as economic violence. The age groups survey reports and datasets were not available. This
of children represented by different surveys also tend to made it impossible to undertake additional analyses or
vary, further complicating comparisons across sources. to recalculate the estimates to make them consistent
While every effort has been made to explain any with findings from other countries.
discrepancies in the measurements and definitions used For some indicators, data were also collected in other
across sources, data should be interpreted cautiously in countries that are not presented in figures, tables or
light of these limitations. maps since results are based on less than 25 unweighted
cases.
While different labels and definitions are found in
the literature to describe the range of victimization
experiences suffered by children, the terms ‘abuse,’ Precision of the estimates
‘violence’ and ‘maltreatment’ are used interchangeably A sampling error is usually measured in terms of the
throughout this report for easier reading. When these standard error for a particular statistic. The standard
umbrella terms are used to describe available data, it error can be used to calculate confidence intervals within
is important to remember that specific definitions may which the true value of the estimate can reasonably be
have been used when collecting the data. A correct assumed to fall. This means that the value of any given
understanding of the precise way in which violence has statistic calculated from a sample survey will fall within
been measured is crucial when interpreting the data. a range. Because some of the indicators presented in
The specific acts covered by the data are indicated in the the following chapters represent only a small number of
text, endnotes and/or notes on the figures and tables. respondents, particularly in the case of sexual violence,
28
the confidence intervals are likely to be relatively wide,

affecting the accuracy of the estimates produced. The
size of the confidence interval also affects the ability
to make meaningful comparisons between groups
of respondents, such as girls and boys. Confidence
intervals are not systematically displayed in figures
for readability. Caution is therefore warranted when
interpreting the findings.
Accuracy of data
A key element of data quality is accuracy – that is, how
closely the data represent the reality of a particular
situation. While accuracy is challenging in any statistical
field, it is particularly so when producing statistics on
topics such as violence against children due to increased
risks for underreporting and other possible sources of bias
(see Chapter 2). For example, it is important to remember
that the data presented on the following pages reflect
the willingness and capacity of respondents to disclose
personal experiences. Furthermore, since disclosure
is also dependent on questionnaire design and on the
ability of interviewers to draw out sensitive information
from respondents, the data are also influenced by the
quality of the data collection tools and process.
Ethical concerns
As explained in detail in Chapter 2, the collection of data
on violence against children is a complex and sensitive
undertaking that poses a number of ethical constraints.
While the collection of data on violence requires
compliance with strict ethical protocols regardless of
the age of those involved, the direct participation of
minors as respondents compounds these challenges
and requires even more careful adaptation and
implementation of procedures and requirements.
Some survey reports included only limited information
on the field procedures used for data collection and
on the ethical protocols employed to obtain consent
from survey participants and to guarantee the privacy,
© UNICEF/SRLA2011-0228/Asselin
confidentiality and safety of all concerned. Moreover,

the actual procedures used to collect data in the field
may diverge from the protocols described in survey
reports. Due to limited information in survey reports, it
was unfortunately not possible to systematically assess
whether data included in this report have been collected
while adhering to core ethical principles.
29
Chapter 3 Physical
violence:
A leading
cause of
injury
and death
among
children
According to General Comment No. 13 on the
Convention on the Rights of the Child, physical
violence takes both fatal and non-fatal forms and
encompasses physical torture, cruel and inhuman
physical punishment, physical bullying and hazing,
and corporal punishment.1 It involves someone
using an object or part of his or her body to physically
harm a child and/or control a child’s actions.
Examples include slapping, spanking, pushing,
punching, kicking, choking, burning, assaulting a
child with a weapon or object, and murder. Physical
violence against children can be committed by
anyone, although adults in positions of trust and
authority (such as caregivers and teachers) are
often cited as perpetrators. Acts of physical force

against children are also common among peers.
Although all children are vulnerable to physical

violence, different factors affect their relative
risk of exposure as well as the consequences.
Young children, for example, are generally most
vulnerable to serious injury from physical violence,
30
with fatal cases often found among infants.2 In suicidal behaviour and self-harm, among others.8
some contexts, girls are at heightened risk for Physical violence has also been shown to impair
infanticide.3 And children with disabilities may be brain development, leading to long-term cognitive,
particularly vulnerable to specific forms of physical language and academic challenges.9 Children who
violence such as forced sterilization and violence in experience physical violence are at risk of truancy and
the form of ‘treatment’ (including electroconvulsive repeating grades and tend to have lower educational
therapy).4 Several studies have found that, in many aspirations and achievement than children who
countries, boys appear to be at greater risk than have not been treated in this way.10 There may also
girls of physical punishment by caregivers and be long-term economic consequences, including
may be subjected to harsher forms.5 While young increased rates of unemployment in adulthood and
children are often at greater risk than older children a greater likelihood of living below the poverty line.11
of physical violence by caregivers, adolescents Physical violence has social ramifications as well,
are prone to such violence by peers, including which can include aggression, social withdrawal
dating partners. Older children are also more likely and difficulty relating to others.12 Although not
to experience violent physical interactions with all children who experience such violence will
strangers, often as a result of crime and gang exhibit future behavioural problems, the possible
violence. In too many instances, such violence consequences include not only drug and alcohol
leads to premature death. abuse, delinquency and risk-taking behaviours13
but also an increased risk of perpetrating violence
Physical abuse is likely to be accompanied by other against others, including physical fights with peers,
forms of violence (see ‘When different forms of dating violence and bullying.14
violence overlap’ on p. 84). For example, a national
survey in the United States found that children who Data sources
were physically assaulted at some point in their
Mortality figures can provide an indication of the
lives were at least six times more likely than other
extent of lethal violence in a particular community
children to have also experienced sexual violence.6
or country. When compared to statistics on other
Another study found that over 60 per cent of
causes of deaths, such data are important in
children who were victims of physical violence at
assessing the relative burden caused by violence.15
home also witnessed the physical abuse of one
family member by another.7 Many children who
Estimates on homicides presented in this chapter are
experience physical violence, therefore, not only
derived primarily from the Global Health Estimates
have to cope with direct physical harm but, in many
produced by the World Health Organization (WHO).16
cases, must also deal with other forms of violence
In a few instances, data from administrative records
committed against them and those they love.
compiled by the United Nations Office on Drugs
and Crime (UNODC) have been used,17 mainly to
Wide-ranging consequences provide additional information on circumstances
Extensive research on the experience of physical surrounding fatalities. Estimates from both sources
violence during childhood provides evidence of need to be interpreted with caution.
the potentially damaging effects it can have on
CHAPTER 3: PHYSICAL VIOLENCE
the physical, psychological and social well-being of The Global Health Estimates figures, for example,
children. It can lead to consequences ranging from are obtained through a standardized statistical
minor bruises and broken bones to head trauma, model that produces cross-country comparable
physical disability and even death. Direct and data on all causes of death, including figures on
indirect exposure to repeated or severe episodes fatalities due to interpersonal violence. The model
of physical violence in childhood are associated is based on a series of underlying assumptions
with an extensive range of mental health problems, and inferences and, as result, estimates are prone
including depression, anxiety, post-traumatic to possible measurement errors. In the absence
stress disorder, dissociation, psychiatric disorders, of alternative figures obtained from actual counts,
31
modelled estimates are used to get an indication of and since then have been updated three times (in
the numbers of deaths due to violence. 2000-2002, 2004 and, most recently, 2012). For any
given disease or injury, DALYs are calculated as the
Registration systems that are operating effectively sum of years of life lost (YLL) due to premature
compile vital statistics on the occurrence of mortality in the population and the years lost due to
births and deaths during a given period. These disability (YLD) for incident cases of the disease or
data are then combined with figures obtained injury. One DALY can be thought of as one lost year
through medical and police records resulting of ‘healthy’ life due either to premature death or to
from the certification of causes of individual being in a state of poor health or disability.20 DALYs
deaths and the investigation of criminal cases. are calculated for all WHO Member States for over
However, in many countries, administrative data 100 different diseases and injuries.
pertaining to intentional injuries and deaths are not
systematically collected, may not be accessible or Updated data for 2012 were recalculated for the
may not be adequately compiled across sources. following major causes of disease and injury:
Calculating reliable figures from these basic counts communicable diseases, maternal, neonatal
is often not possible due to weaknesses in data and nutritional conditions; non-communicable
collection systems, such as incomplete coverage or diseases; unintentional injuries (including road
misrepresentation of the events. The identification injuries, poisonings, falls, exposure to fire/heat/
of causes of death is a complex undertaking even in hot substances, exposure to forces of nature
countries with advanced health systems. And, even and other unintentional injuries); homicides; and
when injuries are identified as the cause of death, other intentional injuries (including self-harm and
the determination of whether the death resulted collective violence and legal intervention21).
from an accidental or wilful act can be challenging,
particularly in situations of conflict and civil unrest, Finally, since physical violence does not always
or when the victim is very young. Additionally, result in injury or death, prevalence data from
differences may exist across countries in the way in surveys are necessary to provide a comprehensive
which some types of killings are treated or classified, picture of the extent to which it occurs. Such
and significant discrepancies in numbers may be data are useful in assessing the magnitude of the
found across different data sources (such as public problem, describing the circumstances surrounding
health and criminal justice records).18 the incidents and shedding light on factors that
can increase children’s risk for violent physical
The production of reliable estimates is further ham- victimization. Demographic and Health Surveys
pered by considerable gaps in the availability of dis- (DHS) provide comparable statistics on specific
aggregated data. Data on victims are often compiled forms of violence for a number of low- and middle-
in broad age categories that do not allow for the cal- income countries. For example, information on the
culation of specific statistics for children. Finally, pop- experience and perpetration of physical violence
ulation data that are necessary to produce homicide is gathered from girls and women of reproductive
rates are often unavailable or unreliable, particularly age (15 to 49 years) and from boys and men aged
in countries affected by rapid population movements 15 and older (depending on the survey, men up to
and where populations are difficult to count.19 the age of 49, 54, 59, 60 or 64 years are included).
To measure physical violence committed by a

Since non-fatal outcomes of physical violence are partner, girls and women and boys and men are
much more common than fatalities, mortality data asked if their spouse/partner committed any of the
need to be combined with figures on the impact of following acts: pushed, shook or threw something
such violence on morbidity. A useful measure of the at her/him; slapped her/him; twisted her/his arm or
impact that poor health, disability and premature pulled her/his hair; punched her/him with her/his fist
death can have on a person’s potential lifespan or with something that could hurt her/him; kicked,
is the disability-adjusted life year (DALY). DALY dragged or beat her/him up; tried to choke or burn
estimates were developed by the WHO in 1990, her/him on purpose; or threatened or attacked her/
32
him with a knife, gun or other type of weapon. This to 6 per cent at age 10 to 14, to 22 per cent at age
question is also asked in reference to experiences 15 to 19 (Figure 3.1A). Among girls, the proportion
that occurred in the 12 months preceding the increases from 0.5 per cent at age 0 to 9, to 5 per
survey. Data from girls and women and boys and cent at age 10 to 14, to 14 per cent at age 15 to 19
men who reported that they experienced various (Figure 3.1B). About 5 per cent of these adolescent
forms of physical violence at the hands of their deaths are due to homicides.
current or most recent spouse or partner are then
combined with information collected from both
ever- and never-married girls and women/boys and The share of deaths due to intentional
men on whether they have ever been hit, slapped, injuries, including homicide, increases as
kicked or hurt physically in another way by anyone children enter adolescence
(other than a spouse or partner) since the age of 15.
The prevalence of any physical violence since age Figure 3.1A
15, committed by anyone, is then reported for all
Percentage distribution of deaths among boys aged 0 to 19
girls and women and boys and men, regardless of
years in 2012, by cause and by age group
their marital status.
15 to 19 years
Given that questions on physical violence focus
on respondents’ experiences since age 15 and on 10 to 14 years
physical victimization by an intimate partner, the data
obtained from the DHS only partially capture corporal 5 to 9 years
punishment by parents and physical bullying, which
are common forms of violence against young 0 to 4 years
children, as illustrated in Chapters 5 and 6. DHS
0 10 20 30 40 50 60 70 80 90 100
data therefore need to be read in conjunction with
other statistics on physical victimization presented Communicable diseases, maternal, Unintentional injuries
neonatal and nutritional conditions Other intentional injuries
in this chapter as well as in other parts of the report
Non-communicable diseases Homicides
to get a sense of how widespread the experience of
violence is among children.
Preventable deaths Figure 3.1B
In 2012 alone, an estimated 8.8 million deaths Percentage distribution of deaths among girls aged 0 to 19
occurred worldwide among boys and girls under age years in 2012, by cause and by age group
20.22 A look at what children and adolescents are
dying of, and where, provides a glimpse of the impact 15 to 19 years
of violence on these young lives. From 0 to 9 years
of age, 85 per cent of children’s deaths are as a result 10 to 14 years
of communicable or non-communicable diseases,
and mortality is not differentiated substantially by 5 to 9 years
sex. As children enter the second decade of their

lives, injuries – both intentional and unintentional – 0 to 4 years
become more prominent, responsible for the deaths
0 10 20 30 40 50 60 70 80 90 100
of 28 per cent and 44 per cent of adolescents aged
10 to 14 and 15 to 19, respectively. At this stage, Communicable diseases, maternal, Unintentional injuries
gender starts to play a role, and mortality patterns neonatal and nutritional conditions Other intentional injuries
due to injuries begin to diverge.23 The proportion of Non-communicable diseases Homicides
deaths due to intentional injuries (out of all causes)

Source: World Health Organization, Global Health Estimates (GHE) Summary Tables:
among boys rises from 0.5 per cent at age 0 to 9, Deaths by cause, age, sex and region, 2012, WHO, Geneva, 2014, recalculated by UNICEF.
33
Homicides Figure 3.2
While most homicide victims are adults, young
people are not exempt. In 2012 alone, homicide took
In 2012, almost one in five homicide victims
the lives of almost 95,000 children and adolescents
worldwide were under the age of 20
aged 0 to 19 – almost one in five of all homicide Percentage distribution of all homicide victims in 2012, by
victims that year (Figure 3.2). age group
Most of these young victims (around 85,000 or 90

per cent) lived in low- and middle-income countries.
The highest homicide levels among children and
adolescents are found in Latin America and the
Caribbean (Table 3.1). In 2012 alone, more than
25,000 homicide victims in that region were below
the age of 20 (representing about one quarter of all 0 to 19 years
homicide victims worldwide). This translates into a 20 and above
homicide rate of 12 victims per 100,000 population,
three times the global average of 4 per 100,000. Source: World Health Organization, Global Health Estimates (GHE) Summary Tables:
Latin America and the Caribbean also has the Deaths by cause, age, sex and region, 2012, WHO, Geneva, 2014, recalculated by UNICEF.
largest global share of homicides among individuals

of all ages: In 2012, about one in three homicides
Table 3.1
worldwide were committed in this region (32 per
cent), and the homicide rate for all ages was almost Latin America and the Caribbean has the
four times the global average (23 per 100,000 versus largest share of homicides among children
6 per 100,000).24 Evidence indicates that this pattern and adolescents in the world
of lethal violence is in part attributable to the illicit
activities of organized criminal groups, the presence Number of homicide victims among children and adoles-
cents aged 0 to 19 years and number of homicide victims
of street gangs and the accessibility of firearms.
among children and adolescents aged 0 to 19 years per
In fact, it is estimated that about 30 per cent of 100,000 population in 2012, by region
all homicides in the region are related to crime
Number of Homicide
and gangs and two thirds are perpetrated with a homicide rate per
firearm.25 victims 100,000
Latin America and
the Caribbean 25,400 12
Boys are at particular risk of homicide, accounting
for 70 per cent of victims under 20 years of age West and Central Africa 23,400 10
compared to 30 per cent who are girls. This higher Eastern and Southern Africa 15,000 6
risk is found in every region of the world, but
South Asia 15,000 2
differences between the two sexes are particularly
striking in Latin America and the Caribbean, where Middle East and North Africa 3,700 2
boys are almost seven times more likely to die due Countries outside of
3,800 2
these regions
to interpersonal violence than girls (Figure 3.3).
Central and Eastern Europe

Differences between the sexes are also seen in and the Commonweath of 1,500 1
terms of perpetrators. Globally, almost half (47 per Independent States (CEE/CIS)
cent) of female homicide victims of all ages are killed East Asia and the Pacific 7,100 1
by family members or intimate partners, whereas the
figure for men is 6 per cent.26 Males on the other hand
World 94,900 4
are more likely to be killed by strangers, which is in
part due to their increased likelihood of participating Note: Figures in this table have been rounded.
in crime-related and other violent activities, such as Source: World Health Organization, Global Health Estimates (GHE) Summary Tables:
Deaths by cause, age, sex and region, 2012, WHO, Geneva, 2014, recalculated according to
gang involvement and street fighting (see Box 3.1).27 UNICEF’s regional classification.
34
Figure 3.3
Homicide rates among boys are higher than

those among girls in every region
Number of homicide victims among children and adoles-
cents aged 0 to 19 years per 100,000 population in 2012, by
sex and by region
25
Boys
Girls
Both sexes
20
15
Homicide rate
10
0
Latin America and the Caribbean
West and Central Africa
Eastern and Southern Africa
South Asia
Middle East and North Africa
Countries outside of these regions
CEE/CIS
East Asia and the Pacific
World
© UNICEF/SRLA2009-0057/Asselin
Deaths by cause, age, sex and region, 2012, WHO, Geneva, 2014, recalculated according to
UNICEF’s regional classification.
The three countries with the highest homicide 10 countries with the largest numbers of homicide
rates in the world among children and adolescents victims, Nigeria has the most, with almost 13,000
are all in Latin America and the Caribbean. These deaths, followed by Brazil with approximately
countries, in descending order, are El Salvador, 11,000 (Figure 3.5).
Guatemala and the Bolivarian Republic of Venezuela
(Table 3.2 and Figure 3.4). However, not all When country data are disaggregated by sex, it
countries in the region are heavily affected: Cuba, becomes evident that boys and girls face different
Peru and Suriname, for example, have considerably risks depending on where they live. The highest
lower levels of homicide than other countries in homicide rates among children and adolescents of
the region. Sub-regional disparities are also found both sexes are found in the Democratic Republic of
in other regions. For example, Myanmar has a the Congo, El Salvador, Guatemala, Haiti, Lesotho
homicide rate among children and adolescents that and Swaziland (Figures 3.6A and 3.6B). In other
is significantly higher than the regional average countries with very high rates, the risk of homicide
(7 versus 1 per 100,000 population). Among the differs for boys and girls.
35
The three countries with the highest homicide rates in the world among children and
adolescents are all in Latin America and the Caribbean
Table 3.2
Number of homicide victims among children and adolescents aged 0 to 19 years per 100,000 population in 2012, by region and by country
Central and Eastern Europe East Asia and the Pacific Eastern and Southern Africa West and Central Africa
and the Commonwealth of Democratic Republic of
Myanmar 7 Lesotho 18 14
Independent States the Congo
Cambodia 4 Swaziland 16
Turkmenistan 4 Nigeria 14
Philippines 3 Rwanda 13
Azerbaijan 3 Sierra Leone 9
Uganda 10
Albania 2 Papua New Guinea 3 Equatorial Guinea 9
South Africa 8
Turkey 2 Lao People's Democratic Côte d'Ivoire 9
3 South Sudan 8
Russian Federation 2 Republic Togo 7
Madagascar 8 Gambia 6
Kazakhstan 1 Solomon Islands 3
Republic of Moldova 1 Botswana 7 Guinea-Bissau 6
Democratic People's
Uzbekistan 1 3 Mozambique 7 Congo 5
Republic of Korea
Tajikistan 1 Ethiopia 6 Guinea 5
Fiji 2
Ukraine 1 Namibia 6 Ghana 5
Timor-Leste 2
Georgia 1 Zambia 6 Mali 5
Armenia 1 Thailand 2 United Republic of Tanzania 6 Central African Republic 5
Bulgaria 1 Mongolia 2 Angola 5 Benin 5
Romania 1 Viet Nam 1 Somalia 5 Burkina Faso 5
Montenegro 1 Burundi 5 Cameroon 5
Republic of Korea 1
Serbia 1 Chad 4
Singapore 1 Comoros 5
Belarus 1 Gabon 4
Indonesia 1 Zimbabwe 5
Kyrgyzstan 0 Liberia 4
Kenya 3
Croatia 0 Malaysia 1 Mauritania 3
Eritrea 3
Bosnia and Herzegovina 0 China 1 Niger 3
Malawi 1 Senegal 1
The former Yugoslav Brunei Darussalam 0
0 Mauritius 1
Republic of Macedonia Cabo Verde 1
Regional average 1 Regional average 1 Regional average 6 Regional average 10
Middle East and North Africa South Asia Latin America and the Caribbean Countries outside of these
Afghanistan 8
regions
Sudan 6 El Salvador 27
Djibouti 5 Pakistan 4 Guatemala 22 United States 4
India 2 Canada 2
Iraq 4 Venezuela (Bolivarian
20 Estonia 1
Jordan 3 Maldives 1 Republic of)
Latvia 1
Yemen 2 Sri Lanka 1 Haiti 19 Belgium 1
United Arab Emirates 2 Bhutan 1 Brazil 17 New Zealand 1
Iran (Islamic Republic of) 2 Bangladesh 1 Panama 15 Finland 1
Algeria 2 Nepal 1 Lithuania 1
Honduras 13
Morocco 1 Israel 1
Colombia 13 Poland 1
Syrian Arab Republic 1 Regional average 2
Jamaica 13 Hungary 1
Egypt 1
Trinidad and Tobago 12 Australia 1
Tunisia 1 Netherlands 0
Lebanon 1 Mexico 7
Switzerland 0
Saudi Arabia 1 Paraguay 7 Germany 0
Bahrain 1 Bolivia (Plurinational State of) 7 Denmark 0
Libya 1 Ecuador 6 Greece 0
United Kingdom 0
Oman 0 Guyana 6
Sweden 0
Kuwait 0 Belize 6 Cyprus 0
Qatar 0 Bahamas 5 France 0
Costa Rica 5 Austria 0

Regional average 2 Italy 0
Nicaragua 4
Spain 0
Dominican Republic 4 Japan 0
Argentina 3 Slovenia 0
Barbados 3 Czech Republic 0
Chile 2 Portugal 0
Slovakia 0
Uruguay 2 Norway 0
Notes: Zeroes appearing in the table do not necessarily mean that there were no homi-
Peru 2 Ireland 0
cide victims in these countries, but rather that the recalculated homicide rates came to
0 after rounding. WHO Global Health Estimates are available for WHO Member States Cuba 2 Iceland 0
whose population were over 250,000 in 2012. Countries in bold have homicide rates Suriname 0 Luxembourg 0
above the regional average. Malta 0
Deaths by cause, age, sex and region, 2012, WHO, Geneva, 2014. Regional average 12 Regional average 2
36
Figure 3.4
Number of homicide victims among children and adolescents aged 0 to 19 years per 100,000 population in 2012, in the 10 countries
with the highest homicide rates among this population group
30
Homicide rate
20
10
0
El Salvador Guatemala Venezuela Haiti Lesotho Brazil Swaziland Panama Democratic Nigeria
(Bolivarian Republic of the
Republic of) Congo
Figure 3.5
More than half of all young homicide victims worldwide were killed in 10 countries
Number of homicide victims among children and adolescents aged 0 to 19 years in 2012, in the 10 countries with the largest
numbers of victims among this population group
Nigeria
Brazil
India
Democratic Republic of the Congo
Mexico
Ethiopia
United States
Pakistan
Colombia
Venezuela (Bolivarian Republic of)
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000
The highest homicide rates among both boys and girls are found in the Democratic Republic of
the Congo, El Salvador, Guatemala, Haiti, Lesotho and Swaziland
Figure 3.6A Figure 3.6B

Number of homicide victims among boys aged 0 to 19 years Number of homicide victims among girls aged 0 to 19 years
per 100,000 population in 2012, in the 10 countries with the per 100,000 population in 2012, in the 10 countries with the
highest homicide rates among this population group highest homicide rates among this population group
50
40
30
20
10
0
El Salvador
Venezuela (Bolivar-
ian Republic of)
Guatemala
Brazil
Haiti
Panama
Lesotho
Swaziland
Colombia
Democratic Republic
of the Congo
Lesotho
Nigeria
Guatemala
Rwanda
Honduras
Haiti
El Salvador
Swaziland
Democratic Republic
of the Congo
Uganda
Homicide rate
Source for all figures on this page: World Health Organization, Global Health Estimates (GHE) Summary Tables: Deaths by cause, age, sex and region, 2012, WHO, Geneva, 2014.
37
The risk of dying as a result of homicide varies Box 3.2), where the homicide rate among children
depending on a child’s age. Data from 2012 show aged 0 to 9 is below 1 per 100,000 but increases
that about 30,000 children under age 10 were killed to 32 per 100,000 among adolescents aged 10 to
by homicide (an estimated 6 per cent of homicides 19, with levels more than 10 times higher among
in people of all ages that year) (Figure 3.7). Two in males compared to females (58 per 100,000 and
three of these children were less than 5 years old. 5 per 100,000). In these two countries, as well as
Among adolescents aged 10 to 19, the death toll is in five other Latin American and Caribbean nations,
even higher, with approximately 65,000 victims (or homicides kill more adolescent boys than any other
13 per cent of total homicides). The largest share cause of death, and many of the advances made in
of homicides among young victims (57 per cent) is child survival are offset by losses due to homicide
found among adolescents aged 15 to 19, followed (Figures 3.9A and 3.9B).
by children under age 5.
Compared to boys in early adolescence (aged 10 Figure 3.7

to 14), boys in late adolescence (aged 15 to 19)
experience a dramatic increase in their risk of dying More than half of young homicide victims
by homicide (Figure 3.8). This is true particularly in are between the ages of 15 and 19
Latin America and the Caribbean. In fact, around Number of homicide victims among children and adoles-
4 per cent of the world’s homicide victims are cents aged 0 to 19 years in 2012, by age group
adolescent boys aged 15 to 19 living in this region. 15 to 19 years
10 to 14 years
Country data also confirm that the risk of dying 54,000 5 to 9 years
0 to 4 years
by homicide increases when children enter
adolescence (Maps 3.1A and 3.1B). In the Bolivarian
Republic of Venezuela, for instance, the homicide 11,000
rates rise from 1.2 per 100,000 among children 11,000
less than 10 years old to 39 per 100,000 among
adolescents aged 10 to 19, with significant sex 19,000
differences: 74 per 100,000 among adolescent boys
Note: Figures in this table have been rounded.
and 3 per 100,000 among adolescent girls. Striking Source: World Health Organization, Global Health Estimates (GHE) Summary Tables:
disparities by age and sex are also found in Brazil (see Deaths by cause, age, sex and region, 2012, WHO, Geneva, 2014, recalculated by UNICEF.
Figure 3.8
Homicide rates increase dramatically in late adolescence, particularly among boys

Number of homicide victims among children and adolescents aged 0 to 19 years per 100,000 population in 2012, by sex, age group
and region
Boys Girls
Latin America and the Caribbean 15 to 19 years
West and Central Africa 10 to 14 years
Eastern and Southern Africa 5 to 9 years
0 to 4 years
South Asia
CEE/CIS
World
70 60 50 40 30 20 10 0 0 10 20 30 40 50 60 70
Homicide rate Homicide rate
Source: World Health Organization, Global Health Estimates (GHE) Summary Tables: Deaths by cause, age, sex and region, 2012, WHO, Geneva, 2014, recalculated according to UNICEF’s regional
classification.
38
Map 3.1A
Number of homicide victims among children aged 0 to 9 years per 100,000 population in 2012, by country
20 and above
10 - 19.99
5 - 9.99
0.00 - 4.99
Countries with no comparable
data on homicides from WHO
Map 3.1B
Number of homicide victims among adolescents aged 10 to 19 years per 100,000 population in 2012, by country
20 and above
10 - 19.99
5 - 9.99
0.00 - 4.99
data on homicides from WHO
Notes: These maps are stylized and not to scale. They do not reflect a position by UNICEF on the legal status of any country or territory or the delimitation of any frontiers. The dotted line represents
approximately the Line of Control in Jammu and Kashmir agreed upon by India and Pakistan. The final status of Jammu and Kashmir has not yet been agreed upon by the parties. The final boundary
between the Sudan and South Sudan has not yet been determined. The final status of the Abyei area has not yet been determined. WHO Global Health Estimates are available for WHO Member
States whose population were over 250,000 in 2012.
Source for both maps on this page: World Health Organization, Global Health Estimates (GHE) Summary Tables: Deaths by cause, age, sex and region, 2012, WHO, Geneva, 2014.
39
Figure 3.9A
In seven countries in Latin America and the Caribbean, homicide is

the leading cause of death among adolescent boys
Percentage distribution of deaths among boys aged 10 to 19 years in 2012, by cause, in the
seven countries with the highest homicide rates worldwide among this population group
Homicides Unintentional injuries Communicable diseases, maternal,

neonatal and nutritional conditions
Other intentional injuries Non-communicable diseases
Panama
El Salvador
Trinidad and Tobago
Brazil
Guatemala
Colombia
0 10 20 30 40 50 60 70 80 90 100
Source: World Health Organization, Global Health Estimates (GHE) Summary Tables: Deaths by cause, age, sex and region, 2012, WHO,
Geneva, 2014.
Figure 3.9B
In some countries, advances in child survival have been offset by
deaths due to homicide
Average number of children under age 5 whose lives were saved each year between 2000 and
2012 and number of homicide victims among children and adolescents aged 0 to 19 years in
2012, in the seven countries with the highest homicide rates worldwide among boys aged 10
to 19 years
Lives saved Lives lost
Brazil
Guatemala
Colombia
El Salvador
Panama
Trinidad and Tobago

35000
30000
25000
20000
15000
10000
5000
5000
10000
15000
20000
25000
30000
35000
Note: Figures in this chart have been rounded.

Source: For data on lives saved: UNICEF analysis based on the UN Inter-agency Group for Child Mortality Estimation (IGME) 2013. For data
on lives lost: World Health Organization, Global Health Estimates (GHE) Summary Tables: Deaths by cause, age, sex and region, 2012, WHO,
Geneva, 2014.
40
© UNICEF/NYHQ2012-0086/Zaidi
41
National per capita income is Most countries with high income levels and low inequality
an important variable that may have low homicide rates
influence the homicide levels In low- and middle-income countries and those with higher
within a country, including among levels of inequality, significant variations in homicide rates
the youngest segment of the are found
population. Other factors include
economic or social inequalities, Figure 3.10A
the availability of firearms, poor Number of homicide victims among children and adolescents aged 0 to 19 years
education standards, ineffective per 100,000 population in 2012 and gross national income (GNI) per capita in US$
in 2012, in high-income countries
governance and rule of law, high
30
levels of crime and the presence
of gangs.28 Among cultural
factors that may have an effect
on homicide rates are social 20
norms that condone violence
Homicide rate
as an acceptable way to resolve

Trinidad and Tobago
conflicts, that give priority to
10
parental rights over child welfare, Equatorial Guinea
and that sanction male dominance Bahamas

over women and children.29 United States
Luxembourg Switzerland Norway
0
12,000 22,000 32,000 42,000 52,000 62,000 72,000 82,000 92,000
Available data show that the
GNI per capita, US$
homicide rate among children
Notes: Each dot represents a country. For the 2014 fiscal year, countries are classified as high-income by the World Bank if
and adolescents in high-income they have a GNI per capita, calculated using the World Bank Atlas method, of $12,616 or more in 2012.
Source: Data on GNI per capita are from the World Bank, 2012 (available at: <http://data.worldbank.org/indicator/NY.GNP.
countries is very likely to be PCAP.PP.CD>, accessed 5 June 2014). Data on homicides are from World Health Organization, Global Health Estimates (GHE)
Summary Tables: Deaths by cause, age, sex and region, 2012, WHO, Geneva, 2014.
below 5 per 100,000 population.
Notable exceptions include the
Bahamas, Equatorial Guinea Figure 3.10B
and Trinidad and Tobago (Figure
Number of homicide victims among children and adolescents aged 0 to 19 years
3.10A). However, in low- and per 100,000 population in 2012 and gross national income (GNI) per capita in US$
middle-income countries, the in 2012, in low- and middle-income countries
relationship between income 30
and homicide rates becomes El Salvador
murky and inconsistent: Some
Guatemala
of these countries have very high Haiti Venezuela (Bolivarian Republic of)
20
rates and others very low (Figure Lesotho
Brazil
3.10B). Similar findings can be Panama
observed when homicide rates
Homicide rate
Jamaica Colombia
are analysed in conjunction with 10

economic inequality. Lower rates
of homicide can be observed
in countries with lower levels
0
of inequality, while significant 0 2,000 4,000 6,000 8,000 10,000 12,000
variations in homicide rates are GNI per capita, US$
Notes: Each dot represents a country. For the 2014 fiscal year, countries are classified as low- and middle-income by the
found among countries with World Bank if they have a GNI per capita, calculated using the World Bank Atlas method, of $12,615 or less in 2012.
higher levels of inequality (Figure Source: Data on GNI per capita are from the World Bank, 2012 (available at: <http://data.worldbank.org/indicator/NY.GNP.
PCAP.PP.CD>, accessed 5 June 2014). Data on homicides are from World Health Organization, Global Health Estimates (GHE)
3.10C). Summary Tables: Deaths by cause, age, sex and region, 2012, WHO, Geneva, 2014.
42
Figure 3.10C contributors to the disease

burden. Among boys and girls
Number of homicide victims among children and adolescents aged 0 to 19 years between the ages of 5 and 9,
per 100,000 population in 2012 and Gini coefficient in 2012
non-communicable diseases
as well as unintentional injuries
30
grow in importance as causes
El Salvador
of mortality and morbidity.
Guatemala
Neuropsychiatric disorders and
Venezuela (Bolivarian Republic of) intentional injuries (especially
20 Haiti
Lesotho homicide in certain regions)
Brazil
Panama begin to play a more prominent
Homicide rate
Honduras role in early and late adolescence

Colombia
10 (between the ages of 10 and 19).
Afghanistan South Africa
Namibia
Comoros Even though violence affects
Sweden
children more as they move
0
25 30 35 40 45 50 55 60 65 towards adolescence, the
Gini coefficient youngest children (those under
age 5) are not immune to its
Notes: Each dot represents a country. The Gini coefficient is a measure of the deviation of the distribution of income (or con- impact. In fact, among the
sumption) among individuals or households within a country from a perfectly equal distribution. A value of 0 represents absolute
equality; a value of 100 indicates absolute inequality. youngest boys and girls in all
Source: Data on homicides are from World Health Organization, Global Health Estimates (GHE) Summary Tables: Deaths by
cause, age, sex and region, 2012, WHO, Geneva, 2014. Data on the Gini coefficient are from World Bank, World Development regions, the largest proportion of
Indicators 2012, Washington, DC, 2012, available at <http://data.worldbank.org>,accessed 5 June 2014. DALYs due to intentional injuries
are attributable to homicides
Impact of interpersonal violence on mortality rather than other intentional
and morbidity injuries caused by either self-
harm or collective violence.
While the statistics on homicide are appalling in themselves, they At this early stage of life, sex
capture only part of the story. The full impact of homicide on children differences in disease burden
reaches beyond the tragedy of individual deaths and can mean the loss are practically non-existent
of a parent or friend, thereby damaging the familial and social bonds that across all regions. In West and
hold a child’s world together. Moreover, children and adults who survive Central Africa, around 7 years of
a homicide attempt may sustain serious injuries or suffer emotional ‘healthy’ life are lost per 1,000
trauma that lingers long after the immediate incident has passed. boys and girls under age 5 due
to homicide as compared to 0
To glean some understanding of how violence affects children, it is years of ‘healthy’ life lost per
important to look at its impact in the broader context of how injuries 1,000 boys and girls due to other
and disease change the quality of their lives. From early to mid- intentional injuries. However,
this region has the highest
childhood and into adolescence, there is a shift in both the overall
disease burden due to homicide
levels of the disease burden as well as the major causes of mortality
among children aged 5 to 9 (5.72
and morbidity. According to WHO, the ‘disease burden’ in this context
DALYs per 1,000 boys and 6.96

is a measurement of the gap between current health status and an
DALYs per 1,000 girls). In the
ideal situation where everyone lives into old age, free of disease and
Middle East and North Africa,
disability. As can be seen in Figures 3.11A to 3.11D, among all children other intentional injuries caused
and adolescents under age 20 the average global disease burden is by either self-harm or collective
highest among girls and boys under age 5; this translates into about violence contribute more to the
1,000 years of ‘healthy’ life lost per 1,000 children aged 0 to 4 years. disease burden than homicides
This situation is largely driven by the fact that mortality rates are for children of both sexes aged
highest among the youngest children, particularly in the first few years 5 to 9, although DALY rates are
of life. As children grow older, there is a noticeable shift in the leading higher among boys than girls.
43
Once children reach early adolescence (10 to 14 and Central Africa, for instance, around six years of
years old), the impact of homicide on mortality and ‘healthy’ life are lost per 1,000 boys due to homicide
morbidity, as well as related sex differences in the compared to about three years lost due to either self-
burden, begin to emerge, and these patterns become
harm or collective violence. On the other hand, in
even more pronounced during late adolescence.
Central and Eastern Europe and the Commonwealth
Among children aged 10 to 14 in West and Central
of Independent States (CEE/CIS), South Asia, and
Africa and boys in Latin America and the Caribbean,
the largest share of the disease burden due to the Middle East and North Africa, DALY rates due to
intentional injuries is caused by homicide rather than self-harm or collective violence are higher than DALY
other intentional injuries. For boys of this age in West rates for homicides.
As children age, the increasing risk of homicide affects the number of healthy years they can
expect to live
Figure 3.11A
Major causes of disease burden in disability-adjusted life years (DALYs) per 1,000 children aged 0 to 4 years, by sex and by region
Boys Girls
South Asia
Communicable diseases, maternal, Middle East and North Africa
Non-communicable diseases
Unintentional injuries Latin America and the Caribbean
Other intentional injuries CEE/CIS
Homicide
World
3000 2500 2000 1500 1000 500 0 0 500 1000 1500 2000 2500 3000
Figure 3.11B
Boys Girls
South Asia
Communicable diseases, maternal,
neonatal and nutritional conditions Latin America and the Caribbean
Non-communicable diseases CEE/CIS
Unintentional injuries
Other intentional injuries
Homicide Countries outside of these regions
World
600 550 500 450 400 350 300 250 200 150 100 50 0 0 50 100 150 200 250 300 350 400 450 500 550 600
44
This regional pattern continues among youth among 15- to 19-year-olds is particularly striking in
between the ages of 15 and 19, and the burden of Latin America and the Caribbean. There, about 58
disease due to both homicides and other intentional years of ‘healthy’ life are lost per 1,000 boys aged 15
injuries is systematically higher among boys than to 19 due to intentional injuries, compared to about 9
girls in nearly every region. In CEE/CIS for example, years lost per 1,000 girls in the same age group. This
DALY rates due to self-harm or collective violence are difference is largely driven by the burden caused by
higher than those due to homicide for both boys and homicide among boys in this region, where the DALY
girls of this age, but the number of DALYs among rate for this cause was 51 per 1,000 boys aged 15 to
boys for both causes are about double that of girls. 19 compared to about 5 per 1,000 girls in the same
The sex difference in disease burden due to homicide age group.
Figure 3.11C
Boys Girls
South Asia
Communicable diseases, maternal, Latin America and the Caribbean
Unintentional injuries East Asia and the Pacific
Homicide
World
400 350 300 250 200 150 100 50 0 0 50 100 150 200 250 300 350 400
Figure 3.11D
Major causes of disease burden in disability-adjusted life years (DALYs) per 1,000 adolescents aged 15 to 19 years, by sex and by region
Boys Girls
Communicable diseases, maternal, South Asia

Unintentional injuries East Asia and the Pacific
Homicide
World
400 350 300 250 200 150 100 50 0 0 50 100 150 200 250 300 350 400
Notes: Disability-adjusted life years (DALYs) are a summary measure combining years of life lost due to premature mortality (YLLs) and years lost due to disability (YLDs) for incident cases of the
disease or injury. One DALY represents the loss of the equivalent of one year of full health. The data have been recalculated according to UNICEF’s regional classification.
Source for all figures on pages 44 and 45: World Health Organization, Global Health Estimates (GHE) Summary Tables: Disability-Adjusted Life Years (DALYs) by cause, age, sex and region, 2012,
WHO, Geneva, 2014.
45
Adolescent involvement in gang violence
Box 3.1
Gang violence weighs heavily on the life opportunity.34 There may also be been shown to be associated with
lives of adolescents throughout the the perception that gang peers are early initiation of antisocial and de-
world. Gang members are typically in like family, offering emotional support linquent behaviour as well as alcohol
their teens or early twenties (between and protection. Some additional fac- and drug use.39
the ages of 12 and 24), although some tors that have been found to increase
members are older.30 Early adoles- a young person’s risk of joining a gang Estimates of the numbers of gangs
cence (13 years) is the average age for include low levels of school commit- and gang members are limited and it
entry. It is relatively well-documented ment and academic achievement, as- is difficult to know the extent of youth
that girls are less likely to be involved sociation with delinquent peers and
involvement. Nevertheless, around
in gangs than boys; some research poor parental management practices,
273,875 children (under age 18) were
has also indicated that older adoles- including limited supervision.35 Chil-
believed to be involved in 29,900
cents, particularly those aged 15 and dren are often recruited into gangs by
above, are more likely to report gang physical force, intimidation or threats, gangs in the United States in 2011.40
membership than younger teens.31 and once a member (regardless of the In that country, organized crime/gang-
motive), there is usually no turning related homicide rates have been quite
Children join gangs for a variety of back. stable at below 0.3 per 100,000 since
reasons. A substantial body of litera- 2007.41 In North, South and Central
ture supports the notion that youth Young people who become involved America and the Caribbean, consid-
involvement in gang life results from in gangs are substantially more likely ered some the most violent regions in
a complex interplay of risk factors at to face negative health outcomes, in- the world, about 3 in 10 homicides are
the community, school, peer, fam- cluding violent victimization and even
related to organized criminal groups
ily and individual levels.32 Community death.36 Some research has noted that
and gangs.42 Within the context of
risk factors can include poverty, social girls are less likely than boys to expe-
relatively high overall homicide levels
exclusion, lack of jobs and educa- rience serious gang victimization or to
tional opportunities, the instability of be killed in a gang-related homicide.37 in Central America and the Caribbean,
residential neighbourhoods and the However, recent qualitative research countries such as the Bahamas, Be-
availability of firearms and/or drugs.33 in the UK has revealed that girls in- lize, El Salvador and Honduras report
In the absence of better alternatives volved in gang life can be at particular an increasing trend since 2005 in the
and in extreme conditions of exclu- risk of sexual violence and exploita- numbers of homicides linked to gangs
sion, joining a gang can be seen as a tion.38 Being part of a gang has also and organized criminal groups.43
Tackling the problem of

Box 3.2
adolescent homicides in Brazil

Violence and crime have increased 12 times more homicides among ado- As the availability of firearms in Brazil
dramatically in Brazil in recent dec- lescent boys than adolescent girls and increases, youth are particularly vul-
ades, particularly in large urban cen- three times more among adolescents nerable to homicides involving these
tres.44 Rising homicide rates have of African descent than white teens.46 lethal weapons. Roughly 90 per cent
been attributed to growing inequality Recent years, however, have also seen of homicides in males aged 15 to 44
within the country, greater availability a troubling increase in the number involve firearms,48 and adolescents
of firearms, increased drug use and a of girl victims of homicide, particu- are six times more likely to be killed by
substantial growth in the youth popu- larly among those of African heritage. firearms than by any other means.49 A
lation.45 Adolescent boys of African de- Between 1997 and 2007, homicides 2012 study projected that 36,000 of the
scent face the greatest risk of homicide among girls aged 10 to 19 with African country’s adolescents would likely be
in Brazil. In 2010, there were roughly roots rose by 55 per cent.47 killed as a result of homicide by 2016
46
unless proper prevention measures devising strategies to address it. state departments and 44 municipal
are adopted.50 These findings speak Local and national seminars were departments. It revealed that, despite
to the need to incorporate gender and also organized to bring public officials the country’s very high adolescent
race as critical components of preven- together with representatives of civil homicide rates, programmes concen-
tion initiatives and the importance of society organizations to exchange trating specifically on reducing lethal
gun-control policies. ideas and to further PRVL-initiated violence are scare.
research on adolescent homicide.
In response to the alarming levels of The PRVL is also supporting the devel-
youth violence, UNICEF Brazil is sup- Between 2011 and 2012, a guide was opment of national homicide preven-
porting a Programme for Reducing created to support municipalities in tion policies. In 2012, the Government
Lethal Violence (PRVL), together with better understanding the local dynam- launched the Youth Alive National Plan
the Favela Observatory, the Secretariat ics of lethal violence against young with the goal of reducing the expo-
for Human Rights and the Violence people, develop assessments of the sure and vulnerability of adolescents
Analysis Laboratory. The programme, problem and get prevention initia- and young people to violence in urban
which is operational in 16 metropoli- tives under way. The Municipal Guide areas.54 The plan includes actions
tan areas, has the goal of combating for Preventing Lethal Violence against ranging from creating opportunities
lethal violence against adolescents Adolescents and Young People, 52 pro- for social inclusion and autonomy to
in large urban centres.51 In each me- duced by UNICEF and partners, focuses supplying equipment, utilities and
tropolis, the PRVL prioritizes urban on 238 municipalities, particularly those living spaces in areas with a high
areas most affected by high rates with high adolescent homicide rates. concentration of homicides and rais-
of adolescent homicides, including ing the awareness of public officials
slums and other impoverished com- about the problem. UNICEF and part-
Production of indicators
munities and surrounding suburbs. ners contributed substantially to the
Since its inception in 2007, the PRVL In addition to advocacy, the PRVL has
drafting of the plan based on lessons
has had three main objectives: (1) to developed a system for monitoring
learned through the PRVL. In addi-
sensitize and mobilize civil society and adolescent homicide rates, with the
tion, the methodologies and products
the Government to address the issue ultimate aim of supporting impact
generated by the PRVL – including the
of adolescent homicides, (2) to de- evaluations of lethal violence preven-
Municipal Guide, the annual reports of
velop annual indicators that allow for tion policies. To achieve this goal, the
the Adolescent Homicide Index55 and
systematic monitoring of adolescent programme devised an Adolescent
analyses of local experiences in pre-
homicide rates and the effectiveness Homicide Index, which has been pub-
venting violence56 – were incorporated
of related policies, and (3) to dissemi- lished annually since 2009. The index
into the national plan.57 According to
nate methodologies to reduce killings measures trends in homicides among
the National Secretariat for Youth, the
of adolescents in Brazil. In each pillar adolescents aged 12 to 18, focusing on
Brazilian cities with populations over plan is expected to benefit 132 munici-
of the programme, actions are carried
100,000. For a group of 1,000 adoles- palities that, up until 2010, accounted
out in a networked fashion, as de-
cents, estimates are calculated of the for 70 per cent of all homicides among
scribed below.
number of boys or girls above age 12 adolescents of African descent.58
who are likely to be killed by homicide
Political advocacy, awareness-raising In 2013, UNICEF also launched the
and social mobilization before age 19. The objective in pub-
fourth edition of the Municipal Seal
lishing these figures is to keep Bra-
The PRVL has focused on political of Approval initiative59 and the sec-
zilian society abreast of the problem
advocacy at the highest levels of gov- ond edition of the Platform for Urban
and its escalation and to prioritize the
ernment to raise public awareness Centres,60 its two main mobilization
specific needs of children in this age
and to bring civil society and govern- strategies for promoting equity and
group in policies and programmes
ment together to address the issue of reducing social disparities. Along with
designed to reduce the number of vic-
homicides among adolescents. One tims of urban violence in the country. other measures of youth exposure to
result is that the issue became part violence, homicide rates are now be-
of the Multi-Year Government Action ing used as impact indicators in these
Methodologies for violence prevention initiatives. The Municipal Guide is

Plan. The programme has also en-
couraged the participation of young The third pillar of the PRVL strategy one of the main tools used for train-
people themselves in developing local is identifying, analysing and dissemi- ing managers in urban areas enrolled
violence prevention policies and has nating local- and state-level policies in the Municipal Seal of Approval ini-
strengthened the capacity of munici- for reducing lethal violence against tiative. Currently, it involves 611 mu-
pal governments to address lethal vio- adolescents and young people. An ini- nicipalities in the Amazon region and
lence. Workshops with young people tial survey was conducted to analyse 1,131 municipalities in the semi-arid
from 11 metropolitan areas were held violence prevention efforts developed Northeast region, representing 75 per
to raise awareness of the problem by state and local departments in 11 cent of eligible municipalities in these
of violence and to engage youth in metropolitan areas,53 covering 163 two areas.
47
Figure 3.12 Non-fatal physical
violence among
Over half of adolescent girls report incidents of physical
adolescent girls
violence since age 15 in the Democratic Republic of the
Congo and Uganda Comparable DHS data from 42
Percentage of girls aged 15 to 19 years who experienced any physical violence
low- and middle-income countries
since age 15 and percentage of girls aged 15 to 19 years who experienced any reveal that physical violence is
physical violence in the last 12 months common in the lives of many
adolescent girls. The proportion
Democratic Republic of the Congo of adolescent girls who reported
n/a
Bolivia (Plurinational State of)
Uganda experiencing some form of
Cameroon physical violence since age 15
Equatorial Guinea varies widely across countries,
Egypt
ranging from 4 per cent of girls
Liberia
Gabon aged 15 to 19 in Kazakhstan to
Marshall Islands over 50 per cent of girls of the
Côte d’Ivoire same age in the Democratic
Zambia
Republic of the Congo and
Namibia
Ghana Uganda (Figure 3.12).
Jordan
Pakistan Substantial regional variations can
Timor-Leste
also be observed. The proportions
Nigeria
Kenya of adolescent girls who reported
Haiti experiencing physical violence are
Sao Tome and Principe highest in West and Central Africa,
United Republic of Tanzania
Zimbabwe
Eastern and Southern Africa, and
Mozambique East Asia and the Pacific. In these
Burkina Faso three regions, at least one in nine
Cambodia girls were reportedly subjected
India
Malawi to physical violence since age 15
Rwanda n/a in every country for which data
Cabo Verde are available. In three countries
Honduras n /a
in West and Central Africa
Philippines
Peru n /a
(Cameroon, the Democratic
Republic of Moldova Republic of the Congo and
Dominican Republic Equatorial Guinea), over 40 per
Colombia n /a
cent of adolescent girls said they
Comoros
Experienced physical violence since age 15 experienced physical violence
Nepal
Experienced physical violence in the last 12 months
Azerbaijan since age 15. In Eastern and
Tajikistan Southern Africa, the comparable
Ukraine
share is over half in Uganda
Kyrgyzstan
Kazakhstan and one third in Zambia. In East
0 10 20 30 40 50 60 70 80 90 100 Asia and the Pacific, over one
third of adolescent girls reported
Notes: Data on the proportions of adolescent girls who experienced any physical violence in the last 12 months are not being subjected to physical
available for Colombia, Honduras, Peru and Rwanda. Data for the Plurinational State of Bolivia, Cambodia, Egypt, Jordan
and Pakistan refer to ever-married girls aged 15 to 19 years. Data on the proportion of adolescent girls who experienced violence since age 15 in the
any physical violence since age 15 are not available for the Plurinational State of Bolivia. Data for Colombia and Peru refer
only to physical violence committed by someone other than the current or most recent spouse or partner. Data for the Mar-
Marshall Islands. In contrast, the
shall Islands are based on 25-49 unweighted cases. Data for Kazakhstan are from MICS 2010-2011, which used an adapted prevalence of physical violence is
version of the DHS module on domestic violence.
Source: UNICEF global databases, 2014, based on DHS and MICS, 2005-2013. lower among adolescent girls in
48
CEE/CIS, ranging from 4 per cent in Kazakhstan to 14 comparable proportion among girls and women aged
per cent in the Republic of Moldova. 15 to 24 in Turkey was about 22 per cent (based on
a 2008 survey).65
When looking at recent victimization among
adolescent girls, rates are alarmingly high in several The Violence Against Children Surveys (VACS) in
countries. The highest rates of reported physical Kenya,66 Swaziland,67 the United Republic of Tanzania68
violence in the last year are found in sub-Saharan and Zimbabwe69 also collected information on the
Africa. In Eastern and Southern Africa, at least 12 per
prevalence of physical violence during childhood
cent of girls aged 15 to 19 reported incidents of physical
among adolescents and young adults aged 13 to
violence in the past 12 months in all nine countries for
24. In the case of Swaziland, questions were limited
which data are available, except Comoros. In Uganda
to girls and women. While these surveys generally
and the United Republic of Tanzania, more than one
in five adolescent girls reported past-year incidents addressed many of the same forms of physical
of physical violence. In West and Central Africa, the violence covered in the DHS, there were some
proportion is at least 1 in 14 girls in each of the 11 variations. The surveys in Kenya and Zimbabwe asked
countries with available data; the share reaches as respondents if parents, adult relatives or persons of
high as 26 per cent in Cameroon and 42 per cent authority (such as teachers and police) ever punched,
in the Democratic Republic of the Congo. Outside kicked, whipped or beat them with an object or used
of sub-Saharan Africa, the countries with the highest or threatened to use a knife or other weapon against
reported prevalence of physical violence in the past them. Additionally, questions were also asked about
12 months include the Marshall Islands (30 per cent), whether respondents were ever slapped, pushed,
Timor-Leste (25 per cent), Pakistan (24 per cent) and punched, kicked, beaten with an object or had a
Egypt (22 per cent). Adolescent girls living in CEE/
weapon used or threatened to be used against them
CIS reported the lowest rates of physical violence in
by a current or previous romantic partner. The same
the past year, ranging from 2 per cent in Kazakhstan
questions were also asked in the United Republic
to 7 per cent in the Republic of Moldova.
of Tanzania, except that respondents there were
Information from other nationally representative not asked if they were beaten with an object by
surveys confirms that physical violence is anyone and, in the case of romantic partners, they
widespread among adolescent girls. For instance, in were also asked if they were whipped. In Swaziland,
a 2008-2009 survey conducted in Guatemala,61 about the survey asked girls and women if they were ever
one in four adolescent girls aged 15 to 19 said they kicked, bitten, slapped, punched or threatened with a
were beaten or physically mistreated before age weapon by an adult during their childhood.
15. The same question was also asked in a 2006-
2007 survey in Nicaragua,62 which found that about 8 According to the survey results, 66 per cent of women
per cent of adolescent girls were reportedly beaten aged 18 to 24 in Kenya and 64 per cent in Zimbabwe
or physically mistreated since age 15. A recent reported incidents of physical violence prior to age
survey in Madagascar63 showed that 15 per cent of 18. The most common forms in both countries were
girls aged 15 to 19 experienced at least one of the
being slapped, pushed, punched, kicked, whipped
following acts of physical violence in the 12 months
or beaten with an object. In the United Republic of

preceding the survey: kicked or punched; slapped
Tanzania, 74 per cent of females aged 13 to 24 said
or had their arm twisted; dragged on the ground;
they experienced physical violence before age 18 at
shaken or had something thrown at them; strangled
or burned; threatened or attacked with a knife, gun or the hands of a relative, authority figure or intimate
other weapon; or spat on. In a 2009 national survey partner.70 Here, the most commonly reported forms
in Vanuatu,64 more than one third (37 per cent) of of physical violence were being punched, whipped or
adolescent girls aged 15 to 19 reported incidents of kicked.71 In Swaziland, a quarter of 13- to 24-year-old-
being beaten or physically mistreated by someone females reported experiences of physical violence
other than a husband or partner since age 15. The committed by an adult prior to age 18.
49
Perpetrators of physical violence against girls were the main perpetrators at 31 per cent, while in
In the DHS, girls who reported experiences of Zimbabwe it was another relative (27 per cent).
physical violence since age 15 were asked to
identify the perpetrator, including both partners Physical violence during pregnancy
and non-partners. Parents (mothers or fathers) and
For adolescent girls, pregnancy is a risky time.
other caregivers (stepmothers or stepfathers) were
Pregnant girls must cope with the stress of
the most commonly reported perpetrators in the
transitioning to parenthood and may not be
majority of the 36 countries with available data (Table
physically, mentally or emotionally ready to become
3.3).72 In Azerbaijan, Cambodia, Haiti, Kyrgyzstan and
mothers. Violence can be extremely damaging to
Timor-Leste, for instance, over half of girls named
the health and survival of both an adolescent mother
their mother or stepmother as perpetrators. In the
and her child, resulting in miscarriage, stillbirth,
Plurinational State of Bolivia, Egypt, Mozambique,
premature labour or delivery, and low birth weight.73
Nepal, Pakistan and Zimbabwe, current husbands or
In extreme circumstances, violence can lead to
partners were most often cited.
maternal mortality.74 For instance, a study in Uttar
Pradesh, India found that infants born to women who
The identity of the most common perpetrator
experienced partner violence during pregnancy faced
varied, however, by the marital status of the victim.
double the risks for perinatal and neonatal mortality
Not surprisingly, among ever-married girls who
than infants born to women who did not experience
experienced physical violence since age 15, a current
such violence. The authors of this study concluded
or former partner was cited most often in all of the
countries with available data (results not shown). The that approximately one fifth of stillbirths and deaths
proportion is over 70 per cent in India, Mozambique, during the first month of a newborn’s life might
Nepal, Pakistan, the United Republic of Tanzania possibly have been prevented if partner violence had
and Zambia (see Chapter 7 for an in-depth overview not occurred during pregnancy.75
of intimate partner violence among adolescents).
However, never-married girls were most likely Currently available DHS data from 30 countries76
to report physical violence at the hands of family reveal the prevalence of physical violence against
members, friends/acquaintances and teachers. girls during pregnancy to range from 1 per cent in
The most commonly reported perpetrator was the Cabo Verde to 17 per cent in Pakistan (Figure 3.13).
victim’s mother or stepmother, with over half of More than 1 in 10 ever-pregnant girls experienced
single girls reporting this in Azerbaijan, Cambodia, physical violence during pregnancy in six countries:
Haiti, India, Liberia, Sao Tome and Principe, and Cameroon, the Democratic Republic of the Congo,
Timor-Leste. Some exceptions to this pattern include Equatorial Guinea, Gabon, Haiti and Pakistan.
the Democratic Republic of the Congo, Honduras and Reported rates of violence during pregnancy are
Peru, where the most commonly cited perpetrators generally lower among adolescent girls than adult
were fathers or stepfathers. In four sub-Saharan women. However, the opposite is true in several
African countries, teachers were the most commonly countries, including Haiti and Pakistan, where the
reported perpetrators among single girls: Ghana (34 rates of physical violence during pregnancy for
per cent), Kenya (47 per cent), Uganda (58 per cent) adolescents are substantially higher than those
and the United Republic of Tanzania (39 per cent). among older women. In many cases, the violence is
In Rwanda, neighbours or community members perpetrated by the victim’s partner.77
50
Table 3.3
In almost all countries, parents and other caregivers are the most commonly cited perpetrators
of physical violence against adolescent girls
Percentage of girls aged 15 to 19 years who experienced any physical violence since age 15, by perpetrator
Persons who committed physical violence against girls
Current/former
Mother-in-law
Brother/sister
Daughter/son
Other relative
Police/soldier
band/partner
band/partner
acquaintance
Father-in-law
Current hus-
Other in-law
Former hus-
someone at
stepmother
Employer/
stepfather
boyfriend
Mother/
Teacher
Father/
Friend/
Other
work
Azerbaijan 11 3 0 52 20 0 29 5 0 0 0 - 0 0 0 0
Bolivia (Plurinational State of) 55 6 - 3 0 4 1 1 - 4 3 0 - - 1
Burkina Faso 7 2 3 36 34 0.1 8 24 0.3 0 2 4 3 0 - 16
Cambodia 7 0 - 63 30 1 29 1 0 0 3 0 3 0 0 5
Cameroon 20 5 7 26 22 0 24 9 0.1 0 1 1 16 0 0 11
Colombia 18 11 13 42 0 - 18 - - 1 - - - - 13
Comoros 5 3 5 30 14 0 33 15 2 0 2 - 12 0 0 8
Côte d’Ivoire 13 2 5 36 32 0 26 - 0 0 11 2 1 1 - 10
Democratic Republic of the Congo 23 3 9 37 35 1 45 0 2 0 0 14 11 0.2 1 12
Dominican Republic 26 18 - 24 23 0 7 6 0 0 - 13 0 0 - 5
Egypt 49 3 - 26 48 0 25 - 1 0 - 0 3 0 - 0
Gabon 16 4 13 28 26 0 26 21 0 0.4 0.1 - 3 0 0 5
Ghana 6 0 1 27 14 0 18 12 0 0 0.4 - 33 1 0 22
Haiti 13 2 4 52 40 0 17 16 0.1 1 0.2 - 11 0 0.2 4
Honduras 25 14 2 29 21 0 15 11 0 0 0 - 0 0.2 0 6
India 33 2 0.2 41 18 0.4 25 2 1 0.4 1 - 11 0 0 0.4
Jordan (28) (5) - (43) (42) (0) (23) - (0) (0) - (0) (1) (0) (0) (1)
Kenya 10 2 1 37 23 0 13 10 0.1 0.1 0.1 - 42 0 0.1 18
Kyrgyzstan 4 1 0 55 15 2 30 5 0 0 0 - 3 0 0 12
Liberia 17 6 12 46 39 0.2 12 13 0 0.1 1 - 8 0 2 4
Malawi 22 6 1 14 5 0 21 9 0 0 1 - 3 0 0 31
Mozambique 37 7 5 20 11 0.1 17 10 0 0 1 - 5 0.2 0 8
Nepal 44 0 0 17 22 0 18 7 0 0 0 - 4 0 0 1
Nigeria 9 2 4 41 32 1 30 18 0.2 0 1 7 32 1 0.1 1
Pakistan (85) (1) - (10) (4) (0) (6) (0) (15) (12) (1) - (0) (0) (0) (0)
Peru 22 6 - 33 34 0 22 9 0 0 2 2 0 0 - 7
Philippines 15 3 0 33 28 0 20 11 0 0 0.3 0 0 0.4 0 13
Republic of Moldova 9 6 9 17 38 0 11 6 0 0 0 - 0.3 3 0 13
Rwanda 7 4 3 11 13 0 16 - 0 0 5 3 12 0 2 36
Sao Tome and Principe 25 2 3 44 32 0 9 - 0 0 10 4 0 0 0 5

Tajikistan 13 0 1 47 12 3 30 7 0 0 0 - 4 0 0 1
Timor-Leste 7 0.3 0.3 63 51 2 28 8 0 0 0 - 12 0 0 1
Uganda 12 4 7 26 21 0 10 7 0 0 1 - 48 2 0 11
United Republic of Tanzania 23 4 5 17 3 0 18 13 0 0 0.2 - 28 3 0 12
Zambia 20 3 6 19 12 0 18 14 0 0 2 - 10 0 0 19
Zimbabwe 30 5 4 15 8 0 10 21 1 0 0.4 - 12 1 0 10
Notes: Data for the Plurinational State of Bolivia, Cambodia, Egypt, Jordan and Pakistan refer only to ever-married girls aged 15 to 19 years. Data for Jordan and Pakistan are based on 25-49
unweighted cases.
Source: UNICEF global databases, 2014, based on DHS and MICS, 2005-2013.
51
Figure 3.13
More than one in seven adolescent girls experienced physical violence during pregnancy in
Cameroon, the Democratic Republic of the Congo and Pakistan
Percentage of ever-pregnant girls aged 15 to 19 years and percentage of ever-pregnant women aged 20 to 49 years who
experienced physical violence during pregnancy
Pakistan
Cameroon 15 to 19 years
Democratic Republic of the Congo 20 to 49 years
Haiti
Equatorial Guinea
Gabon
Zambia
Malawi
Marshall Islands
Nepal
Uganda
Jordan
Zimbabwe
Dominican Republic
Sao Tome and Principe
Nigeria
Rwanda
Philippines
Comoros
Côte d’Ivoire
Mozambique
Timor-Leste
Azerbaijan
Tajikistan
Cambodia
Ghana
Burkina Faso
Kyrgyzstan
Cabo Verde
0 10 20 30 40 50
Notes: Data for Jordan and Pakistan refer to ever-married girls and women only. Data for the Marshall Islands and Rwanda are based on 25-49 unweighted cases. Data from the DHS 2005 are used for
Rwanda since information on physical violence during pregnancy was not collected in the most recent DHS, conducted in 2010.
Source: UNICEF global databases, 2014, based on DHS, 2005-2013.
non-fatal physical violence physical violence is found in Uganda, where over half
among adolescent boys (54 per cent) of adolescent boys reported incidents
of physical violence since age 15, and over one

Data regarding adolescent boys’ exposure to physical
violence are limited in comparison to those for girls. third (34 per cent) also said they experienced such
In each of five low- and middle-income countries violence in the past 12 months. The country with
with comparable DHS data, at least one in four the lowest reported prevalence of physical violence
adolescent boys aged 15 to 19 said they experienced among adolescent boys is Mozambique. Even so, 28
physical violence since age 15 (Figure 3.14). The per cent of adolescent boys in that country reported
reported prevalence exceeds 40 per cent in all but incidents of physical violence since age 15 and 13
two of the countries (Ghana and Mozambique) for per cent said they experienced such violence within
which data are available. The highest prevalence of the last year.
52
: © UNICEF/NYHQ2012-0710/Holt
Figure 3.14
In five countries, at least one in four adolescent boys report incidents of physical violence since age 15
Percentage of boys aged 15 to 19 years who experienced any physical violence since age 15 and percentage of boys aged 15 to 19
who experienced any physical violence in the last 12 months
100
Experienced physical violence since age 15
90
Experienced physical violence in the last 12 months
80
70
60
50
40
30
20
10
n/a
0
Uganda Ukraine Cameroon Ghana Mozambique Bolivia (Plurina-
tional State of)
Notes: Data on the proportion of adolescent boys who experienced any physical violence since age 15 are not available for the Plurinational State of Bolivia. Data for Ukraine refer to any physical violence
committed by someone other than the boys’ current spouses or partners.
53
Table 3.4
The most common perpetrators of physical violence against boys are friends, teachers and
‘other’ perpetrators
Percentage of boys aged 15 to 19 years who experienced any physical violence since age 15, by perpetrator
Persons who committed physical violence against boys
Current/former
Mother-in-law
Brother/sister
Daughter/son
Other relative
Police/soldier
Current wife/
acquaintance
Father-in-law
Former wife/
Other in-law
someone at
stepmother
Employer/
stepfather
girlfriend
Mother/
Teacher
partner
partner
Father/
Friend/
Other
work
Bolivia (Plurinational State of) 10 3 - 0 0 - 2 2 0 0 0 22 1 0 - 33
Cameroon 2 1 0.6 18 20 0 22 10 0 0 1 24 21 1 1 29
Ghana 0 0 2 10 11 0 16 17 0 0 0 43 19 0 0 7
Mozambique 0 1 6 18 27 0 27 7 0 0 8 - 29 1 3 13
Uganda 0.2 1 0 10 20 1 10 9 0 0 0 - 34 1 1 46
Note: Data for the Plurinational State of Bolivia refer to physical violence experienced by boys in the last 12 months.
Apart from DHS data, information on physical The experience of being punched, whipped or
violence against boys is also available from the kicked was more commonly reported among males
VACS in Kenya, the United Republic of Tanzania and than being threatened with a weapon.
Zimbabwe. In Kenya and Zimbabwe, 18- to 24-year-
old men were asked about their experiences of
Perpetrators of physical violence against boys
physical violence prior to age 18, using the same
definitions of physical violence referenced earlier for As with girls, boys aged 15 to 19 who experienced
girls. In the United Republic of Tanzania, information physical violence since age 15 were asked to
on the prevalence of childhood physical violence identify the perpetrator, which could include both
against males aged 13 to 24 was reported using the partners and non-partners. The most commonly
same definition outlined previously.78 reported perpetrators varied across the five
countries with available DHS data: friends or
In both Kenya and Zimbabwe, reported rates of acquaintances (Ghana), teachers (Mozambique)
physical violence during childhood among men are and ‘other’ perpetrators (the Plurinational State of
higher than those recorded for women. In Kenya, Bolivia, Cameroon and Uganda) (Table 3.4).
73 per cent of men said they experienced physical
violence prior to age 18 (compared to 66 per cent It was not possible to compare perpetrators based
of women), with most reporting that they were on the marital status of male respondents since
slapped, pushed, punched, kicked, whipped or the number of ever-married boys who reported
beaten with an object. In Zimbabwe, 76 per cent experiencing physical violence since age 15 was
of men aged 18 to 24 were subjected to physical too small to produce reliable estimates. However,
violence prior to age 18 (as opposed to 64 per cent of among never-married boys in Mozambique,
females of the same age). Again, men in Zimbabwe teachers, siblings and fathers/stepfathers were the
were most likely to say they were slapped, pushed, most frequently cited offenders (at 28 per cent, 27
punched or hit with an object. In the United Republic per cent and 25 per cent, respectively). A slightly
of Tanzania, a slightly lower proportion of boys and different pattern emerges in Cameroon, where other
men aged 13 to 24 reported incidents of physical perpetrators (29 per cent), friends (25 per cent),
violence before age 18 than girls and women of the siblings (22 per cent) and teachers (21 per cent)
same age (72 per cent and 74 per cent, respectively). were most often named by never-married boys. In
54
Uganda, other perpetrators were most commonly

cited (47 per cent), followed by teachers (35 per
cent). In Ghana, on the other hand, male friends
were the most frequently reported perpetrators
among never-married boys (43 per cent), followed
by teachers (27 per cent).
Information on the identity of perpetrators of physical

violence against boys and men in childhood was also
collected through the VACS carried out in Kenya, the
United Republic of Tanzania and Zimbabwe. Among
men aged 18 to 24 in Kenya, more than half said
they were punched, kicked, whipped or beaten with
an object before turning 18 by a parent or other
adult relative (57 per cent) or an authority figure (58
per cent). Among the former, the most commonly
cited perpetrators were fathers and mothers, while
among the latter, nearly all (96 per cent) reported it
was a teacher. Less than 5 per cent of respondents
said they were threatened or attacked with a
weapon by either a parent/adult relative or other
authority figure. Reports of physical violence prior to
age 18 at the hands of a current or former intimate
partner were infrequent among men in Kenya.
In the United Republic of Tanzania, boys and men

aged 13 to 24 cited relatives as the most common
perpetrators of childhood physical violence both
among those living on the mainland (57 per cent)
and Zanzibar (63 per cent). A significant proportion
of males living in both areas of the country also
reported incidents of childhood physical violence
at the hands of a teacher (51 per cent from the
mainland and 43 per cent from Zanzibar). Violence
at the hands of intimate partners or authority figures
other than teachers was relatively uncommon in
both areas of the country.
In Zimbabwe, the largest proportions of men

aged 18 to 24 who said they experienced physical
violence before age 18 by a parent or other adult
relative named their father (46 per cent) or mother

© UNICEF/NYHQ2011-2241/Dormino
(43 per cent) as perpetrators. The second most

commonly reported family members to perpetrate
physical violence against boys were uncles and
brothers. Among those who cited authority figures,
95 per cent said they were victimized by teachers.
Slightly less than 2 per cent of adult men reported
being physically violated before age 18 by a current
or former intimate partner.
55
Emotional violence
Emotional violence, also referred to Short- and long-term In both countries, data correspond to
as psychological or mental abuse, in- consequences experiences of emotional violence
cludes scaring, terrorizing, threatening, before age 18 among young adult
Unlike physical abuse, which causes males and females aged 18 to 24.
exploiting, rejecting, isolating, ignoring,
injuries that can be seen on a child’s In the United Republic of Tanzania,
insulting, humiliating and ridiculing a
body, emotional violence is more emotional abuse was defined as being
child. Denying emotional responsive-
subtle and can be difficult for both called bad names, being made to feel
ness, neglecting the health or educa-
the victim and outsiders to identify. unwanted or being threatened with
tional needs of a child, exposing a child
Nonetheless, its consequences can abandonment by an adult. Data from
to domestic violence, and psychologi-
be just as devastating.5 Children who that country represent experiences
cal bullying and hazing are also classi-
experience recurrent episodes of of emotional violence prior to age 18
fied as emotional violence. Extreme
emotional violence often grow up among males and females aged 13 to
types of such violence include sub-
thinking they are deficient in some way. 24. In Swaziland, emotional violence
jecting a child to solitary confinement They may blame themselves for the
or other degrading conditions of de- was measured by asking respondents
abuse, internalizing the negative words whether they had received emotional
tention.1 While distress or emotional and aggressive actions against them.6
harm often result from the experience ill treatment such as name-calling or
Research has shown that childhood
of physical or sexual violence, children saying mean things that made them
exposure to emotional abuse has a
can also be specifically victimized by feel scared or “really bad”. Research
range of long-term effects that can
targeted acts of verbal or psychological here was limited to females only, and
persist well into adulthood, including
aggression. data refer to emotional violence before
depression, anxiety, post-traumatic
age 18 among females aged 13 to 24.
stress, low self-esteem, isolation and
Emotional violence against children
estrangement from other people, Findings from these surveys suggest
is most commonly perpetrated by
insecure attachment and difficulty that the experience of emotional
people with whom they have a close with relationships, among others.7 violence is slightly more common
personal relationship or attachment. Moreover, one study revealed that the among males than females. In
In fact, parents and caretakers are consequences of emotional violence Zimbabwe, 38 per cent of young men
frequently cited as the most common vary according to the form it takes: reported incidents of emotional abuse
offenders.2 According to the Fourth Children who are repeatedly terrorized in childhood versus 29 per cent of
National Incidence Study of Child by their caregivers tend to develop young women. Roughly 30 per cent
Abuse and Neglect in the United anxiety and somatic complaints in of young men in both Kenya and the
States, 73 per cent of all cases of child adulthood, while those who are United Republic of Tanzania were also
emotional abuse that were officially ignored and degraded may develop subjected to emotional violence as
reported in that country from 2005 to borderline personality disorders.8 children versus 26 per cent and 24 per
2006 were perpetrated by a biological cent of young women, respectively.
parent.3 Although research on other In Swaziland, 30 per cent of females
perpetrators of emotional violence has Four studies from Africa
reported experiences of emotional
historically been limited, a few studies The Violence Against Children Surveys abuse prior to age 18.
in several other countries, including (VACS) carried out between 2007 and
Botswana, Canada, Portugal, Sweden, 2011 in Kenya, Swaziland, the United Consistent patterns emerged across
the United Kingdom and Zimbabwe, Republic of Tanzania and Zimbabwe countries in regards to the types of
have begun to explore such violence asked young adults about their violent acts most commonly reported
committed by teachers, peers and experiences of emotionally violent acts by respondents. Among both males
dating partners.4 before the age of 18.9 However, there and females in Kenya and males in
are differences in the ways in which Zimbabwe, humiliation by an adult was
Despite these examples, emotional such violence was measured across the cited most often. Roughly one quarter
violence remains an under-studied countries and in the composition of the of men in both countries reported
Emotional violence
topic, and relevant statistics are sample for which data are presented. being humiliated by an adult, along with
sorely lacking. This can be attributed In Kenya and Zimbabwe, emotional 18 per cent of women in Kenya and
largely to the fact that it is difficult to violence was defined as an adult 13 per cent of women in Zimbabwe.
conceptualize and operationalize all saying or doing something on purpose In the United Republic of Tanzania,
possible manifestations of emotional to humiliate the child in front of others, calling a child bad names was the most
violence and to quantify its myriad making the child feel unwanted and/ commonly reported form of emotional
ramifications. or threatening to abandon the child. abuse, with about one in five males and
56
females reportedly victimized in this among females and 65 per cent among say they had been humiliated by
way. Many respondents of both sexes males). In Swaziland, non-parental teachers and neighbours. Similarly, in
also reported feeling unwanted by an relatives were the most frequently Zimbabwe, humiliation by a teacher or
adult or other caretaker – including cited offenders: 29 per cent of girls uncle was most commonly reported
about 17 per cent in Zimbabwe, 15 and women said they experienced by men, while aunts or mothers
per cent in Kenya and 8 per cent in the emotional abuse by a female relative were most frequently cited among
United Republic of Tanzania. Across all other than a mother or stepmother women. Among Zimbabweans who
countries, threatening abandonment (such as an aunt or grandmother), while reported feeling unwanted, most
was the least reported form of 24 per cent were victimized by a male (close to a third of both females and
emotional violence. Even so, 12 per relative other than a father, stepfather males) pointed to an aunt or uncle of
cent of respondents were threatened or husband. While victimized within the same sex as the source of those
in this way in Zimbabwe, along with their own families, many young people feelings. In Kenya, fathers were most
6 per cent and 9 per cent of female were also subjected to emotional likely to make respondents of both
and male respondents in Kenya, abuse by perpetrators outside the sexes feel unwanted (31 per cent of
respectively. A smaller proportion in home. For instance, even though
males and 24 per cent of females),
the United Republic of Tanzania (5 most respondents in the United
while parents of both sexes and
per cent of males and 4 per cent of Republic of Tanzania were emotionally
aunts were most likely to threaten
females) said they were threatened abused by a relative, many said they
abandonment.10 Males in Kenya were
with abandonment as children. were also victimized by neighbours,
especially vulnerable to threats of
dating partners and other non-family
While respondents in all four countries abandonment by their fathers, with
authority figures. Among those aged
were subjected to emotional abuse over half reporting this particular
13 to 24 in this country, 34 per cent of
from a wide range of caregivers and form of emotional abuse. Women
males and 19 per cent of females who
people in positions of authority, it is in Zimbabwe most often reported
experienced emotional abuse reported
clear that many children are exposed to that a neighbour was the perpetrator. threats of abandonment from their
emotional violence in the family home. mothers, grandparents or uncles,
For example, in the United Republic of Perpetrators also varied according to while the most common perpetrators
Tanzania, the most frequently reported the type of emotionally violent acts of this type of emotional abuse
perpetrator of emotional violence was committed. In Kenya, for example, mentioned by men were fathers,
overwhelmingly a relative (80 per cent young adults were most likely to uncles or mothers.
references
1 United Nations Committee on the Rights of the Child, General Comment across the Life Cycle: What theory and research tell us’, Journal of Ag-
No. 13 (2011): The right of the child to freedom from all forms of violence, gression, Maltreatment and Trauma, vol. 19, no. 1, 2010, pp. 5-51; Riggs,
UN document CRC/C/GC/13, Office of the High Commissioner for Human S. A., and P. Kaminski, ‘Childhood Emotional Abuse, Adult Attachment,
Rights, Geneva, 18 April 2011. and Depression as Predictors of Relational Adjustment and Psychological
Aggression’, Journal of Aggression, Maltreatment and Trauma, vol. 19, no.
2 Lamont, A., Who Abuses Children? Australian Institute of Family Studies,
1, 2010, pp. 75-104; Chapman, D. P., et al., ‘Adverse Childhood Experiences
Melbourne, 2011.
and the Risk of Depressive Disorders in Adulthood’, Journal of Affective
3 Abuse included close confinement (for example, tying or binding), verbal Disorders, vol. 82, no. 2, 2004, pp. 217-225; Spertus, I. L., et al., ‘Childhood
assaults, threats of sexual abuse (without contact), threats of other Emotional Abuse and Neglect as Predictors of Psychological and Physical
maltreatment, terrorizing, administering prescribed substances and other Symptoms in Women Presenting to a Primary Care Practice’, Child Abuse &
unclassified abuse. See: Sedlak, A., et al., Fourth National Incidence Study Neglect, vol. 27, no. 11, 2003, pp. 1247-1258.
of Child Abuse and Neglect (NIS-4): Report to Congress, US Department
of Health and Human Services, Administration for Children and Families, 8 Allen, B., ‘An Analysis of the Impact of Diverse Forms of Childhood Psycho-
Washington, DC, 2010. logical Maltreatment on Emotional Adjustment in Early Adulthood’, Child
Maltreatment, vol. 13, no. 3, 2008, pp. 307-312.
4 Leen, E., et al., ‘Prevalence, Dynamic Risk Factors and the Efficacy of
Primary Interventions for Adolescent Dating Violence: An international 9 United Nations Children’s Fund Kenya Country Office, Division of Violence
review’, Aggression and Violent Behavior, vol. 18, no. 1, 2013, pp. 159-174; Prevention, National Center for Injury Prevention and Control, US Centers
Theoklitou, D., N. Kabitsis and A. Kabitsi, ‘Physical and Emotional Abuse for Disease Control and Prevention, and the Kenya National Bureau of
of Primary School Children by Teachers’, Child Abuse & Neglect, vol. 36, Statistics, Violence against Children in Kenya: Findings from a 2010
no. 1, 2012, pp. 164-170; Shumba, A., ‘The Nature, Extent and Effects of national survey – Summary report on the prevalence of sexual, physical and
Emotional Abuse on Primary School Pupils by Teachers in Zimbabwe’, Child emotional violence, context of sexual violence, and health and behavioral
Abuse & Neglect, vol. 26, no. 8, 2002, pp. 783-791; Osei-Hwedie, K., and A. consequences of violence experienced in childhood, Nairobi, 2012; United
K. Hobona, ‘Secondary School Teachers and the Emotional Abuse of Chil- Nations Children’s Fund, US Centers for Disease Control and Prevention
Emotional violence
dren: A study of three secondary schools in Gaborone, Botswana’, Journal and Muhimbili University of Health and Allied Sciences, Violence against
of Social Development in Africa, vol. 16, no. 1, 2001, pp. 143-163. Children in Tanzania: Findings from a national survey 2009, UNICEF Tanza-
nia, Dar es Salaam, 2011; Zimbabwe National Statistics Agency, United Na-
5 Norman, R. E., et al., ‘The Long-term Health Consequences of Child Physi-
tions Children’s Fund and Collaborating Centre for Operational Research and
cal Abuse, Emotional Abuse, and Neglect: A systematic review and meta-
Evaluation, National Baseline Survey on Life Experiences of Adolescents,
analysis’, PLOS Medicine, vol. 9, no. 11, 2012, p. e1001349.
2011, Harare, 2013; United Nations Children’s Fund Swaziland, A National
6 Iwaniec, D., E. Larkin and S. Higgins, ‘Research Review: Risk and resilience Study on Violence against Children and Young Women in Swaziland,
in cases of emotional abuse’, Child & Family Social Work, vol. 11, no. 1, UNICEF Swaziland, Mbabane, 2007.
2006, pp. 73-82.
10 In Swaziland and the United Republic of Tanzania, data on the specific
7 Riggs, S. A., ‘Childhood Emotional Abuse and the Attachment System perpetrator by type of emotional abuse are not available.
57
Violence against children
in institutional care:
The case of Kazakhstan 1
In countries around the world, public the ages of 7 and 19, structured inter- cational institutions for children with
and private orphanages and other resi- views with institution directors, an ob- antisocial behaviour, 21 per cent in
dential institutions have been estab- servational checklist, and semi-struc- orphanages and 16 per cent in shel-
lished for the care of children without tured interviews with graduates (ages ters. About 19 per cent, 17 per cent
family support. Yet according to some 17 to 23) of state-run institutions. The and 13 per cent of children, respec-
studies, rates of violence against chil- study involved 30 institutions in three tively, said they wore dirty or torn
dren in care institutions are six times regions of the country, representing clothes or clothing that was not
higher than those of children living in six different types of facilities.3 Al- appropriate for the weather or was
family-based foster care.2 Institutional- though the findings do not claim to be wrongly sized. Medical neglect was
ized children may be subjected to vio- representative of all state-run residen- particularly common among children
lence by their peers, particularly when tial institutions for children in Kazakh- in educational institutions, where
conditions are poor and proper super- stan, they do provide a general indica- 12 per cent of children said they
vision is lacking, or even victimized tion of the nature and prevalence of did not receive adequate care when
by the very staff and officials charged violence against children within such sick. Roughly 5 to 6 per cent of chil-
with their safekeeping. establishments. The study measured dren indicated they did not receive
the following forms of violence: bully- enough food to eat in all three kinds
Central and Eastern Europe and the ing, harsh verbal abuse, psychological of institutions. Lack of supervision
Commonwealth of Independent States abuse, physical violence and neglect was the least commonly reported
has the highest prevalence of insti- (including lack of adequate food, cloth- form of neglect, with 2 per cent of
tutional care of children in the world, ing, medical care and supervision).4 children in each kind of institution
and Kazakhstan has one of the highest reporting being locked in his or her
rates per capita. Historically, state-run • Of the 997 girls and boys surveyed room at night without adult supervi-
institutions in that country encouraged in educational institutions for chil- sion.
families facing difficulties to give their dren with antisocial behaviour, or-
children up to residential care – even if phanages and shelters, 14 per cent, • Surveys of 284 staff working in in-
one or both parents were still alive. It 7 per cent and 5 per cent, respec- fant homes (housing children under
was a solution deemed reasonable by tively, said they were afraid of other age 5) revealed that over 18 per cent
all parties concerned, except perhaps children in the institution. Similarly, were witnesses to staff members
the children themselves. As a result, 13 per cent, 6 per cent and 7 per using moderate or severe physical
the majority of children in Kazakhstan cent, respectively, said they were violence on children. Moreover, 62
Violence against children in institutional care
with physical or mental disabilities are afraid of the staff. As many as one per cent of staff in infant homes said
in state-run institutions, often because in two children in these institutions they did not know if their institutions
of stigmatization or because parents reportedly witnessed incidents of had an official written document reg-
lack the resources or capacity to care violence between child residents. ulating the conduct of staff.
for them. In 2012, around 30,000 of In orphanages and educational in-
the country’s 5.3 million children were stitutions, more than one in three • Among 349 staff working in spe-
living in state and non-state institutions children witnessed staff using some cial correctional institutions or with
for orphans and children without pa- form of violence against children; disabled children, as many as one
rental care. one in four witnessed incidents of in three witnessed bullying and well
severe violence by staff. Among over 50 per cent witnessed physical
In 2011, a study carried out by the Na- the 997 children surveyed, 11 per violence between child residents.
tional Centre for Human Rights and cent said they had run away from Over 50 per cent of staff also ob-
UNICEF Kazakhstan revealed, for the their respective institutions at least served staff members using differ-
first time, the extent of violence suf- once and 9 per cent said they had ent forms of violence against the
fered by children in the country’s state- purposely hurt themselves because children. In institutions for children
run residential institutions. The study they were unhappy. with psycho-neurological and severe
design employed multiple methods, disabilities, nearly 23 per cent of
including anonymous surveys of staff Some form of neglect was also re- staff reported witnessing incidents
and institutionalized youth between ported by 26 per cent of those in edu- of harsh physical violence by staff
58
against residents; nearly 54 per cent children tend to have higher rates of grow up in a family environment un-
of staff in these institutions held alcohol and drug use than non-institu- less it is in the child’s best interests
positive views regarding the use of tionalized children and are more likely to do otherwise. The increased risk
corporal punishment. to come into conflict with the law.6 of violence against children while in
Both girls and boys who have resided institutional care adds to the State’s
Studies worldwide have consistently in institutions are also more vulnerable obligations to ensure that effective leg-
shown the negative impact of institu- to violence and exploitation, including islation and other measures are taken
tionalization on children, ranging from intimate partner violence, sexual ex- to protect these children. The Conven-
compromised physical health and at- ploitation and human trafficking.7 tion specifically addresses the rights of
tachment disorders to developmental children with disabilities, saying that
delays and potentially irreversible psy- The Convention on the Rights of the segregation and the institutionalization
chological damage.5 Institutionalized Child recognizes that children should of such children cannot be justified.
references
1 Unless otherwise noted, the source for the information in neglect entailed locking a child in his or her room at night
this box is: Haarr, R. N., Violence against Children in State- without adult supervision. Bullying referred to instances
run Residential Institutions in Kazakhstan: An assessment, in which a child perpetrated repeated acts of harsh verbal
Commissioner for Human Rights in the Republic of Kazakh- abuse, psychological abuse and/or physical violence against
stan and the United Nations Children’s Fund, Kazakhstan, other children. According to the study report, specific ques-
2011. tions were not asked of children or staff about sexual abuse
due to taboos surrounding the subject.
2 Barth, R. P., Institutions vs Foster Homes: The empirical
base for a century of action, University of North Carolina 5 Zeanah, C., et al., ‘Institutional Rearing and Psychiatric Dis-
School of Social Work, Jordan Institute for Families, Chapel orders in Romanian Preschool Children’, American Journal
Hill, 2002, cited in: Pinheiro, P. S., World Report on Violence of Psychiatry, vol. 166, no. 7, 2009, pp. 777-785; Johnson,
against Children, UN Secretary-General’s Study on Violence R., K. Browne and C. Hamilton-Giachritsis, ‘Young Children
against Children, United Nations, Geneva, 2006, p. 183. in Institutional Care at Risk of Harm’, Trauma Violence &
Abuse, vol. 7, no. 1, 2006, pp. 34-60; Roy, P., M. Rutter
3 The types of facilities included infant homes, orphanages, and A. Pickles, ‘Institutional Care: Associations between
shelters, specialized institutions of education for children overactivity and lack of selectivity in social relationships,
with ‘deviant’ behaviour, institutions for children with Journal of Child Psychology and Psychiatry, vol. 45, no. 4,
psycho-neurological and severe disabilities, and youth 2004, pp. 866-873.
homes.
6 Browne, K., The Risk of Harm to Young Children in Insti-
4 Acts of psychological abuse included breaking or ruining a tutionalized Care, Save the Children UK and Better Care
child’s things on purpose, acting in a way that made a child Network, London, 2009; Ryan, J. P., et al., ‘Juvenile Delin-
afraid he or she might be physically hurt/injured, threaten-
Violence against children in institutional care

quency in Child Welfare: Investigating group home effects’,
ing to physically harm or hit a child, locking a child in a room Children and Youth Services Review, vol. 30, no. 9, 2008,
or small place for a long time, tying a child up or chaining pp. 1088-1099; Harwin, J., Children of the Russian State:
him/her, preventing a child from using the toilet and giving 1917-95, Ashgate Publishing Company, Avebury, 1996.
a child onerous tasks around the institution. Forms of harsh
verbal abuse included swearing at, cursing or calling a child 7 Browne, op. cit.; Carter, R., Family Matters: A study of
names and saying mean things to hurt a child’s feelings. institutional childcare in Central and Eastern Europe and the
Acts of physical violence included twisting a child’s ear or former Soviet Union, EveryChild, London, 2005; Reilly, T.,
arm, pinching a child, shaking a child, slapping a child in the ‘Transition from Care: Status and outcomes of youth who
face or on the head, buttocks, back, leg or arm, throwing or age out of foster care’, Child Welfare, vol. 82, no. 6, 2003,
knocking a child down, pushing or grabbing a child, hitting, pp. 727-746.
kicking or physically injuring a child, hitting a child hard
enough to create marks, hitting or attacking a child with a
hard object or weapon and burning a child with hot items.
Neglect covered four areas: clothing, medical care, nutrition
and supervision. Clothing neglect included making a child
wear dirty or torn clothes or clothing that was not appropri-
ate for the weather or that was the wrong size. Medical
neglect involved not taking care of a child when sick (such
as not taking him or her to the doctor or not providing medi-
cine). Nutritional neglect meant not giving a child enough
food to eat so that she or he went hungry. Supervision
59
Chapter 4 Sexual
violence:
Not limited
to girls
Sexual violence is one of the most unsettling of
children’s rights violations. As such, it is the subject
of dedicated international legal instruments aimed
at protecting children against its multiple forms (see
Box 4.1). Acts of sexual violence, which often occur
together and with other forms of violence, range
from direct physical contact to unwanted exposure
to sexual language and images. Even when not
accompanied by physical force or restraint, the sexual
victimization of children resulting from emotional
and psychological manipulation, intimidation and
verbal threats, deception or entrapment can be
equally intrusive and traumatic.1
The experience of sexual violence encompasses

situations in which a child is forced to perform a
sexual act by a caregiver or neighbour, pressed to
have unwanted sexual intercourse by a dating partner,
exposed to sexual comments or advances by a peer
or an adult, impelled to engage in sex in exchange
for cash, gifts or favours, coerced to expose her or
his sexual body parts, including in person or online,
subjected to viewing sexual activities or sexual body
parts without his or her consent, or raped by a group
of persons as part of a ritual, a form of punishment
or the cruelty of war.2 Indeed, ‘sexual violence’ is
often used as an umbrella term to cover all types

© UNICEF/IDSA2010-00028/Purnomo
of sexual victimization, including exploitative as well

as non-exploitative forms.3 Sexual abuse becomes
exploitative when money or other material or non-
material gains are given or promised in exchange
for sexual activities – irrespective of whether they
occur with occasional partners or as part of a stable
relationship.
60
Box 4.1
International legal frameworks

protecting children from sexual
abuse and sexual exploitation
Children have the right to be pro- tion of a child victim of any form of International Labour Organization
tected from all forms of violence, in- neglect, exploitation or abuse, and Convention 182 on the Worst Forms
cluding sexual abuse and exploita- stipulates that such recovery and of Child Labour (1999); the Protocol
tion, under articles 19, 34, 35 and 39 reintegration shall take place in an to Prevent, Suppress and Punish
of the Convention on the Rights of environment that fosters the health, Trafficking in Persons, especially
the Child (CRC). Article 34 of the CRC self-respect and dignity of the child. Women and Children, supplement-
specifically protects children from ing the United Nations Convention
The Optional Protocol to the Con-
sexual abuse and sexual exploita- against Transnational Organized
vention on the Rights of the Child
tion, including the inducement or Crime (2000); the African Charter
on the sale of children, child pros-
coercion of a child to engage in any on the Rights and Welfare of the
titution and child pornography was
unlawful sexual activity; the exploit- Child (1990); and the Council of
adopted in 2000 and entered into
ative use of children in prostitution Europe Convention on Protec-
force in 2002.4 It further refines the
or other unlawful sexual practices; tion of Children against Sexual
protections offered under the CRC
and the exploitative use of children Exploitation and Sexual Abuse
by providing detailed definitions of
in pornographic performances and (2007). The latter, although a regional
the sale of children, child prostitu-
materials. Additionally, article 35 instrument, is open to accession by
tion and child pornography, and
provides protection against the non-Member States of the Council
requires States Parties to criminal-
abduction, sale or trafficking of of Europe. It is the first human rights
ize these child rights violations as
children for any purpose or in any offences, and to provide adequate treaty that provides a definition
form, including sexual exploitation. support services to child victims. of sexual abuse and, among other
Article 39 requires States Parties things, protects children from being
to take all appropriate measures to Other international conventions solicited for sexual purposes, a
promote physical and psychologi- protecting children from sexual process more commonly known as
cal recovery and social reintegra- exploitation and abuse include the ‘grooming’.5
The particular vulnerability children may be particularly vulnerable to abuse by

of children adults or other caregivers they trust and on whom
they depend – at home or in other care settings
No one factor can adequately explain why children
where they are likely to spend a majority of their
are at risk of sexual victimization, although it is
time. Adolescents, on the other hand, may be more
CHAPTER 4: SEXUAL VIOLENCE
clear they are more vulnerable to abuse due to their prone to victimization outside the home through
age.6 In general, children have less experience, exposure to both strangers and peers, the latter
knowledge, maturity and, in many cases, physical within the context of both friendship and intimate
strength than adults, which puts them at increased relationships.7
risk of being perceived as easy targets. Although
children of every age are vulnerable, the specific Children’s perceptions of what constitutes abuse
risks they may be exposed to are likely to vary and their ability to comprehend the experience are
across developmental stages. For instance, younger also likely to be influenced by their age and evolving
61
capacities. For example, younger children may be Researchers have also found a connection between
especially susceptible to manipulation, coercion and early experiences of sexual violence and adverse
‘grooming’8 by older peers and adults since they are behavioural outcomes. Common coping strategies
probably unaware of perpetrators’ motives or the adopted by children, such as running away from
nature of the acts experienced.9 home, skipping school, engaging in risky sex and
taking drugs or alcohol, can isolate them socially
The far-reaching consequences and put them at further risk of future abuse or
of sexual violence exploitation.20 Children who have experienced
sexual violence are also more likely to be arrested,
Experiences of sexual violence in childhood hinder including in adulthood.21 In addition, research has
all aspects of development: physical, psychological found that girls who have been sexually abused
and social.10 Apart from the physical injuries that are at higher risk of experiencing intimate partner
can result, exposure to HIV and other sexually violence22 and of being involved, or exploited, in
transmitted infections, along with early pregnancy, sex work later in life.23 It is commonly thought that
are also possible outcomes.11 Other physical children who are sexually abused are more likely
consequences of sexual violence include a range of to become offenders themselves once they get
self-harming behaviours, such as the development older. However, most research suggests that only a
of eating disorders, like bulimia and anorexia.12 small proportion of those children who are sexually
Children who have been abused are also more likely abused go on to sexually violate others.24
to attempt suicide; the more severe the violence,
the greater the risk.13 Children who are abused sexually are also at
heightened risk of being re-victimized, both while
Researchers have consistently found that the sexual they are still children and as adults.25 Moreover,
abuse of children is associated with a wide array of experiences of early sexual violence often increase
mental health consequences, including symptoms the risk of being subjected to other types of
of depression and panic disorder.14 Anxiety and violence26 and diminish children’s ability to resist
nightmares are also commonly observed in younger unwanted sexual advances in the future.27 Children
children who have experienced such violence.15 who experience multiple forms of violence face the
The psychological impact of sexual violence can poorest outcomes, exhibiting heightened risk of
be severe due to the shame, secrecy and stigma complex traumatic stress disorders, higher levels of
that tend to accompany it, with child victims often depression, lower self-esteem, higher involvement
having to find ways to cope in isolation.16 The risk in life-threatening and risky behaviours, increases
of developing adverse mental health outcomes has in suicidal thoughts and attempts, more frequent
been found to increase in relation to the frequency delinquent conduct and increased use of alcohol and
and severity of children’s exposure to sexual violence drugs.28
and to exert a lasting impact.17
What remains hidden in most
In addition to physical and psychological data collection efforts
consequences, childhood experiences of sexual Although estimates can be found of the number
violence result in considerable social harm.
of children who have experienced sexual violence,

Particularly when violated by a caregiver or one of the biggest challenges in this field is
trusted person, children may develop insecure or underreporting, which stymies efforts to generate
disorganized attachments to others and may face accurate statistics. In fact, some research has shown
difficulty building and sustaining relationships later in that between 30 and 80 per cent of victims do not
life.18 Children who are subjected to sexual violence disclose experiences of childhood sexual abuse
may experience heightened levels of fear and until adulthood,29 while many others (a number
arousal and feel an intensified perception of threat impossible to quantify) remain silent for their entire
or hostility from other people.19 lives (see Box 4.2).
62
Some studies have suggested that boys are even however, certain experiences of violence are likely to
less likely than girls to report incidents of sexual evade any data collection efforts. For these reasons,
abuse.30 It can be particularly difficult for boys to any estimates at the global, regional and country
report and seek help for fear of being viewed as level of the prevalence of sexual violence should
vulnerable or helpless, notions that run contrary to be interpreted with extreme caution and assumed
a common cultural definition of men as being strong to be an underestimate of the actual number of
and self-reliant.31 In addition, boys may be reluctant children victimized.
to admit they have been victims of sexual abuse,
particularly if the offender is male, for fear of being The availability of reliable, comprehensive and
labelled as homosexual.32 comparable data on child sexual abuse also suffers
from variations in definitions, study designs,
Even when victims find the strength to accurately samples and questions used to elicit information. A
report what occurred, other challenges may affect number of systematic reviews and meta-analyses of
the availability of comprehensive and reliable self-reported data from hundreds of surveys across
statistics. For example, figures obtained from several countries reveal a very mixed picture on
child protection services and agency reports, or the extent of the problem.38 In their review of the
data gathered during humanitarian emergencies prevalence of child sexual abuse in 55 studies from
in designated sentinel sites or through rapid 24 countries, Barth and colleagues found figures
assessment exercises, rely on selected samples of ranging from 8 per cent to 31 per cent among
informants that are successfully reached or that are females and from 3 per cent to 17 per cent among
independently able to contact available services.33 males.39 Other studies found lifetime prevalence
of sexual violence among girls and boys to range
Population-based surveys that rely on random from 0 per cent to more than 50 per cent, even
samples and use properly designed and validated across studies from the same country or region;40
questionnaires are uncovering violence that is not higher rates were obtained in studies using more
officially reported. A recent and comprehensive comprehensive definitions of violence and more
meta-analysis conducted by Stoltenborgh and detailed questionnaires.41
colleagues showed rates of sexual abuse to be more
than 30 times higher in studies relying on self-reports
Data sources
than in official reports, such as those based on data
from child protection services and the police.34 Comparable data on specific forms of sexual violence
That said, population surveys also face the risk of are available for a number of low- and middle-income
underreporting, depending on the approach used countries through Demographic and Health Surveys
to gather data. For instance, substantial differences (DHS). Questions on experiences of sexual violence
were found in levels of child sexual abuse when are posed to girls and women of reproductive age
respondents were questioned about the issue in (15 to 49 years) and to boys and men aged 15 and
face-to-face interviews versus anonymous, self- older (depending on the survey, men up to the age
administered questionnaires.35 Similarly, dedicated of 49, 54, 59, 60 or 64 are included). To measure
surveys on violence that involve specialized sexual violence committed by a partner, girls and
training of interviewers and employ well-designed women and boys and men are asked if their spouse/
questionnaires that enable interviewers to establish a partner committed any of the following acts:
rapport with respondents have demonstrated better physically forced her/him to have sexual intercourse
results in facilitating the disclosure of victimization when she/he did not want to; physically forced her/
experiences.36 Finally, qualitative approaches based him to perform any other sexual acts when she/he
on non-structured interviews, particularly when did not want to; and forced her/him with threats or
administered in repeated sessions with trained in any other way to perform sexual acts when she/
personnel, have demonstrated an increased ability he did not want to.42 This question is also asked in
to capture victimization experiences not recorded reference to experiences that occurred in the 12
through regular surveys.37 Even in these cases, months preceding the survey.
63
Box 4.2
Interviewing children about sexual

violence: Ethical and methodological
considerations
Gathering data on violence, to collect recent information on the fear of the perpetrator.48 According
particularly the most severe forms, exposure to risks and on victims’ to some studies, the typical chain
always carries the risk of potential access to services. The assumption of events is as follows: The child
harm, and the overriding need is that interviewing adults (even feels compelled to keep the sexual
is to protect respondents from young adults) on a situation that abuse a secret, often due to feelings
unnecessary danger. While these may have happened in the distant of helplessness and fear of not
concerns apply to reporting on past might be less relevant since being believed. If the child does
all forms of violence, they are risk patterns can change over time. disclose the incident(s), the failure
particularly relevant in cases of Additionally, the availability of of family members or professionals
sexual abuse, given its highly services for referrals and follow-up to protect and support him or her
sensitive nature. Moreover, the are also likely to have changed. adds to the child’s distress and may
dynamics of sexual abuse among lead to retraction of the disclosure.49
young children differ from those While the need for recent data is
Research also suggests that children
of adolescent and adult sexual certainly justifiable, there is little
who are sexually abused by a close
abuse in ways that may affect cross-cultural evidence on whether
family member are particularly
the child’s ability and willingness obtaining information from children
hesitant to reveal the abuse.50
to disclose his/her experience on their experience of violence is
Finally, children who experience the
during interviews.43 For example, more reliable than asking young
most severe forms of sexual abuse
in cases of sexual abuse among adults retrospectively. The issue
and those who are subjected to
young children, the perpetrator is here relates both to reporting bias
repeated victimization over a longer
often known to the child and is a (that is, whether truthful disclosure
period are the least likely to disclose
trusted caregiver. Rather than using is more likely among children
their experience.51
physical force, the perpetrator or young adults) and recall bias
tends to exploit the child’s trust and, (whether children are more likely to
For these reasons, the disclosure
over a period of time, cultivate a accurately remember and recount
of child sexual abuse is better seen
relationship that gradually becomes events that have happened in the
as a process rather than a single
sexual and more aggressive.44 recent past as opposed to young
event,52 requiring the use of special
adults, who may have experienced
age-sensitive questionnaires and
Given the potential risks and some the violence years before the
interviewing techniques by trained
of the circumstances known to be interview).45 Obvious limitations
personnel to facilitate disclosure.53
associated with experiences of in interviewing younger children
Critical questions that arise include
sexual abuse, a key issue is whether about situations of abuse are their
how to collect data on child sexual
the involvement of young children ability to engage in conversation,
abuse within the context of general
is necessary or even appropriate comprehend what is being asked of
surveys based on structured inter-
in research on the subject. In other them and respond appropriately.46
views between children and generic
words, is it possible to obtain And, even once they are able to
reliable information on violence in report, children have been found to interviewers and whether this re-
childhood by interviewing young underreport or report with delays.47 search is truly in the best interests
adults about their past experiences? of the children participating. Further
The literature suggests, for instance, research is urgently needed to build
One of the arguments used to justify that children who are victims of a strong evidence base to support
the need for direct interviews with sexual abuse often do not disclose methodologies and choices related
children revolves around the need the incident immediately due to to these ethical issues.
64
Data from those who reported that they experienced representative household surveys from several
various forms of sexual violence at the hands of other countries, particularly from Latin America and
their current or most recent spouse or partner are the Caribbean and the Pacific. Additional findings
then combined with information collected from both from selected low- and middle-income with non-
ever-married and never-married girls and women/ comparable data are provided at the end of the
boys and men on whether, at any time in their chapter (see Box 4.6).
lives (as children or adults), anyone ever forced
them – physically or in any other way – to have Although the chapter focuses on low- and middle-
sexual intercourse or to perform any other sexual income countries, this should in no way be
acts against their will. This question is also asked interpreted to suggest that sexual violence is not
in reference to experiences that occurred in the also found in high-income nations. In fact, data on
12 months preceding the survey. Girls and women sexual violence against children is likely to exist
and boys and men who responded ‘yes’ to either of for many high-income countries, since most of
these questions are then asked how old they were them routinely carry out national surveys of crime
the first time this happened and their relationship victimization (and typically have well-established and
to the perpetrator (father/stepfather, friend, other functioning systems for police-recorded incidents).
relative, in-law, teacher, current/former partner, However, the age groups covered, types of sexual
employer/co-worker, police/soldier, stranger, etc.). violence measured and methodologies employed for
The violence may have been experienced either data collection vary widely and render cross-country
repeatedly or as an isolated event. The prevalence of comparisons problematic. Data on sexual violence
any sexual violence, committed by anyone, is then from a selection of high-income countries are also
reported for all girls and women and boys and men, included at the end of the chapter (see Box 4.7).
regardless of their marital status.
Child sexual exploitation is minimally covered in this
While the same questionnaire is used in DHS to chapter since statistics are extremely limited. This is
collect information on the experiences of sexual largely due to the fact that these activities, by their
violence among both sexes, data for males are nature, are kept hidden, and it is therefore extremely
available for a significantly smaller number of difficult to capture the true magnitude of the problem
countries. This paucity of data can be explained by (see Box 4.3). In addition, most available data are
the assumption that sexual violence against boys is based on small-scale anecdotal studies and surveys
relatively uncommon and may, in turn, have fuelled that cannot be considered representative and that
the erroneous perception that boys are immune pose unique considerations with regards to both data
from it. While it is fairly well recognized that, in collection and interpretation. Findings from selected
most parts of the world, girls are at increased risk surveys that have attempted to collect prevalence
of sexual violence, boys are victimized as well data on sexual exploitation are presented in the last
– but their experiences of abuse remain largely section of this chapter.
undocumented.54
Sexual violence against girls
Given current data availability, the information Comparable data from 40 low- and middle-income
presented in this chapter covers a few specific countries confirm that exposure to certain forms
forms of sexual violence: forced sexual intercourse of sexual violence is not uncommon in the lives
(or rape) and other types of sexual activities in which of many girls. The proportion of those aged 15
a child has been forced to participate against her to 19 who have ever experienced forced sexual
or his will, with an emphasis on girls’ experiences. intercourse or other sexual acts in their lifetime varies
Apart from the DHS, other sources of data used widely across countries, ranging from no reported
here include the Violence Against Children Surveys experiences among adolescent girls in Kyrgyzstan
(VACS) in Kenya, Swaziland, the United Republic to 22 per cent among such girls in Cameroon
of Tanzania and Zimbabwe,55 along with nationally (Figure 4.1).
65
Box 4.3
Methodological challenges in
collecting data on the sexual
exploitation of children
Reliable statistics on the sexual ex- Collecting reliable data on child providers, but while important in
ploitation of children are exception- sexual exploitation is challenging other respects, these data cannot be
ally difficult to obtain and are there- for many reasons. Data on the is- generalized to the broader popula-
fore extremely limited in coverage sue are rarely captured in generic tion of sexually exploited children.58
and scope. While global estimates household surveys. Moreover,
on the prevalence of child sexual given the underground nature of Creating a strong evidence base on
exploitation can be found, there is certain forms of sexual exploitation, child sexual exploitation will require
conceptual clarity and consistency
considerable uncertainty about how many children who are victims nev-
in the definition(s) used. It will also
such estimates were calculated.56 er have the opportunity to be identi-
mean integrating validated ques-
Many reports do not include a clear fied. Children who are exploited in
tions into dedicated data collection
explanation of the underlying meth- sexual activities for remuneration or
efforts relying on sampling meth-
odology and studies often differ in any other form of compensation –
ods designed to capture hard-to-
the definitions used, calling into whether in an established setting
reach populations.
question the comparability of the such as a brothel or in an informal
findings. Furthermore, available es- context – often do so in conditions of
timates often combine women and secrecy and illegality. Many studies
children in one category.57 rely on data collected from service
Significant regional variations can also be observed. reported victimization in the last year. The highest
In all of the countries of CEE/CIS with comparable rates of past-year victimization are found in the
data except the Republic of Moldova, less than 1 Democratic Republic of the Congo (10 per cent),
per cent of adolescent girls reported sexual abuse Uganda (9 per cent) and Equatorial Guinea and
at some point in their lives. In 13 of the 18 countries Zambia (8 per cent each).
in sub-Saharan Africa with available data, prevalence
rates of 10 per cent or more are found. This includes As mentioned earlier, there are many kinds of sex-
both conflict and non-conflict countries (see Box ual victimization. These include forced penetration
4.4). At least one in eight adolescent girls in all West and other forced sexual acts, such as those covered
and Central African countries except two (Nigeria in the DHS, as well as sexual touching or fondling.
and Sao Tome and Principe) with available data They also include non-contact sexual abuse, such
reported experiences of forced sexual intercourse as verbal or written sexual harassment or threats or
or other sexual acts at some point during their lives. online sexual victimization (see Box 4.5). Surveys
Prevalence rates of sexual violence are above 10 that ask respondents whether they have experi-
per cent in all countries of Eastern and Southern enced a diverse range of acts will naturally generate
Africa with available data, except for Comoros and higher overall prevalence rates since they will be
Mozambique. capturing a wider array of sexually violent acts. On
the other hand, surveys that restrict questions to
When it comes to recent exposure to sexual the most violent forms of sexual violence, such as
violence, up to 12 per cent of adolescent girls forced intercourse, will produce lower prevalence
aged 15 to 19 in all countries with available data rates since these are not likely to be reported as
66
often as other forms, such as sex- Figure 4.1

ual touching or unwanted sexual
advances. In addition, some surveys Forced sexual intercourse and other forms of sexual
only inquire about acts that have coercion are not uncommon in the lives of many girls
been completed while others might
Percentage of girls aged 15 to 19 years (or otherwise noted) who ever expe-
also ask about, and include in prev-
rienced forced sexual intercourse or any other forced sexual acts (including in
alence rates, situations in which a childhood), and percentage of girls aged 15 to 19 years (or otherwise noted)
particular act was attempted but who experienced forced sexual intercourse or any other forced sexual acts in
the last 12 months
not completed. This would also
affect overall prevalence levels. Cameroon
Uganda
For instance, in the 2007 Violence Zimbabwe
Against Children Survey (VACS) car- Malawi
ried out in Swaziland,59 about one in Equatorial Guinea
Ghana
three girls and women aged 13 to 24 Zambia
reported experiencing at least one Gabon
of the following incidents before age United Republic of Tanzania
Liberia
18: unwanted sexual touching (such Rwanda n/a
as kissing, grabbing or fondling), at- Jamaica n /a
tempted unwanted intercourse and Kenya
Nepal
physically forced intercourse or co-
Haiti
erced intercourse (in which a man or Mozambique
boy persuaded or pressured the re- Sao Tome and Principe Experienced sexual violence ever
Dominican Republic
spondent to have sexual intercourse Experienced sexual violence in the last 12 months
Republic of Moldova
against her will). Of these four Nicaragua n /a
forms of sexual violence, the most El Salvador n /a
Honduras
commonly reported was attempted
Ecuador n /a
unwanted intercourse. Nigeria
Philippines
In the 2010 DHS conducted in the Bolivia (Plurinational State of) n /a
Comoros
United Republic of Tanzania, which India
was restricted to forced sex or other Colombia n /a
forced sexual acts, the prevalence Guatemala n /a

Timor-Leste
rate in the previous 12 months was Paraguay n /a
5 per cent among girls aged 15 Kazakhstan
to 19. The 2009 VACS60 gathered Tajikistan
Azerbaijan
information on additional forms
Ukraine
of sexual violence (such as Cambodia
attempted unwanted intercourse Kyrgyzstan
n /a
Côte d’Ivoire
and unwanted sexual touch-
0 10 20 30 40 50
ing), which resulted in a higher

proportion (14 per cent) of girls Notes: Data on the proportions of girls who experienced forced sexual intercourse or any other forced sexual acts
in the last 12 months are only available for a selection of countries. Data for the Democratic Republic of the Congo
aged 13 to 17 who reported at refer to girls aged 18 to 19 years who experienced only forced sexual intercourse. Data for Côte d’Ivoire refer only to
ever-married girls aged 15 to 19 years; there are no lifetime prevalence data available for the country. Data for the
least one incident in the year Plurinational State of Bolivia and Ecuador include only forced sexual intercourse. In El Salvador, Guatemala, Nicaragua
preceding the survey.61 Among and Paraguay, sexual violence committed by a spouse or partner among ever-married girls included forced sexual
intercourse or agreeing to have sexual intercourse when they did not want to for fear of what their partner might
the different types of sexual vio- do; sexual violence committed by anyone among all girls and women included only forced sexual intercourse. Data
for Colombia include only girls raped by someone other than a spouse or partner. Data for Jamaica refer to girls and
lence explored in the survey, the women aged 15 to 24 years who experienced only forced sexual intercourse. Zeroes appearing in the figure do not
most commonly reported form was necessarily mean that there were no victims of sexual violence in these countries but rather that the estimates came
to 0 after rounding.
unwanted sexual touching. Source: UNICEF global databases, 2014, based on DHS, MICS and other nationally representative surveys, 2004-2013.
67
Similarly, the 2008-2009 DHS in Kenya found that certain forms of sexual violence in the 12 months
about 4 per cent of girls aged 15 to 19 experienced prior to the survey. In the subsequent 2011 VACS,63
forced sex or other forced sexual acts in the last 12 9 per cent of girls aged 13 to 17 reported forced or
months. In the 2010 VACS in Kenya,62 11 per cent coerced sex, attempted unwanted sex or unwanted
of girls aged 13 to 17 said they experienced forced sexual touching at least once during the same
or coerced sex, attempted sex or unwanted sexual time frame. As in Kenya and the United Republic
touching at least once in the previous year. Again, of Tanzania, unwanted sexual touching was the
unwanted sexual touching was most commonly most commonly reported form of sexual violence.
reported. These findings illustrate how prevalence rates of
sexual violence are affected by both the diversity
Finally, the 2010-2011 DHS in Zimbabwe found that and number of specific sexual acts asked about in
about 7 per cent of girls aged 15 to 19 experienced a questionnaire.
Box 4.4
Sexual violence in emergencies

Representative and robust data on crowded and lack privacy as well as the Congo on the extent of conflict-
sexual violence against children safe spaces for children.68 related sexual violence range from
in the context of emergencies are 18 per cent to 40 per cent among
challenging to gather. Neverthe- In addition to conflict-specific victim- girls and women and 4 per cent to
less, it is well documented that the ization, certain forms of violence – 24 per cent among boys and men,
risk of exposure to certain forms such as intimate partner violence and depending on the approach used to
of sexual victimization intensifies other types of abuse against children gather the data.73
during civil unrest and armed con- (including abuse of a sexual nature)
flict.64 In fact, rape has been used in – can become more pronounced dur- Prevalence rates on sexual violence
some contexts as a weapon of war, ing times of war.69 This is most likely during emergencies are not only af-
with children and women most of- the result of heightened stress, the fected by a number of methodologi-
ten the targets.65 Adolescent girls breakdown of social structures and cal issues; they also differ due to the
are particularly vulnerable and, weakened judicial and law enforce- varying scope and nature of such
in some cases, are abducted and ment systems. A study conducted in violence in different conflicts and in
used for sexual purposes by armed eastern Democratic Republic of the the periods before, during and after
groups.66 Congo on a hospital-based cohort an active conflict or crisis. Findings
of care-seekers for alleged sexual from a survey carried out in 2008 in
During emergencies, children violence reported cases of milita- selected rural communities of Côte
are also more likely to become rized rape of children and youth, d’Ivoire support this view.74 The
separated from their families and with similar features of cruelty and reported prevalence of being sub-
caregivers, thus increasing their extreme violence as the incidents jected to forced sex by someone
risk of certain forms of violence. reported among adults.70 However, other than an intimate partner since
Evidence suggests that unaccom- the study also indicated that the ma- the age of 15 was higher before and
panied and separated children, as jority of perpetrators of child sexual during the active crisis than it was in
well as internally displaced and abuse were described as civilians (81 the 12 months following the end of
refugee children, are particularly per cent) and known to the family (74 the conflict among both females and
susceptible to sexual victimization, per cent).71 males aged 15 to 49. Furthermore,
exploitation and trafficking.67 Al- rates of forced sex by an intimate
though camps can shelter children Due to the increased risks of victimi- partner among ever-partnered girls
from several risks, they can also zation by both civilians and military and women and were higher (29 per
create conditions in which children perpetrators, certain conflicts have cent) before and during the crisis.
become vulnerable to sexual vio- been associated with large-scale in- Rates were lower, although still high
lence by a different group of per- cidents of sexual violence.72 Reports (15 per cent) even within a year
petrators since they are often over- from the Democratic Republic of afterwards.
68
© UNICEF/NYHQ2012-0711/Holt
69
Box 4.5
Online sexual abuse and exploitation

of children
Worldwide, children are becoming communication can also escalate keep an online connection with as
increasingly avid users – and in- into in-person sexual abuse and/or many as 200 youths at a time, all of
novators – of the Internet, typically exploitation if children meet some- whom may be in different stages of
surpassing their parents and other one they have been communicating the grooming process.82
adults in their knowledge and skills. with online.78 Below is an overview
While this exposure can provide of a few common forms of child Evidence suggests that children,
novel opportunities for learning and sexual abuse that can occur online. especially girls, are at the greatest
social interaction, it also presents risk of being groomed online.83 Re-
new dangers. Online grooming search has also shown that girls in
their late adolescent years may be
Research from a number of coun- Solicitation of children for sexual more likely than younger girls to
tries has found that young people purposes, or grooming, is defined engage in risky behaviour with peo-
frequently report feeling safer shar- in article 23 of the Council of Europe ple they meet on the Internet. One
ing private and sensitive informa- Convention on Protection of Chil- study, for example, found that 17- to
tion about themselves online than dren against Sexual Exploitation 18-year-old girls were almost four
through other forms of communi- and Sexual Abuse as “the intention- times more likely to agree to meet
cation.75 Being behind a computer al proposal, through information a stranger they met online than
screen, often in one’s own home, and communication technologies, 12- to 13-year-old girls.84 Although
can provide a perceived level of of an adult to meet a child…for the several studies have found that chil-
safety.76 However, children engaged purposes of [future sexual exploita- dren are generally aware that they
in online conversations in chat tion or abuse]”.79 A long-standing should not meet or divulge personal
rooms, social networking sites and body of research has documented
information to people they do not
other electronic forums may un- that sexual offenders often groom
know, they seem to make a distinc-
knowingly expose themselves to children and adolescents over a
tion between ‘strangers’ and ‘virtual
a global audience. Once personal period of time to lay the ground-
friends’. In many cases, children
images and information are posted work for future sexual advances.80
are aware that they are conversing
online, children may find out too Grooming is used by offenders to
with an adult online. And in some
late that they are unable to control build a relationship of trust, or a
instances, when children meet their
how these are used.77 friendship, with a child in order to
pursuer in person, they may view
gain access to her or him; this typi-
any sexual activities they engage in
Online sexual abuse of children can cally involves subtle behaviours that
as consensual.85
take many forms and ranges in its victims may not recognize as steps
severity. Children may be contacted towards abuse.81 Online grooming
Child abuse images online
online by an adult or another young can occur over periods varying from
person for the purpose of creating hours to months, based on the ob- The sexual abuse and exploitation
and distributing images and videos jectives of the offender and the re- of children through images involves
of a sexual nature, including those sponse of the child. Some research the representation of a child en-
documenting sexual abuse. Online has indicated that perpetrators may gaged in real or simulated sexual
70
activities or of his or her sexual ed and exchanged between paedo- Sexual cyber-bullying
body parts for a sexual purpose.86 philes without sale94 and some are
Sexual cyber-bullying involves the
National and international legisla- distributed through their interna- use of sexually loaded terms to in-
tion usually refers to such images tional online networks.95 The ready sult someone, tease someone be-
as pornography. However, law access and assumed anonymity as- cause of his or her sexual behaviour,
enforcement and child protection sociated with the Internet may en- make threats or jokes about sexual
agencies are increasingly using the courage sex offenders to seek such assault or rape, spread rumours
term ‘child abuse images’, since it images of children online.96 about a person’s sexual behaviour
better connotes their exploitative
or pressure someone to engage in
and harmful nature.87 Research has established the long-
sexual activities.98 Sexual cyber-
term harm associated with being a
Although the exact number is un- bullying also includes the distribu-
victim of this type of online abuse.
known, millions of child abuse im- tion of sexually explicit photos and
The impact, in fact, can mirror the
ages are estimated to be circulating videos taken of other children for
effects seen in children who have
on the Internet.88 The exchange and the purposes of shaming or causing
experienced other forms of sexual
sale of such images were occur- emotional distress to the victim. Re-
violence. However, children whose
ring long before the emergence of search has found that some of the
images appear online face addi-
the web. However, the Internet has most common methods of cyber-
tional issues due to the documen-
removed some of the communica- bullying include ‘sexting’ (that is,
tation and public distribution of creating and sending sexually ex-
tion barriers and social sanctions their abuse. When sexual images
regarding pornography and pae- plicit images over mobile phones)
are distributed on the Internet, it is and posting inappropriate photos
dophilia.89 In an instant, child abus-
virtually impossible to destroy them or videos online.99 Although forms
ers can gain access to thousands
or prevent them from continuing of sexual bullying can also be com-
of exploitative images of children,
to be accessed – either online or mitted offline, the speed and ease
including live videos.90 Additionally,
offline in other places where they afforded by the Internet and mobile
children themselves are now able
may also be stored. Children whose devices has the potential to intensi-
to easily access pornographic im-
images have been used must carry fy the severity of the consequences.
ages through the Internet, including
the burden of not knowing whether
those depicting other children.91
these will continue to be used by
Many perpetrators are involved in their abuser and/or other offend-
the production, distribution, pur- ers. Young people who experience
chase and viewing of this form of this type of violence are often con-
online child sexual abuse. Some cerned that offenders will show
are also seeking access to children their images to other children as a
and adolescents, both online and form of abuse. They may also be
offline, with the intention of cre- concerned that other people, includ-
ating sexually explicit images of ing their friends, peers and parents,
them.92 Images of child abuse can may become aware of the images.
have substantial commercial value, For some, the possibility of expo-
and the industry growing up around sure can become a constant source
them is becoming more profitable.93 of anxiety that persists well into
However, such images are also trad- adulthood.97
71
Figure 4.2
Boys also report experiences of sexual violence, but to a lesser extent than girls
Percentage of boys aged 15 to 19 years who ever experienced forced sexual intercourse or any other forced sexual acts (including
in childhood), and percentage of boys aged 15 to 19 years who experienced forced sexual intercourse or any other forced sexual
acts in the last 12 months
Experienced sexual violence ever Experienced sexual violence in the last 12 months
7% 2% 3% 1% 2% 0% 1% n/a
Uganda Mozambique Cameroon Bolivia (Plurinational
State of)
Notes: Data on the proportions of boys who experienced forced sexual intercourse or any other forced sexual acts in the last 12 months are not available for the Plurinational State of Bolivia; lifetime
prevalence data for this country refer to forced sexual intercourse only.
Sexual violence against boys sex. In the United Republic of Tanzania, around one
in seven boys and men aged 13 to 24 living on the
Comparable data on forced sexual intercourse
mainland said they were sexually victimized before
and other forced sexual acts among boys are only
age 18, compared to about one in 10 of those living
available for four countries (Figure 4.2).100 In Uganda,
on the island of Zanzibar.101 Among adolescent boys
adolescent boys are nearly two times less likely than
aged 13 to 17 on the mainland and in Zanzibar, about
their female counterparts to report experiences of
6 per cent and 4 per cent, respectively, said they
forced sexual intercourse or other forced sexual
experienced at least one form of sexual violence
acts during their lifetimes. Similarly, in Mozambique,
within the past year. The most commonly reported
a much smaller proportion of adolescent boys than
form of sexual violence against males in both areas
adolescent girls said they were the victims of sexual
of the country was unwanted sexual touching. In
violence (3 per cent versus 9 per cent, respectively).
Zimbabwe, 9 per cent of men aged 18 to 24 said they
were subjected to some form of sexual violence prior
The VACS in Kenya, the United Republic of Tanzania
to age 18, while around 2 per cent of boys aged 13 to
and Zimbabwe included information on reported
17 said they were victims of sexual violence within
experiences of sexual violence among boys and
the previous year. Again, the most prevalent form of
men. In those surveys, sexual violence was defined
sexual violence was unwanted sexual touching.
as any sexual act perpetrated against one’s will and
included a range of offences, including physically
forced or coerced intercourse, attempted unwanted
Age at first experience of
intercourse and other forms of abusive sexual sexual violence among girls
contact (such as unwanted touching of a sexual Comparable data are available for 21 countries on
nature). Among men aged 18 to 24 in Kenya, around the percentage of adolescent girls (aged 15 to 19)
18 per cent indicated that they were a victim of who were subjected to sexual violence by the age at
sexual violence prior to age 18; about 4 per cent which they first experienced it.102 These data need
of adolescent boys aged 13 to 17 reported having to be interpreted with caution and bear some level
been the victim of sexual violence within the past of uncertainty since there are significant proportions
12 months. The most commonly reported forms of those who could not recall the exact age at which
of sexual violence against Kenyan boys were they first experienced sexual violence and of missing
unwanted sexual touching and unwanted attempted data in most countries.
72
As shown in Figure 4.3, in all of these countries share of them experienced their first incident of
except three (Gabon, Honduras and Uganda), the sexual violence when they were adolescents. In
majority of girls reported that they were victimized fact, in most of the countries in which respondents
for the first time between the ages of 15 and 19. were able to recall the age at which this occurred,
However, a substantial proportion of adolescent it was most often between the ages of 15 and 19.
girls experienced sexual violence for the first time In Ghana, for instance, half of the respondents said
at younger ages. In all of the countries except
they were first victims of sexual violence at this age.
India, Liberia, the Republic of Moldova, Sao Tome
and Principe, and Zimbabwe, at least one in five
The same pattern is also observed in countries with
adolescent girls who reported sexual violence said
nationally representative surveys other than DHS.
it occurred for the first time between the ages of
This includes Paraguay,103 where 41 per cent of girls
10 and 14. In Comoros, Honduras and Uganda, the
largest proportion of girls reported they experienced and women aged 15 to 44 who experienced forced
sexual violence for the first time between the ages sex reported that it happened for the first time
of 10 and 14. Among adolescent girls in Gabon, 32 between the ages of 15 and 19. In Jamaica,104 the
per cent were victimized before age 10. proportion was the same: 41 per cent of girls and
women aged 15 to 24 who were victims of forced
An analysis of data from the entire sample of girls sexual intercourse said it occurred for the first time
and women aged 15 to 49 confirms that a large when they were 15 to 19 years old.
Figure 4.3
A significant proportion of adolescent girls were first-time victims of sexual violence before age 15
Percentage distribution of girls aged 15 to 19 years who ever experienced forced sexual intercourse or any other forced sexual acts,
by age at first incident of the violence
Under age 10 10 to 14 years 15 to 19 years Don’t know/missing
Gabon
Honduras
Dominican Republic
Uganda
Nigeria
Rwanda
Zambia
Philippines
Republic of Moldova
Comoros
Ghana
Cameroon
Malawi
India
Zimbabwe
Haiti
Kenya
Mozambique
Liberia
0 10 20 30 40 50 60 70 80 90 100
Note: Data for Comoros are based on 25-49 unweighted cases.

73
In a national survey conducted in Turkey,105 a Age at first experience of
distinction was made between experiences of sexual violence among boys
sexual violence before and after age 15. In that
Comparable DHS data on the age at which adolescent
country, a larger share of young women (aged 15
boys (aged 15 to 19) first experienced sexual violence
to 24) reported having been forced to have sex
were only available for four countries. Moreover, the
or perform other sexual acts against their will by
number of boys who reported experiences of sexual
someone other than an intimate partner prior to the
violence was not sufficient to produce reliable
age of 15 than after that age (11 per cent versus 6
estimates. An analysis of the entire sample of boys
per cent). National surveys conducted in a number
and men aged 15 to 49 reveals that, like girls and
of Pacific island countries, including Kiribati,106
women, their first experiences of sexual violence
Solomon Islands107 and Vanuatu,108 also confirm
most often occurred between the ages of 15 and
that the experience of sexual violence among girls
19. In both Mozambique and Uganda, at least 4 in
and women is very likely to occur at young ages.
10 boys and men said they were sexually victimized
In Kiribati and Solomon Islands, respectively, 19 per
for the first time when they were 15 to 19 years old.
cent and 37 per cent of girls and women aged 15 to
In Cameroon, on the other hand, the largest share of
49 reported that they were touched sexually or made
boys and men reported being victimized for the first
to do something sexual that they did not want to do
time after age 20 (results not shown).
by someone other than a husband or partner before
age 15. Among adolescent girls aged 15 to 19 in
The VACS also included questions on the age
Vanuatu, nearly 3 in 10 (29 per cent) reported either
at which boys and men first experienced sexual
being touched sexually or made to do something
victimization (that is, physically forced or coerced
sexual they did not want to do by someone other
intercourse, attempted unwanted intercourse or
than a husband or partner before age 15.
unwanted sexual touching). As can be seen in
Figure 4.4, 38 per cent of men aged 18 to 24 in
Kenya experienced their first incident of sexual
Figure 4.4 violence at ages 14 or 15; a slightly lower proportion
(37 per cent) first experienced sexual violence at age
In Kenya and Zimbabwe, most men say their
16 or 17. Around one in four men reported that the
first incident of sexual violence took place
first incident occurred before age 14. In the survey
before age 16
conducted in the United Republic of Tanzania, nearly
Percentage distribution of men aged 18 to 24 years who 6 in 10 men (58 per cent) aged 18 to 24 from the
experienced physically forced intercourse or coerced
intercourse, attempted unwanted intercourse or unwanted mainland who experienced sexual violence before
sexual touching, by age at first incident of the violence age 18 said the first incident occurred when they
were 16 or 17 years old; more than one in four (27
Under age 14 14 to 15 years 16 to 17 years
per cent) said it happened when they were 14 or
15. An additional 16 per cent reported that they
Zimbabwe were victims of sexual violence for the first time
before age 14. Among adult men aged 18 to 24 in
Kenya Zimbabwe who experienced some form of sexual

violence prior to age 18, 46 per cent reported that
United Republic of Tanzania the first incident occurred when they were 16 or 17
years old. A similarly high proportion (43 per cent)
0 10 20 30 40 50 60 70 80 90 100 said they were victims of sexual violence for the
Note: Data for the United Republic of Tanzania refer only to the mainland and do not
first time when they were 13 or younger, with the
include Zanzibar. remaining 11 per cent reporting their first incident at
Source: UNICEF global databases, 2014, based on the VACS from Kenya (2010), the United
Republic of Tanzania (2009) and Zimbabwe (2011). 14 or 15 years of age.
74
Forced first sex among were more likely to report that it was forced than
girls and women girls and women who had sex for the first time at
age 20 to 24 (Figure 4.5). In Nepal, for instance,
In 17 comparable surveys, girls and women who
nearly half (47 per cent) of girls and women aged
reported ever having sexual intercourse were asked 15 to 49 who had sex for the first time before age
whether their first experience of sex was forced 15 said they were forced to do so; the proportion
(against their will).109 Among girls and women aged is 29 per cent for girls whose sexual debut was
15 to 49, the percentage of those who reported between the ages of 15 and 19, and 19 per cent for
forced sexual initiation ranges from 1 per cent in those whose first experience of sex was between
Timor-Leste to 29 per cent in Nepal (results not the ages of 20 and 24. In a few countries, such as
shown). The likelihood of being forced into sexual Cambodia, the Democratic Republic of the Congo
intercourse varies by age at first sex: In some and the United Republic of Tanzania, the proportion
countries, girls and women who first had intercourse of girls and women reporting forced first sex is
before age 15 or between the ages of 15 and 19 roughly the same across all three age groups.
Figure 4.5
Sexual initiation of girls and women is more likely to be forced if it occurs at younger ages
Percentage of girls and women aged 15 to 49 years who ever had sexual intercourse who say their first experience of intercourse
was forced against their will, by age at first sexual intercourse
Nepal
Cameroon
Zimbabwe
Ghana
Republic of Moldova
Kenya
Before age 15
Malawi 15 to 19 years
Philippines 20 to 24 years
Liberia
Mozambique
Dominican Republic
Timor-Leste
Cambodia
0
Azerbaijan
0 10 20 30 40 50
Note: Data for Azerbaijan and the Republic of Moldova on the proportions of girls and women who reported having sex for the first time before age 15 are based on 25-49 unweighted cases.
75
Forced first sex among in the Plurinational State of Bolivia, the Dominican
boys and men Republic, Guatemala, Kenya, the Republic of
Moldova, the United Republic of Tanzania and
DHS data on forced first sex among boys and men
Uganda also reported being victimized by a friend
are only available for Mozambique, and less than 1
or acquaintance. With the exception of Guatemala,
per cent overall reported their first experience of sex
relatively few girls reported being sexually violated
as being forced.110 However, information on this topic
by their father or stepfather. Other relatives were
was also collected through the VACS in Kenya, the
reported perpetrators in a significant number of
United Republic of Tanzania and Zimbabwe. Among
instances in Colombia, Comoros, Gabon, Guatemala
men aged 18 to 24 in Kenya who first had sexual
and Honduras. Reports of sexual victimization by
intercourse before age 18, 9 per cent reported that it
an in-law, teacher, employer or someone else at
happened unwillingly. Of those boys and men aged work, priest or other religious leader, and police
13 to 24 who had sexual intercourse for the first time officers or soldiers are relatively uncommon across
before age 18, 18 per cent from mainland United all countries. The prevalence of reported sexual
Republic of Tanzania and 13 per cent from Zanzibar violence by a stranger was highest in the Republic
reported that they were unwilling partners in their of Moldova (38 per cent), followed by Comoros (33
first sexual encounter (that is, they were either forced per cent), Nigeria and Zambia (both 27 per cent), the
or coerced into sex). In Zimbabwe, 7 per cent of men Dominican Republic (23 per cent), Rwanda (20 per
aged 18 to 24 who had their sexual debut prior to cent), Uganda (19 per cent), and Colombia and Kenya
age 18 said they were forced, pressured, tricked or (both 16 per cent). In countries including Kenya and
threatened to engage in sexual intercourse. Among Rwanda, a substantial proportion of teenage girls
adolescent boys aged 13 to 17 in that country who described the perpetrator of sexual violence against
had sexual intercourse for the first time in the 12 them as ‘other’. In almost half of the cases in both
months preceding the survey, 5 per cent said the Liberia and Sao Tome and Principe, information was
experience was unwanted. not reported on the identity of the perpetrators.
In sexual and reproductive health surveys undertaken The widespread perpetration of sexual violence
in four African countries in 2004, adolescent boys by intimate partners is further confirmed in other
aged 12 to 19 who had sexual intercourse were surveys. In the Swaziland VACS, for instance, slightly
asked if they engaged in it willingly.111 Overall, 13 per more than one third (36 per cent) of girls and women
cent of the boys in Ghana, 7 per cent in Malawi, 5 per aged 13 to 24 who experienced either physically
cent in Burkina Faso and 4 per cent in Uganda said forced or coerced sex, attempted unwanted sex
they were not willing partners. Moreover, younger or unwanted sexual touching before age 18 said it
boys (aged 12 to 14) were more likely than older boys occurred at the hands of a husband or boyfriend.
(aged 15 to 19) to report that they did not willingly A 2006-2007 national survey in Nicaragua recorded
engage in their first experience of intercourse. that around one quarter of girls and women aged 15
to 24 experienced forced sex at some point in their
Perpetrators of sexual lives by a current spouse or partner.113
violence against girls
Girls and women who reported experiences of
Perpetrators of sexual
violence against boys
sexual victimization were asked to describe their

relationship to the perpetrators, including both Comparable DHS data on perpetrators of forced
partners and non-partners. Table 4.1 presents this sexual intercourse or other forced sexual acts on
information for adolescent girls aged 15 to 19 in 25 boys are only available for Uganda, where the most
countries with comparable data.112 By far the most commonly reported perpetrators were identified as
commonly reported perpetrators of violence against strangers.114
girls across all countries are intimate partners,
defined as either a current or former husband, Reports of sexual violence by intimate partners of
partner or boyfriend. Significant proportions of girls boys and men were most common in the VACS
76
Table 4.1
Among adolescent girls who have been subjected to sexual violence, the most likely
perpetrator was an intimate partner
Percentage of girls aged 15 to 19 years who ever experienced forced sexual intercourse or any other forced sexual acts,
by perpetrator
Persons who committed sexual violence against girls
religious leader
Current/former
Employer/some-
Other relative
Police/soldier
band/partner
band/partner
acquaintance
Family friend
Current hus-
Father/step-
Former hus-
Own friend/
stepbrother
Neighbour/
community
one at work
boyfriend
Stranger
Brother/
member
Missing
Teacher
Priest/
In-law
father
Other
Bolivia (Plurinational State of) 14 - 17 3 - 9 1 26 - 0 1 - - 14 - 4 3
Cameroon 26 21 7 0 - 5 0.3 14 7 0 1 0 0.2 - - 7 9
Colombia - 6 8 8 2 17 0.1 15 - 0 - - - 16 - 9 -
Comoros (2) (5) (22) (0) (10) (14) (0) (6) (11) (0) (0) (0) (0) (33) - (2) (2)
Dominican Republic 7 12 8 1 - 5 0.4 31 4 0 0 0 0 23 - 5 7
Gabon 15 3 3 4 1 41 0.2 1 7 7 7 7 7 7 - 13 0
Ghana 10 43 3 0 - 5 1 11 6 5 0 0 0 10 - 0 10
Guatemala 19 14 3 19 - 13 - 27 - - - - - 9 - 1 -
Honduras 13 7 18 5 2 22 0 14 7 0 1 0 0.3 14 - 1 -
India 77 3 6 0.3 - 6 1 4 2 0 0.4 0 0 3 - 0.3 0
Jamaica 16 36 - - - 12 - 19 - - - - - 15 - 2 -
Kenya 7 29 6 0 - 4 1 23 1 0 0 0 0 16 - 14 3
Liberia 15 12 14 2 - 0 1 5 8 0 0 6 0 5 - 1 45
Malawi 27 20 5 1 - 5 1 5 2 2 0 0 0 11 - 4 26
Mozambique 44 25 14 4 - 1 0 11 2 0 0 0 0 9 - 0 0
Nepal 27 6 0.4 0 0 0 0 1 0 0 0 0 0 3 - 2 0
Nigeria 22 17 5 0 0.2 9 0.3 14 10 1 0.2 0 0 27 - 0 0
Philippines 25 40 7 6 - 2 0 6 2 0 3 0 0 1 3 0 3
Republic of Moldova 1 25 3 1 - 1 0 28 2 0 0 0 0 38 - 0 4
Rwanda 1 11 2 0 0 7 0 9 9 0 0.3 0 0 20 15 22 3
Sao Tome and Principe (26) (17) (6) (0) - (1) (0) (3) (0) (0) (0) (0) (0) (0) - (0) (48)
United Republic of Tanzania 24 14 5 0 - 9 0 27 3 0 0 0 0 9 - 6 7
Uganda 29 3 10 1 1 8 4 21 3 2 0 0 0 19 - 5 1
Zambia 14 9 2 0 - 12 3 4 6 2 1 0 0 27 - 5 0.1
Zimbabwe 47 25 12 0 0 10 0 1 3 0 0 0 0 5 - 4 0
Notes: Data for Colombia include only girls who were raped by someone other than a current spouse or partner. In Guatemala, sexual violence committed by a spouse or partner among ever-married
girls included forced sexual intercourse or agreeing to have sexual intercourse when they did not want to for fear of what their partner might do; sexual violence committed by anyone among all girls
and women included only forced sexual intercourse. Data for Jamaica refer to girls and women aged 15 to 24 years who experienced only forced sexual intercourse. Data for Comoros and Sao Tome
and Principe are based on 25-49 unweighted cases. For the Plurinational State of Bolivia, the category ‘Father/stepfather’ also includes ‘Mother/stepmother’.
Source: UNICEF global databases, 2014, based on DHS and other nationally representative surveys, 2005-2012.
77
conducted in Kenya, the United Republic of Tanzania incident occurred if they experienced sexual violence
and Zimbabwe. Among men aged 18 to 24 in Kenya, more than once). In the United Republic of Tanzania,
nearly half (43 per cent) reported that the first incident respondents were asked to report on the location
of sexual violence prior to age 18 was perpetrated of their first and most recent experience of sexual
by a girlfriend or romantic partner, while about one violence. In Swaziland, girls and women were asked
in five (21 per cent) identified a neighbour. Almost where the first incident of sexual violence occurred.
half (49 per cent) said the perpetrator was about the
same age as they were, while around 36 per cent In all four countries, most sexual violence took
reported the offender as 10 years older or more. place at home – either the victim’s own home, the
home of the perpetrator or the home of another
In the United Republic of Tanzania, dating partners person. In Kenya, 22 per cent of women and 32
(that is, current or former girlfriends) were the most per cent of men (aged 18 to 24) who reported
frequently cited perpetrators of sexual violence before experiences of sexual violence in childhood said
age 18 against boys and men aged 13 to 24; the rate the first incident occurred in their own homes; in
was much higher among those living in Zanzibar Zimbabwe, the proportions were about 32 per cent
(71 per cent) than those living on the mainland (48 for women and 31 per cent for men. In Swaziland,
per cent). A significant proportion of males living in over half of girls and women aged 13 to 24 who
both areas of the country reported that the sexual experienced sexual violence before age 18 said
violence was committed by a stranger, neighbour it occurred in someone’s home; one third said
or member of the community. Of those males aged it occurred in their own home and close to one
13 to 24 from the mainland who experienced sexual quarter said it happened in another person’s home
violence prior to age 18, a majority (58 per cent) (that of a friend, relative or neighbour). In the United
reported that the perpetrator was about the same Republic of Tanzania, nearly half of both females
age as they were. Among males from Zanzibar, a and males aged 13 to 24 reported that at least one
higher proportion (51 per cent) reported that at least experience of sexual violence before age 18 took
one of their experiences of childhood sexual violence place in someone’s home. In Kenya and Zimbabwe,
was perpetrated by someone older, while about 39 girls were significantly more likely than boys to
per cent said the perpetrator had been about the experience sexual violence in the perpetrator’s
same age. home. About 26 per cent of women in Kenya and
31 per cent of women in Zimbabwe reported that
In Zimbabwe, 27 per cent of men aged 18 to 24 said their first experience of sexual violence before age
18 occurred in the perpetrator’s home, in contrast
their first experience of sexual violence before age 18
to less than 10 per cent of men in both countries.
was committed by a girlfriend. However, neighbours
were the most commonly reported perpetrators (33
Experiences of sexual violence also appear to be
per cent). Among men aged 18 to 24, slightly more
common at school or on the way to or from school.
than one third (35 per cent) said the perpetrator of
In Kenya, about one in five women and men (aged 18
sexual violence against them prior to age 18 was at
to 24) who experienced sexual violence before age
least 10 years older.
18 reported that the first incident occurred at school.
However, women were significantly more likely to
Typical settings where
say it happened while travelling on foot compared

sexual violence occurs to men the same age (27 per cent and 14 per
Data from the VACS in Kenya, Swaziland, the United cent, respectively). Similar patterns are also found
Republic of Tanzania and Zimbabwe show that in Zimbabwe. There, women aged 18 to 24 were
adolescents are at risk of sexual violence in many significantly more likely than their male peers to
locations they frequent on a day-to-day basis. In report that their first experiences of sexual violence
Kenya and Zimbabwe, respondents who reported at before age 18 occurred on the way to or from school
least one incident of sexual violence before age 18 (19 per cent and 7 per cent, respectively). In the
were asked where it happened (or where the first United Republic of Tanzania, experiences of at least
78
one incident of sexual violence during childhood of sexual violence before age 18 occurred outside –
while travelling to or from school were reported by in a field, the bush, river or roadway. In Swaziland,
23 per cent of females and 15 per cent of males aged about one in five girls and women (aged 13 to 24)
13 to 24. experienced sexual violence before age 18 in a public
area, field or open area of land. In Zimbabwe, 19 per
Although girls are generally more likely to experience cent of men and 15 per cent of women (aged 18
sexual violence while travelling to and from school, to 24) reported that childhood sexual violence took
both sexes are at risk in a range of locations outside place for the first time outdoors. In Kenya, less than
the home and school. For example, about one in 5 per cent of both women and men (aged 18 to 24)
four males and females aged 13 to 24 in the United said their first encounter of sexual violence before
Republic of Tanzania reported that at least one incident age 18 occurred outdoors or in the bush.
Box 4.6
Additional findings on sexual violence

in selected low- and middle-income
countries with non-comparable data
In the 2005 Global School-based Stu- and female students in grades 7 and however, that research has consist-
dent Health Survey (GSHS) in Leba- above in the former Yugoslav Re- ently found that most sexual assaults
non, 20 per cent of boys aged 13 to public of Macedonia reported being are not reported to the police, and
15 said they were exposed to sexual physically forced to have sex when even when they are, the age of the
comments, touched in a sexual way they did not want to at some point reported victim is often unknown.120
or the victim of an attempted forced in their lives in the 2007 GSHS (6 per Researchers are also unclear which
sexual relationship at some point in cent and 5 per cent, respectively). kinds of sexual assault are the most
their lives, compared to 15 per cent of likely to be underreported.121
girls the same age. South Africa is reported to have one
of the highest rates of sexual violence Moving beyond police-reported sta-
In a 2012-2013 national survey car- in the world.116 Several small-scale tistics, a 2011-2012 nationally repre-
ried out in Madagascar,115 girls and studies have found that adolescent sentative survey found that around
women aged 15 to 49 were asked girls there are at particular risk for 1 in 20 secondary school students (5
whether they experienced any of the experiencing forced sex – with es- per cent) reported at least one act of
following in the 12 months preceding timates ranging from 39 per cent to unwanted sexual contact (regardless
the survey: non-consensual sexual 66 per cent.117 Despite this, nationally of whether penetration occurred or
intercourse, made to do other non- representative data on the prevalence not) at school in the year preceding
consensual sexual acts, had their of sexual violence against children re- the survey.122 Girls were much more
intimate parts touched in a non- main limited. Some indications of the likely to report having been recently
consensual way, made to have sex-
extent of the problem are available, sexually violated at school than boys
ual intercourse with other people for
however. For example, an analy- (8 per cent and 1 per cent, respec-
money or other favours not granted,
sis of the 2010/2011 police records tively). Another study conducted in
or made to live with or get married
found that a total of 28,128 sexual the Eastern Cape and KwaZulu-Natal
to someone without their consent.
offences against children under the provinces found that 28 per cent of
Around 14 per cent of girls aged 15 to
age of 18 were reported to the police, men reported previously raping a
19 said they experienced one of the
representing just over half of all the woman (whether alone or with ac-
above forms of sexual violence in the
reported crimes committed against complices).123 Reports of rapes by
previous year.
children that year.118 An analysis by multiple perpetrators were especially
In the 2009 GSHS in Maldives, 18 per victims’ age using the 2008/2009 high, with 20 per cent of men report-
cent of adolescent boys in grades 8 to police-recorded data reveals that ing that they participated in a gang
10 said they were physically forced to around 6 in 10 of the reported sexual rape. However, only 5 per cent of men
have sex when they did not want to at offences committed against children admitted that they had raped a child
some point in their lives, compared to that year affected those below the under the age of 15. Primary reasons
16 per cent of girls the same age. age of 15 and that about one quarter cited for raping a girl below the age of
of the child victims were under the 18 included boredom, fun and sexual
Nearly equal proportions of male age of 10.119 It is important to note, entitlement.124
79
Box 4.7
Sexual violence in selected high-income

countries
Germany125 The identity of the perpetrator was with the most commonly reported
found to vary by the type of of- perpetrators being acquaintances,
Data from a 2011 study in Germany
fence: For sexual harassment with- relatives and strangers.
on a nationally representative sam-
out physical contact, the highest
ple of individuals aged 16 to 40
proportion was reportedly perpe- Switzerland127
found that about 6 per cent of wom-
trated by male offenders unknown
en and 1 per cent of men reported The Optimus Study,128 carried out
to the victim (47 per cent), followed
incidents of sexual harassment in in Switzerland in 2009, collected
by male family members (33 per
their youth. The definition used was information from adolescent boys
cent) and other males known to
“being forced to see intimate parts and girls (aged 15 to 17) on their
the victim (27 per cent). For sexual
for the sexual arousal of someone experiences of sexual victimization.
violence with physical contact, male
else at least once before the age Around 40 per cent of girls and 20
offenders known to the victim, in-
of 16”. An additional 7 per cent of per cent of boys reported incidents
cluding neighbours or friends, were
girls and women and 1 per cent of of sexual victimization that did not
the most commonly reported per-
boys and men said they were vic- involve physical contact at some
petrators (45 per cent), followed by
tims before the age of 16 of sexual point in their lives. This included
male family members (41 per cent),
violence with physical contact (de- indecent exposure, verbal or writ-
strangers (23 per cent) and female
fined as physical contact of intimate ten harassment (including online)
offenders (4 per cent). For other
parts, vaginal or anal penetration or and exposure to pornography. Of
types of sexual offences, the most
oral sex). Finally, about 2 per cent of those who experienced non-contact
commonly reported offenders were
girls and women and less than 1 per sexual victimization at least once,
male family members (36 per cent),
cent of boys and men reported that more than one quarter (27 per cent)
followed by males unknown to the
they were the victims of some other of girls and 14 per cent of boys said
victim (25 per cent), female offend-
type of sexual violence before the it occurred within the last year. The
ers (9 per cent) and other males
age of 16. Findings also revealed most common form of sexual vio-
known to the victim (3 per cent).
variations by ethnic background: lence among both sexes was cyber-
For example, women with roots in victimization.
the Russian Federation were most Italy126
likely to report sexual harassment, A 2006 national survey collected In terms of sexual victimization
while those of German origin were data from girls and women aged 16 involving physical contact, includ-
most likely to report contact sexual to 70 in Italy on their experiences of ing sexual touching, attempted or
violence. Girls and women of Turk- being touched sexually or forced to completed intercourse or other sex-
ish background registered the low- do any other sexual activity against ual acts, such as oral sex, a much
est reported prevalence rates of their will. Around 7 per cent said higher proportion of adolescent
both sexual harassment and contact they experienced sexual violence girls (22 per cent) said they experi-
sexual violence. by a non-partner before age 16, enced at least one incident in their
80
lifetime than adolescent boys (8 United States of America130 boys.

per cent). Among those who expe-
The second National Survey of
rienced contact sexual victimiza- Around one in five adolescent girls
Children’s Exposure to Violence
tion at some point, 14 per cent of (21 per cent) and 11 per cent of ado-
(NatSCEV II), carried out in 2011 in lescent boys were found to have
girls reported an incident within the
the United States, captured reports been sexually harassed at some
past year compared to 6 per cent
on both lifetime and past-year sex- point during their lives; the propor-
of boys. Around 7 per cent of girls
ual victimization, including sexual tions declined to 14 per cent of girls
and 1 per cent of boys reported the
harassment, indecent exposure, and 5 per cent of boys when they re-
most severe form of contact sexual
attempted or completed rape and ported on incidents during the last
victimization (attempted or com-
other types of sexual offences in- year. Lifetime prevalence rates of
pleted penetration) at some point in
volving physical contact. More than “unwanted Internet sex talk” were
their lives.
one third of adolescent girls (35 per 21 per cent and 4 per cent for girls
cent) and one in five adolescent and boys, respectively; last-year
United Kingdom of Great Britain boys (20 per cent) aged 14 to 17 revictimization rates were 13 per cent
and Northern Ireland129 ported experiencing such forms of and 3 per cent for girls and boys, re-
A nationally representative survey violence at some point in their lives. spectively.
of youth (aged 11 to 17) conduct- The proportions fell when reporting
ed in the United Kingdom in 2009 on sexual victimization during the
found that around 17 per cent ex- past year: 23 per cent of girls and
perienced contact or non-contact 10 per cent of boys reported at least
sexual abuse by an adult or peer one such incident in the previous 12
during their lifetime; 9 per cent ex- months.
perienced such victimization within
the past year. The majority of perpe- Around 17 per cent of adolescent
trators were found to be males who girls and 4 per cent of adolescent
were typically known to the victim. boys reported experiences of sexual
assault at some point in their lives,
Around 7 per cent of girls and 3 defined as attempted or completed
per cent of boys revealed that they rape or other sexual offences in-
experienced some form of contact volving physical contact. Sexual as-
sexual abuse (as defined by crimi- sault by a known adult was reported
nal law) by an adult or peer at some by 6 per cent of girls and less than 1
point during their lives. Most re- per cent of boys; sexual assault by a
ported the perpetrator to be another peer was reported by 11 per cent of
child (under age 18) rather than an girls and 4 per cent of boys. About
adult inside or outside the home. 11 per cent of girls and 2 per cent
Among the 11- to 17-year-olds re- of boys were the victims of a sexual
porting sexual aggression by adults, assault within the last year. Sexual
most said the offender was known assault at the hands of peers within
to them (either a parent or guardian, the previous year was reported by
neighbour or family friend). 7 per cent of girls and 2 per cent of
81
The sexual exploitation of children females and 6 per cent of males aged 18 to 24 who
experienced sexual violence before age 18 reported
Although it is widely acknowledged that sexual
receiving money for sex; 8 per cent of females and
exploitation of children occurs in many parts of the
4 per cent of males reported receiving gifts, food or
world, population-based surveys pertaining to the
subject have only been conducted in a small number favours in exchange for sex. In the VACS conducted
of countries. Moreover, these surveys covered only on mainland United Republic of Tanzania in 2009, 4
some forms of sexual exploitation, using different per cent of girls aged 13 to 17 said they received
questions to elicit children’s experiences. money or goods in exchange for sex at least once
in their lifetime (there were too few cases among
In the VACS carried out in Swaziland in 2007, in boys to report a stable prevalence estimate). In the
which only females were sampled, about 2 per cent 2011 Zimbabwe survey, among both females and
of young women aged 13 to 24 said that a teacher males aged 18 to 24 who reported physical, sexual
or principal offered money, gifts, food, shelter or or emotional violence prior to age 18, about 3 per
better grades in exchange for sex at some point in cent said they received money, gifts, food or favours
their lives; 8 per cent reported that the same was in exchange for sex.
offered by someone other than a teacher or principal
during their lives. A much smaller proportion of the In 2004, a series of nationally representative surveys
respondents said they had sex with others because on adolescent sexual and reproductive health were
they “hoped to receive” money or goods. In the also carried out in Burkina Faso, Ghana, Malawi
2010 VACS conducted in Kenya, 7 per cent of and Uganda.131 The surveys included a question
82
Figure 4.6
In Poland, about one in five boys report they offered sexual services for pay at least once
Percentage of 18-year-olds who reported that they offered sexual services for pay at least once, by sex
50
Boys Girls
40
30
20
10
0
Poland Northwest Russian Lithuania Sweden Norway Estonia
Federation
Source: Baltic Sea Regional Study on Adolescents’ Sexuality, 2003-2004.
on whether adolescents aged 12 to 19 who had adolescents within this age group.
sexual intercourse within the previous 12 months
received gifts or money from their last sex partner in The Baltic Sea Regional Study on Adolescents’
exchange for sex.132 The question was only asked of Sexuality, carried out between 2003 and 2004,
girls who were not married at the time of the survey also incorporated questions relating to sexual
and who reported having sex more than once in the exploitation.133 In this study, nationally representative
year preceding the survey. samples of adolescents in secondary and vocational
schools in Estonia, Lithuania and Poland and
In all four countries, a higher percentage of single
representative samples from large cities in Norway,
adolescent girls than boys who had sex within the
Sweden and Northwest Russian Federation134
previous 12 months reported receiving gifts or
were asked whether or not they ever offered
money for sex from their last sexual partner. Among
sexual services for pay.135 Figure 4.6 presents
single girls aged 12 to 19, the proportions receiving
the percentage of 18-year-old girls and boys who
gifts or money were 79 per cent in Malawi, 75 per
cent in Uganda, 73 per cent in Ghana and 35 per cent reported that they exchanged sexual services for
in Burkina Faso. Among single boys, the proportions pay at least once. The highest proportions among
were 32 per cent in Uganda, 28 per cent in Ghana, both sexes were in Poland, where about one in five
8 per cent in Malawi and 5 per cent in Burkina Faso. boys and one in eight girls (both aged 18) reported
These findings should be interpreted with caution, that they offered sexual services for pay at least
however, since they are restricted to only those once. Rates were somewhat similar for both boys
adolescents not currently married or in union and and girls in Northwest Russian Federation (between
therefore do not represent the experiences of all 8 per cent and 10 per cent, respectively).
83
When different forms of violence
overlap (POLY-VICTIMIZATION)
Poly-victimization among far outweigh those of sexual violence. for adolescent girls to experience a
adolescent girls In some countries, however, physical combination of sexual and physical
violence is only slightly more common violence than sexual violence alone.
It is well documented in the literature than sexual violence among adoles- In Côte d’Ivoire, Timor-Leste, Uganda
that adolescent girls are more likely to cent girls. In Malawi, for example, 21 and Ukraine, for example, over three
experience physical violence than sex- per cent and 18 per cent of adolescent quarters of adolescent girls who were
ual violence. This is confirmed by De- girls reported incidents of physical and subjected to sexual violence also re-
mographic and Health Survey (DHS) sexual violence, respectively.
data from 25 countries.1 In every ported experiences of physical vio-
country with available data, the preva- The data also show that most girls lence. In Uganda, 16 per cent of girls
lence of physical violence among girls who are subjected to physical vio- reported both sexual and physical vio-
aged 15 to 19 exceeds that of sexual lence experience this type of violence lence, while 3 per cent reported sexual
violence. In Timor-Leste, for example, in isolation. However while physical violence only. The prevalence of poly-
about 30 per cent of adolescent girls violence is rarely accompanied by victimization was especially high in a
reported incidents of physical violence, violence of a sexual nature, many ado- handful of countries in sub-Saharan
in contrast to 2 per cent of girls report- lescent girls who are sexually violated Africa. In Cameroon, Gabon, Uganda,
ing sexual violence. Similarly, in Côte also report accounts of physical abuse. Zambia and Zimbabwe, more than 1 in
d’Ivoire, Gabon, Equatorial Guinea and In most of the 25 countries for which 12 adolescent girls reported incidents
Uganda, reports of physical violence data are available, it is more common of both physical and sexual violence.
Data on the overlap of various forms

of physical, sexual and emotional vio-
Most adolescent girls who are victims of sexual violence also lence are also available through the
report physical abuse Violence Against Children Surveys
Percentage of girls aged 15 to 19 years who experienced physical violence only (since (VACS) in Kenya, the United Republic
age 15), sexual violence only (including in childhood), or both physical and sexual violence of Tanzania2 and Zimbabwe.3 In both
Kenya and Zimbabwe, 76 per cent of
Uganda adult women aged 18 to 24 reported
Cameroon some form of violence (physical, sex-
Equatorial Guinea ual or emotional) during childhood.
Gabon
Ghana Among them, the largest proportion
Zambia reported physical violence only prior to
Malawi age 18. In Kenya, a combination of all
Kenya three forms of violence was reported
Côte d’Ivoire
second most often (13 per cent). In
When different forms of violence overlap
Zimbabwe
Haiti Zimbabwe, a combination of physical
Timor-Leste and sexual violence was reported sec-
United Republic of Tanzania ond most often (14 per cent). In both
Nigeria countries, sexual violence among girls
Mozambique
Rwanda during childhood rarely occurred in
India isolation, with less than 6 per cent re-
Physical violence only
Philippines porting incidents of sexual abuse only.
Comoros Physical and sexual violence The same pattern was found in the
Nepal Sexual violence only United Republic of Tanzania, where
Azerbaijan
Tajikistan about 84 per cent of girls and women
Kyrgyzstan between the ages of 13 and 24 who
Ukraine experienced sexual violence in child-
Kazakhstan hood also reported incidents of physi-
0 10 20 30 40 50 60 70 80 90 100 cal violence. In addition, nearly 43 per
cent of females who said they were
victims of sexual violence in childhood
also experienced emotional abuse.
84
Among women in Kenya and Zimbabwe, the most common experience of violence during
childhood was physical, often in combination with emotional or sexual violence
Percentage distribution of women aged 18 to 24 years, by type of violence experienced before age 18
Kenya Zimbabwe
Sexual violence only Sexual violence only
No violence No violence
Sexual, physical and Physical Sexual, physical and Physical

emotional violence violence only emotional violence violence only
Emotional violence only

Physical and emotional violence Emotional violence only Physical and emotional violence
Sexual and emotional violence Sexual and physical violence Sexual and emotional violence Sexual and physical violence
Source: UNICEF global databases, 2014, based on the VACS from Kenya (2010) and Zimbabwe (2011).
Poly-victimization among In Cameroon, Mozambique and Uganda, adolescent boys are far
adolescent boys more likely to experience physical than sexual violence
Considerably less data are avail- Percentage of boys aged 15 to 19 years who experienced physical violence only (since
able on physical and sexual violence age 15), sexual violence only (including in childhood), or both physical and sexual violence
among adolescent boys.4 DHS data 100
from Cameroon, Mozambique and 90
Physical violence only
Uganda reveal that adolescent boys 80 Physical and sexual violence
are also far more likely to experi- 70 Sexual violence only
ence physical violence than sexual 60
violence. In Cameroon, around 80 per 50
cent of adolescent boys aged 15 to 40
19 said they had experienced physi- 30
cal violence since age 15 while about 20
2 per cent reported experiences of
10
sexual violence at some point in their
0
lives. In Uganda, over half of boys Cameroon Uganda Mozambique
aged 15 to 19 reported incidents of Source: UNICEF global databases, 2014, based on DHS, 2011.

physical abuse since age 15, while
6 per cent recounted experiences Data on the overlapping of various sexual – during childhood (9 per cent
of sexual violence. Similarly, in Mo- forms of violence are also available in Kenya and 5 per cent in Zimbabwe).
zambique, the prevalence of physical for males through the VACS in Kenya, In both countries, sexual violence
violence significantly exceeded that the United Republic of Tanzania and among men before age 18 rarely oc-
of sexual violence among adoles- Zimbabwe. In Kenya and Zimbabwe, curred in isolation: Less than 1 per
cent boys, with about one quarter respectively, 79 per cent and 84 per cent in both countries reported the
reporting physical violence and 3 per cent of all men aged 18 to 24 reported experience of sexual violence only.
cent reporting sexual violence. As some form of violence (either physi- The same pattern was found in the
with girls, most boys did not experi- cal, sexual or emotional) at some United Republic of Tanzania, where
ence sexual violence in isolation. In point during their childhood. Among about 83 per cent of boys and men
all three countries, over half of boys them, the largest proportion expe- between the ages of 13 and 24 who
who were subjected to sexual vio- rienced physical violence only prior experienced sexual violence before
lence also reported physical abuse. to age 18. A combination of physical age 18 also related incidents of physi-
Still, the combination of both sexual and emotional violence was reported cal abuse. In addition, about half (51
and physical violence is rare – experi- second most often in both countries. per cent) of males who reported sex-
enced by less than 5 per cent of boys Few men reported all three forms of ual violence also said they been the
in each country. violence – emotional, physical and victims of emotional violence.
85
In some countries, boys are more likely to report experiences of
physical violence while girls more often report sexual violence
in combination with physical abuse
Among men in Kenya and Zimbabwe, the most common form Poly-victimization and sex
of violence experienced during childhood was physical, usually differences
in combination with emotional violence In the two countries for which compa-
Percentage distribution of men aged 18 to 24 years, by type of violence experienced rable DHS data are available for both
before age 18 sexes (Mozambique and Uganda),
overall prevalence rates for violence
Kenya are similar for girls and boys. How-
Sexual violence only ever, the most common forms of
No violence
violence experienced are different.
Boys are more likely to be physically
violated than girls, while girls overall
are more likely to experience sexual
violence, usually in combination with
Sexual, physical and Physical
physical violence. In Uganda, for ex-
emotional violence violence only
ample, half of boys reported incidents
of physical violence only, compared to
39 per cent of girls. Among girls, 19
per cent related incidents of sexual
Physical and emotional violence
abuse as opposed to 6 per cent of
boys; of these, similar proportions re-
Emotional violence only ported experiencing sexual violence
Sexual and emotional violence Sexual and physical violence
in isolation. However, 16 per cent of
girls and 4 per cent of boys in Uganda
experienced both physical and sexual
violence. This pattern was also con-
Zimbabwe
firmed by findings from the VACS in
Sexual violence only
Kenya and Zimbabwe: Men reported
No violence higher rates of physical violence only
during childhood, while women were
Sexual, physical and more likely to report sexual violence
emotional violence only or a combination of sexual and
Physical physical violence.5 Findings from
violence only these two countries also show that

physical violence tends to overlap
with emotional violence in boys to a
greater extent than it does in girls.
Physical and emotional violence However, it is not possible to know
to what extent these differences be-
tween the sexes can be explained by
Sexual and emotional violence Emotional violence only the relatively well-documented pat-
Sexual and physical violence tern that males are less likely to report
experiences of violence. The findings
therefore need to be interpreted with
Source: UNICEF global databases, 2014, based on the VACS from Kenya (2010) and Zimbabwe (2011). some degree of caution.
86
Girls are more likely than boys to be victims of sexual violence, often in combination with
physical abuse
Percentage of girls and boys aged 15 to 19 years who experienced physical violence only (since age 15), sexual violence only (including in
childhood), or both physical and sexual violence
Physical violence only Physical and sexual violence Sexual violence only
Girls
Uganda
Boys
Girls
Mozambique
Boys
0 10 20 30 40 50 60 70 80 90 100
Source: UNICEF global databases, 2014, based on DHS, 2011.
Percentage distribution of women and men aged 18 to 24 years, by types of violence experienced before age 18
Physical violence only Sexual violence only Sexual and emotional violence Sexual, physical and emotional violence
Physical and sexual violence Emotional violence only Physical and emotional violence No violence
Women
Zimbabwe
Men
Women
Kenya
Men
0 10 20 30 40 50 60 70 80 90 100
Source: UNICEF global databases, 2014, based on the VACS from Kenya (2010) and Zimbabwe (2011).

references
1 Data were collected in other countries that conducted a Nations Children’s Fund, US Centers for Disease Control
DHS but are either not presented in the final DHS reports and Prevention and Muhimbili University of Health and
or results are based on fewer than 25 unweighted cases Allied Sciences, Violence against Children in Tanzania: Find-
and therefore are not included here. ings from a national survey 2009, UNICEF Tanzania, Dar es
Salaam, 2011; Zimbabwe National Statistics Agency, United
2 Data on the overlap of various forms of violence are pre-
Nations Children’s Fund and Collaborating Centre for Opera-
sented differently in the United Republic of Tanzania survey
tional Research and Evaluation, National Baseline Survey on
report.
Life Experiences of Adolescents, 2011, Harare, 2013.
3 United Nations Children’s Fund Kenya Country Office,
4 Data were collected in other countries that conducted a
Division of Violence Prevention, National Center for Injury
DHS but are either not presented in the final DHS reports
Prevention and Control, US Centers for Disease Control
or results are based on fewer than 25 unweighted cases
and Prevention and Kenya National Bureau of Statistics,
and therefore are not included here.
Violence against Children in Kenya: Findings from a 2010
National Survey – Summary report on the prevalence of 5 Findings from the DHS are not directly comparable with
sexual, physical and emotional violence, context of sexual those from the Kenya and Zimbabwe surveys, since the
violence, and health and behavioral consequences of specific acts of violence covered differ somewhat between
violence experienced in childhood, Nairobi, 2012; United the two sets of surveys.
87
Disclosing experiences of violence
and seeking help
Research consistently shows that seek help. In societies that sanction may, in some communities, be
most child victims of violence delay male dominance over women, expected to seek help from their or
disclosing their abuse – sometimes children may witness gender-based their partner’s family; typically, the
for relatively long periods of time. violence on a regular basis and see ultimate goal of an intervention is
Many children never tell anyone, nor its occurrence as an integral part of to bring about reconciliation in the
do they seek help to cope with the interpersonal dynamics between relationship rather than protection
experience or to protect themselves the sexes. The perception that for the victim. Additionally, those
from further victimization.1 One some forms of violence, such as
who experience violence at the
reason may be the fear of retaliation. corporal punishment by parents or
hands of their intimate partners
Children may also feel guilty, physical fights with peers, are just an
may be reluctant to come forward
ashamed or confused. Some children ordinary part of growing up can make
children less likely to consider their because of economic dependence on
may lack confidence in the ability or
victimization worth mentioning or to their partners or fear of reprisal.2 To
willingness of others to help them or
think of themselves as in need of help. make matters worse, formal support
be unaware of services available for
victims. Moreover, in some places, cultural services for victims of violence
norms dictate to whom a victim is are lacking in many communities,
Cultural and social norms can also expected to turn for assistance. For creating even more obstacles for
play a role in whether or not children instance, victims of partner violence those who consider seeking help.3
The perception that some forms of violence are an ordinary part

of growing up can make CHILD VICTIMS less likely to think of them-
selves as in need of help
Disclosing episodes of violence Significant proportions of girls and women who experience
In the Demographic and Health physical and/or sexual violence never tell anyone about the abuse
Surveys (DHS), girls and women Percentage of girls aged 15 to 19 years and women aged 20 to 49 years who ever
who reported incidents of physical experienced physical and/or sexual violence and never told anyone about it
and/or sexual violence during their
DISCLOSING EXPERIENCES OF VIOLENCE AND SEEKING HELP
lifetimes were asked whether they 100

15 to 19 years
ever told anyone about the abuse. 90
20 to 49 years
Available data from 20 low- and 80
middle-income countries confirm 70
that many girls and women never 60
disclose their experience.4 In Jordan, 50
Kyrgyzstan, Nepal, Nigeria, Pakistan, 40
Tajikistan and Uganda, more than
30
half of adolescent girls aged 15 to
20
19 who suffered from physical and/
10
or sexual violence said they never
told anyone. In most of the countries, 0
Nepal
Kyrgyzstan
Tajikistan
Uganda
Nigeria
Jordan
Pakistan
Kenya
Haiti
Mozambique
Zimbabwe
Azerbaijan
Zambia
Cameroon
Ghana
United Republic of
Tanzania
Equatorial Guinea
Gabon
Comoros
Malawi
adolescent girls were found to be

as likely as adult women (aged 20
to 49) to keep their experiences of
violence secret. Exceptions include
Kyrgyzstan, where adolescent girls
were significantly less likely than their Notes: Data for Jordan refer to ever-married girls and women. Data for Pakistan refer to ever-married girls and women who have
ever experienced physical violence. Data on adolescent girls aged 15 to 19 years for Jordan and Pakistan are based on 25-49
older counterparts to have confided in unweighted cases.
someone about the abuse. Source: UNICEF global databases, 2014, based on DHS, 2006-2013.
88
When it comes to disclosure of age 18 usually confided in their mother aged 18 to 24, and here again men
different types of violence, female or father (41 per cent) or a friend (36 were most likely to say that they had
victims of sexual violence only were per cent). Among women aged 18 to told a friend about the incident(s). In
least likely to disclose the abuse as 24 in Kenya and Zimbabwe, most told Zimbabwe, the proportion was 45
opposed to those who experienced a relative or friend. per cent and about equal numbers of
either physical violence only or men said they told a friend or relative.
both physical and sexual violence.5 Among boys aged 15 to 19 in
This discrepancy was particularly Mozambique who experienced A comparison of data from countries
pronounced in Zimbabwe. In that physical and/or sexual violence, 51 with information on both sexes
country, 72 per cent of girls and women per cent said they never told anyone, indicates that boys are about as likely
aged 15 to 49 who experienced only according to DHS data.7 In Ghana, 40 as girls to keep their victimization
sexual violence never told anyone, in per cent of adolescent boys aged 15 to themselves. When confidence
contrast to 48 per cent of those who to 19 who experienced only physical intervals are taken into account, the
experienced only physical violence violence chose not to disclose the only difference in the data by sex
and 41 per cent who experienced abuse. As with girls, younger boys in is observed in the United Republic
both physical and sexual violence. both countries were about as likely of Tanzania, where 69 per cent
In a few cases, disclosure was least as adult men (aged 20 to 49) to never of males aged 13 to 24 said they
likely among those who experienced disclose their experience of violence. did not disclose an incident(s) of
only physical violence. In Azerbaijan, Another similarity with girls is the
sexual violence, versus 48 per cent
for example, nearly half (47 per reported rate of disclosure by type of
of females the same age. In some
cent) of girls and women who ever violence: Rates were lowest among
countries, however, significant
experienced physical violence boys and men reporting sexual
differences between the sexes did
only never told anyone about the violence only and were about the
emerge in terms of whom the victims
incident(s), compared to 15 per cent same for those subjected to either
told. In Kenya, for instance, women
of those who experienced sexual physical violence only or to both
aged 18 to 24 who were victims of
violence only. physical and sexual violence.
sexual violence before age 18 were
In the Violence Against Children Disclosure of sexual violence among more likely to have told a relative
Surveys (VACS) carried out in three males was also assessed in the VACS. than men of the same age (70 per
countries in Eastern and Southern In the United Republic of Tanzania, 31 cent versus 22 per cent). On the
Africa, female victims of child sexual per cent of boys and men aged 13 to other hand, in the United Republic of
abuse who disclosed the incidents(s) 24 who experienced childhood sexual Tanzania, males aged 13 to 24 were
were asked whom they told.6 In the violence told someone about the more likely than females of the same
United Republic of Tanzania, girls incident(s), most often talking about it age to disclose childhood experiences
and women aged 13 to 24 who with friends. In Kenya, the proportion of sexual abuse to a friend (71 per
experienced sexual violence prior to was 36 per cent among adult men cent versus 36 per cent).

Generally, girls and boys are equally likely to keep secret their experience of violence
Percentage of females and males of different ages (see notes) who ever experienced physical and/or sexual violence and never told
anyone about it
United Republic of Tanzania (sexual violence) Males

Females
Kenya (sexual violence)
Zimbabwe (sexual violence)
Mozambique (physical and/or sexual violence)
0 10 20 30 40 50 60 70 80 90 100
Notes: Data for Mozambique refer to girls and boys aged 15 to 19 years who ever experienced physical and/or sexual violence.. Data for the United Republic of Tanzania refer to girls and women and boys and
men aged 13 to 24 years who experienced sexual violence before age 18. Data for Kenya and Zimbabwe refer to women and men aged 18 to 24 years who experienced sexual violence before age 18.
Source: UNICEF global databases, 2014, based on DHS, 2008-2011 and the VACS from Kenya (2010), the United Republic of Tanzania (2009) and Zimbabwe (2011).
89
Seeking help or services asked for help. In general, adolescent countries with available data, girls
girls are less likely than adult women and women were most likely to
The DHS also asked girls and women
to seek help in the face of abuse. In seek help when they experienced
whether they sought help from
8 of the 30 countries, reported rates both physical and sexual violence as
any source to end the violence and,
of help-seeking among adult women opposed to one or the other.9 With a
if so, from whom. Findings from
30 countries confirm that most aged 20 to 49 exceeded those of handful of exceptions, reported rates
adolescent girls who experience adolescent girls by more than 10 of help-seeking were higher among
violence do not seek help.8 In all 30 percentage points. The discrepancy those who experienced physical
countries, more than half of girls aged was particularly high in Kyrgyzstan, violence only than among those who
15 to 19 who experienced physical where adult women were 2.5 times experienced sexual violence only.
and/or sexual violence said they did more likely to ask for help in stopping This discrepancy was particularly high
not ask anyone for assistance. Help the violence than adolescent girls. In in the United Republic of Tanzania,
was least sought by adolescent girls Cabo Verde, Colombia and Equatorial where 45 per cent of girls and women
in the Plurinational State of Bolivia, Guinea, reported rates of help-seeking aged 15 to 49 who experienced only
Colombia, Kyrgyzstan, the Philippines, among women were about twice as physical violence said they sought
and Timor-Leste. In those countries, high as those among girls. help from others to stop the abuse,
less than 20 per cent of girls aged 15 versus 18 per cent of those who
to 19 who suffered violence reportedly In the overwhelming majority of experienced only sexual violence.
In some countries, adolescent girls who have been victims of violence are less likely to seek help
than adult women
Percentage of girls aged 15 to 19 years and women aged 20 to 49 years who ever experienced physical and/or sexual violence and
sought help from any source to stop the violence
100
15 to 19 years
90
20 to 49 years
80
70
60
50
40
30
20
10
0
Zambia
Bangladesh
Gabon
Comoros
Rwanda
Peru
Zimbabwe
Ghana
Honduras
Dominican Republic
Burkina Faso
Côte d’Ivoire
Mozambique
Jordan
Uganda
Pakistan
Haiti
Cameroon
Kenya
Tajikistan
Nigeria
Nepal
Equatorial Guinea
Cabo Verde
Timor-Leste
Kyrgyzstan
Philippines
Colombia
Notes: Data for Bangladesh refer to currently married girls and women who experienced physical violence in the last 12 months and received assistance from others. Data for Colombia refer to ever-married
girls and women who experienced physical violence committed by their husband or partner, suffered an injury and visited a medical centre for assistance. Data for Honduras refer to ever-married girls and
women who experienced physical or sexual violence committed by their husband or partner in the last 12 months. Data for Jordan refer to ever-married girls and women. Data for Pakistan refer to ever-married
girls and women who ever experienced physical violence. Data for Peru refer to girls and women who sought help from someone close to them (rather than an institution). Data for the Plurinational State of
Bolivia refer to ever-married girls and women who ever experienced any form of violence committed by their husband or partner and sought help from someone close to them (rather than an institution). Data
for Cabo Verde, Jordan and Pakistan on adolescent girls aged 15 to 19 years are based on 25-49 unweighted cases.
90
When female victims of violence said they actually tried to seek help to 24 who experienced childhood
seek help, most look to their own for their victimization. In Zimbabwe, sexual violence sought professional
families for assistance, according to among women aged 18 to 24 who services. Even among males who
DHS data from 29 countries.10 This experienced emotional, physical or knew where to find professional help,
was the case in all countries except sexual violence as children, 33 per rates of help-seeking were low. In
Rwanda, where friends or neighbours cent, 39 per cent and 38 per cent, Kenya, among men aged 18 to 24
were more commonly turned to for respectively, reportedly knew where who experienced emotional, physical
aid. In India, Jordan, Kyrgyzstan, to go for help; here again, however, or sexual violence as children, 27 per
Nigeria, Pakistan and Tajikistan, over less than 10 per cent of them actually cent, 20 per cent and 13 per cent,
70 per cent of girls and women sought assistance. respectively, reported knowing a
who sought help reported that they place to seek professional assistance.
looked to their own families. After DHS data on help-seeking behaviour
Of these men, fewer than 1 in 9 said
the victim’s own family, in-laws and/ are also available for Ghanaian boys
they actually tried to seek help. In
or family members of the victim’s aged 15 to 19 who experienced
Zimbabwe, the proportions were 38
husband or partner were the second physical violence and for boys the
per cent, 38 per cent and 35 per cent,
most commonly reported source of same age in Mozambique who were
respectively, with less than 10 per
support. This was the case for more subjected to physical and/or sexual
violence.11 In both countries, less than cent seeking professional support.
than one third of female victims in
Kenya, Tajikistan, Ukraine, the United a third of these boys sought help. In
A comparison of data for the two
Republic of Tanzania, Zambia and Ghana, the proportion was around 30
sexes shows that boys are less likely
Zimbabwe. Many girl and women per cent, and this was roughly the
than girls to seek help. In the United
victims also reported going to friends same for older men who sought help
Republic of Tanzania, for example,
and neighbours for help; in Haiti and for physical abuse. In Mozambique,
female victims (aged 13 to 24) of
Nepal, the proportion was over 40 21 per cent of boys sought help,
sexual violence in childhood were
per cent. compared to 37 per cent of men aged
almost twice as likely as male victims
20 to 49. As was the case for girls,
of the same age to seek help (22 per
reported rates of help-seeking among
cent versus 12 per cent). Data for
boys and men were lowest among
Kenya and Zimbabwe indicate that
When female victims those reporting sexual violence only.
only a very small proportion of both
of violence seek Findings from the VACS in the United boys and girls who knew of a place
Republic of Tanzania showed that 12 to go for professional help actually
help, most look to
per cent of boys and men aged 13 sought support (results not shown).
their own families
for assistance

In Mozambique and the United Republic of Tanzania, boys are
less likely to seek help than girls
Overall, victims are much more likely Percentage of females and males of different ages (see notes) who ever experienced
to turn to individuals they know physical and/or sexual violence and sought help from any source to stop the violence
personally for support rather than (Mozambique) or sought services (the United Republic of Tanzania)
to institutions such as the police
department, medical facilities, legal 50
Males
aid, religious groups and/or social Females
40
services. It could be argued that
many girls and women do not seek
30
professional services because they
are unaware of places that offer help. 20
Data from the VACS in Kenya indicate
that among women aged 18 to 24 10
who experienced emotional, physical
or sexual violence as children, 24 per 0
Mozambique (physical and/or sexual violence) United Republic of Tanzania (sexual violence)
cent, 16 per cent and 25 per cent,
respectively, reported knowing a Notes: Data for Mozambique refer to girls and boys aged 15 to 19 years who ever experienced physical and/or sexual violence. Data
for the United Republic of Tanzania refer to girls and women and boys and men aged 13 to 24 years who experienced sexual violence
place to go for professional help. Of before age 18.
these women, less than 10 per cent Source: UNICEF global databases, 2014, based on DHS, 2011 and the VACS from the United Republic of Tanzania (2009).
91
Why children may not what they experienced was a form
seek help of violence or do not see their abuse Most boys remain
as a problem. This was the reason
The four surveys in sub-Saharan Africa given by 57 per cent of women in silent about
sought answers as to why some
victims chose not to seek assistance
Zimbabwe who were subjected to their experiences
physical violence, 32 per cent who
as a result of violence. The surveys
experienced sexual violence and 31 of violence,
included females aged 13 to 24 in
Swaziland and both sexes of this age
per cent who suffered emotional regardless of the
violence before age 18. In Swaziland,
in the United Republic of Tanzania
slightly more than 20 per cent of type, because they
who experienced sexual violence
before age 18 but did not tell anyone
girls and women aged 13 to 24 who do not view it as a
suffered sexual violence as children
about it. In Kenya and Zimbabwe, the problem
did not disclose the incident(s)
surveys included male and female
because they were “not aware that
respondents between the ages of
it was abuse”. In Kenya, one quarter
18 and 24 who experienced physical, Among boys and men aged 13 to 24
of women who were subjected to
sexual and/or emotional violence prior in the United Republic of Tanzania
physical violence, 17 per cent who
to turning 18, knew of a place to get
were victimized emotionally and who experienced sexual violence as
professional help but did not seek
15 per cent who were victimized children, the majority (58 per cent)
services.
sexually as children said they did not cited personal reasons for not telling
Findings from the United Republic seek services because they did not anyone, primarily because they did
of Tanzania show that 34 per cent view the abuse as a problem. This not think it was a problem. Slightly
of girls and women who were was also the response offered by more than one third (36 per cent)
sexually violated in childhood did around one in nine girls and women cited family or community reasons for
not tell anyone for fear of being aged 13 to 24 in the United Republic remaining silent. Among Kenyan men
abandoned by, or separated from, of Tanzania who experienced sexual
aged 18 to 24 who were victims of
their families. In Swaziland, 23 per violence in childhood.
childhood physical abuse, most failed
cent of females did not come forward
to seek help because they did not
for fear of abandonment. About one
consider it a problem. An additional
third of women in Kenya and one
in five women in Zimbabwe who Many victims do not 36 per cent of these young men said
experienced childhood sexual abuse it was because they were afraid of
realize that what getting into trouble. Among adult
did not seek services because they
did not want to embarrass themselves they experienced men of the same age who reported
or their family. experiences of emotional violence
was a form of
before age 18, the largest proportion
violence or do not
Fear of getting into trouble and/or (34 per cent) said they did not seek
getting the perpetrator into trouble
was another commonly cited reason
see the abuse as a help because they did not think it was
a problem; an additional 28 per cent
for not seeking professional help. problem were afraid of getting into trouble.
This was the case for about 15 per
cent of women respondents in Kenya In Zimbabwe, the most commonly
who were subjected to sexual abuse cited reason for not seeking services
In Zimbabwe, just over 25 per among men aged 18 to 24 who
before age 18. In Swaziland, around
cent of women aged 18 to 24 who suffered physical, sexual and/or
one in seven female victims did not
disclose their abuse for fear of getting experienced physical violence as emotional abuse prior to age 18 was
the perpetrator into trouble, and the children reported that they did not
because they did not think it was a
same reason was given by about one seek help because they did not feel
problem. Fear of getting into trouble
in four respondents in Zimbabwe. they needed or wanted it. This was
was the second most commonly
Fear of causing trouble was less a much less common response
cited response among those who
commonly reported by those who among those experiencing sexual
or emotional violence. In Kenya, had been victims of physical and/
endured physical violence.
however, 18 per cent of victims of or emotional violence. Among those
It was also clear from the surveys physical violence and 16 per cent of who suffered sexual violence only,
that many young victims do not those subjected to sexual violence the second most common reason
disclose their experiences or seek did not seek help because they did given was to avoid embarrassment to
help because they do not realize that not want or feel they needed it. themselves or their families.
92
© UNICEF/NYHQ2001-0404/DeCesare
references
1 Paine, M. L., and D. J. Hansen, Factors Influencing United Nations Children’s Fund, US Centers for Disease
Children to Self-disclose Sexual Abuse, Faculty Publica- Control and Prevention and Muhimbili University of
tions, Department of Psychology, Paper 59, University of Health and Allied Sciences, Violence against Children in
Nebraska-Lincoln, Lincoln, 2002. Tanzania: Findings from a national survey 2009, UNICEF
Tanzania, Dar es Salaam, 2011; Zimbabwe National
2 McCleary-Sills, J., et al., Help-seeking Pathways and Statistics Agency, United Nations Children’s Fund and
Barriers for Survivors of Gender-based Violence in Collaborating Centre for Operational Research and
Tanzania: Results from a study in Dar es Salaam, Mbeya, Evaluation, National Baseline Survey on Life Experiences
and Iringa regions, Engender Health/Champion, Dar es of Adolescents, 2011, Harare, 2013; United Nations
Salaam, 2013; Kim, J., and K. A. Gray, ‘Leave or Stay? Children’s Fund Swaziland, A National Study on Violence

Battered women’s decision after intimate partner vio- against Children and Young Women in Swaziland,
lence’, Journal of Interpersonal Violence, vol. 23, no. 10, UNICEF Swaziland, Mbabane, 2007.
2008, pp. 1465-1482.
7 Data were collected in other countries that conducted
3 McCleary-Sills et al., op. cit. a DHS but are either not presented in the final DHS
reports or results are based on less than 25 unweighted
4 Data were collected in other countries that conducted cases and therefore are not included here.
a DHS but are either not presented in the final DHS
reports or results are based on less than 25 unweighted 8 Data were collected in other countries that conducted a
cases and therefore are not included here. DHS but are not presented in the final DHS reports and
therefore are not included here.
5 Although these data are based on all girls and women
between the ages of 15 and 49 years, the general pat- 9 Although these data are based on all girls and women
terns are expected to hold for adolescent girls. between the ages of 15 and 49 years, the general pat-
terns are expected to hold for adolescent girls.
6 United Nations Children’s Fund Kenya Country Office, Di-
vision of Violence Prevention, National Center for Injury 10 Data were collected in other countries that conducted a
Prevention and Control, US Centers for Disease Control DHS but are not presented in the final DHS reports and
and Prevention and Kenya National Bureau of Statistics, therefore are not included here.
Violence against Children in Kenya: Findings from a 2010
National Survey – Summary report on the prevalence 11 Data were collected in other countries that conducted
of sexual, physical and emotional violence, context of a DHS but are either not presented in the final DHS
sexual violence, and health and behavioral consequences reports or results are based on less than 25 unweighted
of violence experienced in childhood, Nairobi, 2012; cases and therefore are not included here.
93
Chapter 5 The most
common
form of
violence
against
children
Teaching children self-control and acceptable behav-
iour is an integral part of child discipline in all cul-
tures. Positive parenting practices involve guidance
on how to handle emotions or conflicts in a manner
that encourages sound judgement and responsibil-
ity and preserves children’s self-esteem, dignity and
physical and psychological integrity. All too often,
however, children are raised using methods that
rely on physical force or verbal intimidation to punish
unwanted behaviours and encourage desired ones.
In many cases, rather than being a deliberate dis-
ciplinary choice, such violent methods are used as
a result of parents’ anger and frustration, or lack of
knowledge of non-violent responses.
Physical discipline, also known as corporal punish-

ment, refers to “any punishment in which physical
force is used and intended to cause some degree
of pain or discomfort, however light”.1 It includes
acts such as kicking, pinching, spanking, shaking or
throwing children, hitting them with a hand or imple-
ment (such as a whip, stick, belt, shoe or wooden
© UNICEF/NYHQ2011-1882/LeMoyne
spoon) or forcing them to ingest something. Violent

psychological discipline involves the use of verbal
aggression, threats, intimidation, denigration, ridi-
cule, guilt, humiliation, withdrawal of love or emo-
tional manipulation to control children.
Both types of violent discipline are violations of chil-

dren’s rights (see Box 5.1). Physical discipline and
94
psychological aggression tend to overlap and fre- development during infancy and childhood and af-
quently occur together, exacerbating the short- and fect the child’s physical, cognitive, emotional and
long-term harm they inflict. Studies have found that social growth.3 Repeated exposure to violent dis-
exposing children to violent discipline has negative cipline increases the proclivity for delinquency and
consequences that vary according to the nature, adult criminal behaviour.4 And children raised by au-
extent and severity of the exposure.2 These range thoritarian parents who regularly employ harsh and
from immediate impacts to long-term harm that chil- punitive disciplinary methods tend to have reduced
dren carry into adult life. Injuries inflicted by a care- self-esteem and lower academic success, are more
giver on a child can result in death or serious dam- hostile and aggressive and less popular with peers,
age, including cognitive and physical impairments. and are less independent than children who have
In addition, exposure to severe, unpredictable or not suffered in this way; such children also engage
prolonged toxic stress can physiologically alter brain in more substance abuse as adolescents.5
Box 5.1
Violent discipline and the Convention on

the Rights of the Child
Violent discipline is a violation of a of the child, appropriate direction in the performance of their child-
child’s right to protection from all and guidance in the exercise by the rearing responsibilities and shall
forms of violence while in the care child of the rights recognized in the ensure the development of institu-
of their parents or other caregivers, present Convention.” tions, facilities and services for the
as set forth in the Convention on the care of children.”
Rights of the Child.6 A fundamental In addition, article 18 emphasizes
that parents have the primary re- While the Convention recognizes
principle of the Convention, con-
sponsibility for the upbringing and and respects the responsibility of
tained in its preamble, is that the
CHAPTER 5: The most common form of violence against children

development of the child and that parents and other caregivers to pro-
family is the natural environment
States Parties shall render appropri- vide “direction and guidance” to
for the growth and well-being of
ate support to parents in the perfor- children, there is an explicit under-
all of its members, particularly chil-
mance of their child-rearing respon- standing that such guidance should
dren. Thus, the Convention recog-
sibilities: not involve any form of violence.
nizes the pivotal role of the family
The Committee on the Rights of the
in protecting children and looking
“1. States Parties shall use their Child in its General Comment No. 8
after their physical and emotional
best efforts to ensure recognition of has stated that the “interpretation
welfare. Article 5 of the Convention of ‘appropriate’ direction and guid-
the principle that both parents have
clearly acknowledges the responsi- ance must be consistent with the
common responsibilities for the up-
bilities, rights and duties of parents bringing and development of the whole Convention and leaves no
and other caregivers in providing child. Parents or, as the case may room for justification of violent or
appropriate direction and guidance be, legal guardians have the prima- other cruel or degrading forms of
in the process of children’s develop- ry responsibility for the upbringing discipline”.7 The Convention also
ment: “States Parties shall respect and development of the child. The mandates that children should be
the responsibilities, rights and du- best interest of the child will be their protected from violent discipline
ties of parents or, where applicable, basic concern. while at school. According to article
the members of the extended family 28(2): “States Parties shall take all
or community as provided for by lo- “2. For the purpose of guaranteeing appropriate measures to ensure that
cal custom, legal guardians or other and promoting the rights set forth in school discipline is administered in
persons legally responsible for the the present Convention, States Par- a manner consistent with the child’s
child, to provide, in a manner con- ties shall render appropriate assis- human dignity and in conformity
sistent with the evolving capacities tance to parents and legal guardians with the present Convention.”
95
Use of disciplinary Figure 5.1
methods
Despite their detrimental and long-
The use of violent discipline in the home is widespread
lasting impact, violent forms of Percentage of children aged 2 to 14 years who experienced any violent disci-
discipline are extremely common. pline (psychological aggression and/or physical punishment) in the past month
Comparable data from 62 countries

or areas show that households use Yemen
Ghana
violent disciplinary practices with Tunisia
Togo
the overwhelming majority of chil- Cameroon
dren: On average, about four in five State of Palestine
Central African Republic
children between the ages of 2 and Democratic Republic of the Congo
Egypt
14 are subjected to some kind of Côte d’Ivoire
violent discipline in the home, with Nigeria
Morocco
percentages ranging from a low of Gambia
Jordan
45 per cent in Panama to a high Liberia
of almost 95 per cent in Yemen Swaziland
Syrian Arab Republic
(Figure 5.1). Algeria
Mauritania
Congo
Because of its pervasive nature, Suriname
Haiti
children in all regions of the world Jamaica
Chad
are at risk of being exposed to vio- Vanuatu
lent discipline at home.8 For exam- Burkina Faso
Lebanon
ple, in all of the countries with avail- Sierra Leone
Niger
able data in both sub-Saharan Africa Guinea-Bissau
and the Middle East and North Kiribati
Iraq
Africa, more than 7 in 10 chil- Tajikistan
dren aged 2 to 14 years are dis- Trinidad and Tobago
Albania
ciplined in a violent manner. In all Azerbaijan

Guyana
but two (Costa Rica and Panama) Lao People’s Democratic Republic
of the 12 countries with available Republic of Moldova
Barbados
data in Latin America and the Afghanistan
Viet Nam
Caribbean, more than half of children Argentina
experience violent discipline. The Djibouti
Solomon Islands
prevalence of violent discipline in Fiji
Belize
Central and Eastern Europe and the Armenia
Commonwealth of Indepen- The former Yugoslav Republic of Macedonia
Saint Lucia
dent States (CEE/CIS) is greater Dominican Republic
Serbia
than 50 per cent in all coun- Georgia
tries except Kazakhstan (at 49 Belarus
Montenegro
per cent). These estimates are Ukraine
Bosnia and Herzegovina
comparable to the levels found Kyrgyzstan
in other countries, including Kazakhstan
Mongolia
some high-income countries Costa Rica
Panama
(see Box 5.2).
0 10 20 30 40 50 60 70 80 90 100
Psychological aggression is slightly Notes: Data for Belarus differ from the standard definition. Data for Fiji, Kiribati and Solomon Islands refer to adult
reports of whether they have used physical discipline on their children. Data for Kyrgyzstan refer to children aged 3
more common than physical pun- to 14 years. Data for Panama refer to children aged 1 to 14 years. For Argentina, the sample was national and urban
ishment, although both forms of (municipalities with a population of more than 5,000), since the country’s rural population is scattered and accounts for
less than 10 per cent of the total.
violent discipline are widespread. Source: UNICEF global databases, 2014, based on DHS, MICS and other nationally representative surveys, 2005-2013.
96
On average, around 7 in 10
Figure 5.2
children experience psycho-
Physical punishment and psychological aggression are logical aggression while 6 in 10
both pervasive experience physical punishment.
The rate of psychological aggres-
Percentage of children aged 2 to 14 years who experienced psychological
aggression and percentage of children aged 2 to 14 years who experienced sion is highest in Yemen, where
physical punishment in the past month it exceeds 90 per cent, while the
Psychological aggression Physical punishment highest rates of physical punish-
ment are found in the Central
Yemen
Tunisia African Republic, Egypt and Ye-
State of Palestine men, where more than 80 per cent
Ghana
Côte d’Ivoire of children are reportedly subject-
Jordan ed to such practice (Figure 5.2).
Cameroon
Togo In only a few countries, including
Algeria Afghanistan, Djibouti and Haiti, are
Syrian Arab Republic children slightly more likely to be
Liberia
Egypt disciplined using physical rather
Mauritania than psychological methods.
Swaziland
Suriname
Gambia
For the most part, households
Nigeria employ a combination of violent
Congo
Burkina Faso disciplinary practices. As shown
Niger in Figure 5.3, most children in a
Iraq
Sierra Leone majority of countries or areas are
Azerbaijan
Vanuatu
exposed to both psychological as
Tajikistan well as physical means of punish-
Jamaica
Chad ment. This confirms that these
Lao People’s Democratic Republic two forms of violence often over-

Albania
Republic of Moldova lap and frequently occur together
Trinidad and Tobago within the context of discipline.
Guinea-Bissau
Armenia Such exposure to multiple types
Guyana
Argentina of violence may exacerbate the po-
Haiti tential harm to a child in both the
Barbados
Afghanistan short and long term.9
Saint Lucia
Serbia
Georgia The most severe forms of physi-
Belarus
Ukraine
cal punishment (hitting the child
Djibouti on the head, ears or face or hitting
The former Yugoslav Republic of Macedonia
Montenegro the child hard and repeatedly) are
Viet Nam less common overall: About 17 per
Belize
Kazakhstan cent of children, on average, are
Kyrgyzstan
subjected to these practices. In 23
Mongolia countries or areas, however, se-
Panama
Costa Rica vere forms of corporal punishment
0 10 20 30 40 50 60 70 80 90 100 are widespread, with more than
one in five children subjected to
Notes: Data for Belarus differ from the standard definition. Data for Kyrgyzstan refer to children aged 3 to 14 years.
Data for Panama refer to children aged 1 to 14 years. For Argentina, the sample was national and urban (municipali-
them (Figure 5.4). Extremely harsh
ties with a population of more than 5,000), since the country’s rural population is scattered and accounts for less than corporal punishment is particularly
10 per cent of the total.
Source: UNICEF global databases, 2014, based on DHS, MICS and other nationally representative surveys, 2005-2013. common in a few countries, such
97
Figure 5.3
Most children experience a combination of violent disciplinary methods

Percentage of children aged 2 to 14 years who experienced both physical punishment and psychological aggression, percentage of
children aged 2 to 14 years who experienced only physical punishment and percentage of children aged 2 to 14 years who experi-
enced only psychological aggression in the past month
Both physical and psychological Only psychological Only physical
Ghana
Tunisia
Togo
Cameroon
State of Palestine
Egypt
Côte d’Ivoire
Nigeria
Gambia
Jordan
Liberia
Swaziland
Algeria
Mauritania
Congo
Suriname
Haiti
Jamaica
Chad
Vanuatu
Burkina Faso
Sierra Leone
Niger
Guinea-Bissau
Iraq
Tajikistan
Trinidad and Tobago

Albania
Azerbaijan
Guyana
Lao People’s Democratic Republic
Republic of Moldova
Barbados
Afghanistan
Viet Nam
Argentina
Djibouti
Belize
Armenia
Saint Lucia
Serbia
Georgia
Belarus
Montenegro
Ukraine
Kyrgyzstan
Kazakhstan
Mongolia
Costa Rica
Panama
0 10 20 30 40 50 60 70 80 90 100
Notes: Data for Belarus differ from the standard definition. Data for Kyrgyzstan refer to children aged 3 to 14 years. Data for Panama refer to children aged 1 to 14 years. For Argentina, the sample
was national and urban (municipalities with a population of more than 5,000), since the country’s rural population is scattered and accounts for less than 10 per cent of the total.
Source: UNICEF global databases, 2014, based on DHS, MICS and other nationally representative surveys, 2005-2013.
98
as Afghanistan, the Central African

Figure 5.4
Republic, Chad, the Democratic
Over 40 per cent of children experience severe physical Republic of the Congo, Egypt, Ni-
punishment in Chad, Egypt and Yemen geria, Vanuatu and Yemen, where
Percentage of children aged 2 to 14 years who experienced severe physical
more than one in three children are
punishment in the past month affected.
Yemen
Egypt Regional variations can be found.
Chad
Afghanistan Countries or areas with the highest
Democratic Republic of the Congo levels of severe physical punish-
Vanuatu ment are generally concentrated in
Nigeria
Tunisia
sub-Saharan Africa and the Middle
Niger East and North Africa. In 16 of the
Guinea-Bissau
Liberia 29 countries or areas in these re-
Mauritania gions with available data, more
Cameroon
Iraq than one in four children aged 2
Congo
State of Palestine
to 14 years have been subjected
Algeria to a harsh form of corporal punish-
Morocco
Syrian Arab Republic ment in the past month. All the
Côte d’Ivoire countries in which less than 10 per
Djibouti
Jordan cent of children were subjected to
Georgia
Burkina Faso
the most severe forms of corporal
Azerbaijan punishment are found in CEE/CIS,
Sierra Leone
Tajikistan East Asia and the Pacific, and Latin
Gambia America and the Caribbean.
Guyana
Togo
Haiti
Reliance on violent discipline does

Albania
Ghana not mean that caregivers do not
Lebanon
Suriname also use non-violent methods. The
Swaziland latter include practices such as
Argentina
Lao People’s Democratic Republic explaining why a behaviour is
Montenegro
wrong, taking away some of a child’s
Barbados
Saint Lucia privileges and distracting a child by
Jamaica
Trinidad and Tobago giving him or her something else
Belize to do. In reality, most children
Bosnia and Herzegovina are disciplined using a combina-
Armenia tion of non-violent and violent
Viet Nam
Kyrgyzstan means, but very few experience
Costa Rica
Mongolia
non-violent discipline exclusively.
Panama The proportion of children receiv-
Kazakhstan
Republic of Moldova ing only non-violent discipline
Dominican Republic ranges from 4 per cent in Camer-
Serbia
Ukraine oon, Ghana and Yemen to 48 per
0 10 20 30 40 50 cent in Costa Rica (Figure 5.5).
However, in only eight countries
Notes: Data for Kyrgyzstan refer to children aged 3 to 14 years. Data for Panama refer to children aged 1 to 14 years. do at least one third of children
For Argentina, the sample was national and urban (municipalities with a population of more than 5,000), since the experience non-violent methods
country’s rural population is scattered and accounts for less than 10 per cent of the total.
Source: UNICEF global databases, 2014, based on DHS, MICS and other nationally representative surveys, 2005-2013. exclusively.
99
Figure 5.5
Children who experience only non-violent forms of discipline are in the minority
Percentage of children aged 2 to 14 years who experienced only non-violent discipline or non-violent discipline combined with
violent discipline (psychological aggression and/or physical punishment) in the past month
Only non-violent discipline Non-violent discipline combined with violent discipline
Belarus
Tunisia
State of Palestine
Albania
Republic of Moldova
Togo
Ukraine
Serbia
Yemen
Azerbaijan
Montenegro
Jordan
Guinea-Bissau
Argentina
Viet Nam
Chad
Gambia
Costa Rica
Mauritania
Belize
Nigeria
Côte d’Ivoire
Iraq
Armenia
Algeria
Trinidad and Tobago

Tajikistan
Suriname
Liberia
Kyrgryzstan
Niger
Ghana
Jamaica
Guyana
Sierra Leone
Kazakhstan
Congo
Djibouti
Cameroon
Swaziland
Barbados
Mongolia
Saint Lucia
Afghanistan
Georgia
Burkina Faso
Panama
Haiti
0 10 20 30 40 50 60 70 80 90 100
100
A closer analysis of the specific methods used almost 7 in 10 children exposed to this. Spanking
with children indicates that certain disciplinary ac- and hitting the child with a bare hand is the most
tions are far more common than others (Figure common form of physical punishment, used on 44
5.6). On average, 81 per cent of children receive per cent of children.
an explanation of why their behaviour was wrong
and 48 per cent have privileges taken away. Few- Additional data from selected countries show that
er experience the other non-violent form (distract- other violent methods are used as well: In Haiti,
ing the child with something else to do). Among more than half of children are punished by making
violent methods measured, shouting, yelling or them kneel on the floor, while in the Congo, one in
screaming is the most widely used form, with three have had their ears pulled (Table 5.1).
Figure 5.6
Certain disciplinary actions are far more common than others

Percentage of children aged 2 to 14 years who experienced any discipline in the past month, by type
100
90
80
70
60
50
40
30
20
10
0
Explained Shouted, Took away Gave Spanked, Called Shook Hit/ Hit on the Hit/slapped Beat with an

why yelled or privileges something hit or dumb/lazy slapped bottom or on face, object/beat
behaviour screamed at else to do slapped with on hand, elsewhere head or as hard as
was wrong bare hand arm or leg with object ears one could
Note: This is a weighted average based on comparable data for 54 countries or areas.
Table 5.1
In Haiti, more than half of children are punished by making them kneel on the floor
Percentage of children aged 2 to 14 years who experienced any violent discipline in the past month, by type, in selected countries
Isolated/not Refused/de- Burned with

talked to/ prived food/ fire or hot Made to sleep Made to
ignored meal instrument outside Bit Pulled ears kneel
Cameroon 22 9 - - - - -
Congo - 12 - - - 35 -
Côte d'Ivoire - 11 0.4 2 - - -
Ghana 11 - - - - - -
Haiti - 2 - - - 15 52
Iraq - - 1 - 3 - -
Jamaica 13 - - - - - -
101
Violent disciplinary Figure 5.7
practices and
children’s In most countries, boys and girls are at about equal risk of
characteristics experiencing violent discipline
Percentage of children aged 2 to 14 years who experienced any violent discipline
Research has found that boys (psychological aggression and/or physical punishment) in the past month, by sex
tend to experience violent of the child
discipline to a greater extent 100
than girls.10 Available figures
90
from comparable surveys
partially support this claim. 80
In just under one third of the 70

countries or areas with data, 60
boys are slightly more likely to be Girls are more likely Boys are more likely
Girls 50 to experience violent to experience violent
subjected to violent disciplinary discipline discipline
practices, including in Bosnia 40
and Herzegovina, Costa Rica, 30

Kazakhstan, Kyrgyzstan and
20
Ukraine (Figure 5.7). In the
10
remaining countries or areas,
however, no difference is found
0 10 20 30 40 50 60 70 80 90 100
in the prevalence of violent Boys
discipline between the sexes. Note: Each dot represents a country.
The prevalence of corporal Source: UNICEF global databases, 2014, based on DHS, MICS and other nationally representative surveys, 2005-2013.
punishment is also similar for

girls and boys, although more Figure 5.8
pronounced sex differences in
In most countries, corporal punishment is also used on
the experience of this form of

violent discipline show up in a boys and girls to a similar extent
few countries or areas, including Percentage of children aged 2 to 14 years who experienced any physical punish-
ment in the past month, by sex of the child
both Costa Rica and Ukraine,
where boys are around one 100
and a half times more likely to 90
be subjected to any physical
80
punishment than girls (Figure
5.8). 70
60
Girls are more likely Boys are more likely
Findings show that violent
Girls
50 to experience physical to experience physical

discipline is more commonly punishment punishment
40
used on younger children,
particularly during middle 30
childhood, than on adolescents.11 20

Indeed, comparable data confirm 10
that, overall, violent discipline
peaks among children aged 5 0 10 20 30 40 50 60 70 80 90 100
to 9 years and falls among older Boys
children aged 10 to 14 years Note: Each dot represents a country.
(Table 5.2). Source: UNICEF global databases, 2014, based on DHS, MICS and other nationally representative surveys, 2005-2013.
102
Table 5.2
Overall, middle childhood (ages 5 to 9) is the period when children are most likely to be
disciplined in a violent manner
Percentage of children aged 2 to 14 years who experienced any violent discipline (psychological aggression and/or physical punish-
ment) in the past month, by age of the child
2 to 4 5 to 9 10 to 14 2 to 4 5 to 9 10 to 14
years old years old years old years old years old years old
Afghanistan 63 78 78 Jordan 90 93 88
Albania 72 81 75 Kazakhstan 48 53 47
Algeria 85 90 87 Kyrgyzstan 49 53 55
Argentina 73 75 69 Lao People's

69 80 76
Democratic Republic
Armenia 67 73 69
Lebanon 85 85 78
Azerbaijan 75 80 75
Liberia 86 93 89
Barbados 78 76 72
Mauritania 85 89 86
Belarus 68 66 60 Mongolia 45 48 44
Belize 71 72 69 Montenegro 65 66 59
Bosnia and Herzegovina 55 59 53 Morocco 89 93 90
Burkina Faso 77 87 81 Niger 79 84 81
Cameroon 92 94 93 Nigeria 89 92 91
Central African Republic 90 94 92 Panama 45 46 44
Chad 79 86 87 Republic of Moldova 76 78 73
Congo 88 89 83 Saint Lucia 69 66 68
Serbia 70 70 62

Costa Rica 56 46 38
Sierra Leone 74 83 85
Côte d'Ivoire 89 91 92
State of Palestine 92 94 91
Democratic Republic
91 94 90
of the Congo Suriname 87 87 85
Djibouti 65 74 74 Swaziland 90 92 86
Dominican Republic 70 70 64 Syrian Arab Republic 84 91 88
Egypt 88 94 91 Tajikistan 69 81 79
Gambia 86 92 92 The former Yugoslav

69 73 66
Republic of Macedonia
Georgia 69 73 62
Togo 91 95 92
Ghana 94 95 91
Trinidad and Tobago 83 79 74
Guinea-Bissau 75 84 83 Tunisia 95 95 91
Guyana 76 79 74 Ukraine 53 63 66
Haiti 86 89 80 Vanuatu 86 86 79
Iraq 76 82 78 Viet Nam 74 76 72
Jamaica 85 86 83 Yemen 91 97 95
103
Figure 5.9
The youngest children are as likely as older children to be exposed to physical violence
Percentage of children aged 2 to 14 years who experienced any discipline in the past month, by type and by age of the child
Shouted, yelled or screamed at

2 to 4 years old
Spanked, hit or slapped with bare hand 5 to 9 years old
10 to 14 years old
Shook
Hit/slapped on hand, arm or leg
Called dumb/lazy
Hit on the bottom or elsewhere with object
Hit/slapped on face, head or ears
Beat with an object/beat as hard as one could
Explained why behaviour was wrong
Took away privileges
Gave something else to do
0 10 20 30 40 50 60 70 80 90 100
Note: This is a weighted average based on comparable data for 54 countries or areas.
Figure 5.10
In Panama, the use of violent discipline is not uncommon even among children as young as
1 year of age
Percentage of children aged 1 year who experienced any discipline in the past month, by type, in Panama
50
40
30
20
10
0.3 0
0
Gave Explained Took away Spanked, Shouted, Hit/ Called Shook Hit on the Hit/ Beat with an
something why privileges hit or yelled or slapped on dumb/lazy bottom or slapped on object / beat
else to do behaviour slapped with screamed hand, arm elsewhere face, head as hard as
was wrong bare hand at or leg with object or ears one could
Source: MICS, 2013.
104
The choice to use particular disciplinary methods as children from poorer households. In the rest of
may be influenced by a child’s age and stage of the countries, the poorest children are more likely
development. For instance, with older children to experience violent discipline than their richest
caregivers may choose to use non-violent peers, although differences overall are quite small.
techniques that rely on communication, such as However, in countries including Albania, Costa
explaining why a behaviour is wrong or revoking a
Rica, Montenegro and Viet Nam, such differences
child’s privileges, because they may perceive such
are more pronounced. Exceptions to this general
practices to be more appropriate at later ages.
pattern include Burkina Faso, Liberia, Mongolia and
Caregivers may also rely more on direct physical
means to control or correct children’s behaviour Yemen, where slightly more rich households report
at younger ages because they cannot think of use of violent disciplinary methods than the poorest
alternative non-violent methods. In reality, research households.
suggests that non-violent methods are suitable even
with young children and can be used effectively to
achieve desired behaviours and teach children life- Figure 5.11
long skills.12 Violent forms of discipline experienced
Children from both rich and poor households
at a young age can be particularly harmful, given the
experience violent forms of discipline
increased potential for physical injuries as well as
children’s inability to understand the violence and Ratio of children aged 2 to 14 years who experienced any
violent discipline (psychological aggression and/or physical
adopt coping strategies to alleviate their distress. punishment) in the past month, by wealth quintiles (poorest
over richest)
Overall, the data indicate that the methods used
1.50
vary only slightly depending on children’s age. The
youngest children are as likely as older children
to be exposed to all forms of physical violence, 1.40
including harsh forms of corporal punishment

(Figure 5.9), while non-violent methods of discipline 1.30
Children from the poorest households

are somewhat less commonly used. In Panama, are more likely to experience violent
discipline
where data on discipline were collected for children 1.20
as young as 1 year of age, the use of corporal
punishment is not unusual, even among the 1.10
youngest children: One in six Panamanian infants
were spanked or hit with a bare hand and an almost
1.00
equal proportion were exposed to yelling and
screaming (Figure 5.10).
0.90
Children from the richest households
are more likely to experience violent
Violent discipline and discipline
0.80
the economic and social
status of households
0.70
Research has revealed an association between
family wealth and parenting practices in certain 0.60
contexts.13 However, comparable data from a
range of countries or areas indicate that the use of
0.50
violent disciplinary practices is not systematically
associated with lower economic and social status
Notes: Each dot represents a country. A ratio of 1.0 (0.95-1.04, grey band) indicates that
(Figure 5.11). In approximately half of countries with violent discipline levels in the two groups (poorest children/richest children) are equal.
Source: UNICEF global databases, 2014, based on DHS, MICS and other nationally repre-
available data, children from wealthier households sentative surveys, 2005-2013.
are equally likely to experience violent discipline
105
Box 5.2
Reports of violent punishment from

other selected countries
Below are key findings on the slapping the child on the face, head Canada
prevalence of violent discipline in or ears and/or beating the child re-
The 2008 Canadian Incidence Study
selected countries. The informa- peatedly with an implement.
of Reported Child Abuse and Ne-
tion presented is not meant to be
Across the overall sample, more glect15 is the third national study
exhaustive but is intended to illus-
trate what is known about corporal than half of all children were report- in that country to examine the in-
punishment and other forms of vio- edly subjected to mild physical pun- cidence of reported child maltreat-
lent discipline in a selection of ad- ishment by at least one parent in the ment and the characteristics of chil-
ditional countries. past month – including 54 per cent dren and families investigated by
of girls and 58 per cent of boys. The child welfare authorities. According
use of severe corporal punishment to the study, an estimated 18,688
China, Colombia, Italy, Kenya, the
was reported less often. Across the cases of physical abuse were sub-
Philippines and Thailand
entire sample, about one in seven
stantiated in 2008, representing a
Through the 2008-2009 Parenting children experienced severe corpo-
rate of 3.1 cases per 1,000 children.
across Cultures Project, mothers ral punishment at the hands of at
and fathers representing a total of In the vast majority of these cases
least one parent in the last month.
1,417 families were asked about (17,212), physical abuse was the pri-
Rates varied among countries. Re-
their use of corporal punishment on ported use of both mild and severe mary form of maltreatment. Investi-
children aged 7 to 10.14 Drawing on punishment was highest in Kenya, gating child protection workers con-
items from the Parent-Child Conflict where almost all children received cluded that almost three quarters of
Tactics Scale, parents were asked a mild form of physical discipline these cases resulted from attempts
whether, in the past month, they while nearly two thirds were sub- to punish a child. Case workers also
had used any of six forms of corpo- jected to severe corporal punish- estimated that over one quarter of
ral punishment, classified as either ment. As shown in the table below, substantiated emotional abuse inci-
mild or severe. Mild physical pun- boys were more likely to experience dents were initiated within the con-
ishment included spanking, hitting corporal punishment than girls in text of punishment.16
or slapping a child with a bare hand; most countries, but statistically sig-
hitting or slapping on the hand, arm nificant sex differences were only
Austria, France, Germany and Spain
or leg; shaking; and/or hitting the found in China and Kenya (for mild
child with an object. Severe physi- punishment) and Italy (for severe In 2007, a multi-country study gath-
cal punishment included hitting or punishment). ered data on the rates of corporal
punishment in Western Europe.
Study participants included a ran-
Percentage of boys and girls aged 7 to 10 years who experienced mild or severe dom sample of 1,000 parents from
corporal punishment in the past month
each nation. Comparisons across
countries revealed considerably
Mild corporal punishment Severe corporal punishment
higher rates of corporal punish-
Girls Boys Girls Boys
ment in France, where 87 per cent
China 48 60 10 15
of parents reported giving a child a
Colombia 68 63 15 4 slap on the bottom, 72 per cent re-
Italy 61 66 12 23 ported giving a child a ‘mild’ slap on
Kenya 82 97 61 62 the face, and 51 per cent reported
Philippines 71 77 9 8 spanking their child’s bottom with
Thailand 58 72 5 3 their hand.17
106
Percentage of parents aged 25 and older who have ever used corporal punishment Findings from the 2008-2009 Parent-
on at least one child under age 1818 ing across Cultures Project seem to
Spanking Resound- Beating confirm that corporal punishment is
Mild slap Slap on bottom ing slap with Severe
on face bottom with hand on face object beating relatively common.22 Around one in
three children aged 7 to 10 (38 per
Austria 50 62 16 18 4 6
cent of girls and 36 per cent of boys)
Germany 43 68 17 13 5 9 were reportedly subjected to mild
France 72 87 51 32 5 12 corporal punishment (defined as
Spain 55 80 54 31 7 4 spanking, hitting or slapping a child
with a bare hand, hitting or slapping
on the hand, arm or leg, shaking and/
United Kingdom of Great Britain Many forms of verbal aggression21 or hitting the child with an object).
and Northern Ireland were reported as common experi- The use of severe forms of physical
ences by the young adults in both violence (including hitting or slap-
Child Abuse and Neglect in the UK
waves of the study. However, this ping the child on the face, head or
Today19 is a nationally representa-
type of violent discipline was report- ears) was reportedly used on 4 per
tive survey conducted in 2009 with
ed significantly less often in the 2009 cent of girls and 5 per cent of boys.
2,160 parents or guardians of chil-
dren under age 11 and 2,275 young survey than in the 1998-1999 survey,
Similar levels of corporal punish-
people aged 11 to 17 and their par- with the exception of being humili-
ment were found with younger
ents or guardians. Information on ated or embarrassed. Respondents
children. For instance, a survey con-
physical punishment was collected who reported one or more experi- ducted in 2007-2008 collected repre-
from the main caregiver of children ences of verbal aggression were sentative data on the use of spank-
under 18 years of age. The survey in- asked whether it had taken place ing from a sample of 2,946 North
cluded two questions on the use of regularly during their childhood. Carolina mothers of children under
physical punishment – smacking the The survey found that regular and 2 years of age. Data were collected
child’s hand or hitting the child with prolonged verbal aggression de- using selected questions from the
a belt, a wooden implement (such clined significantly over time – from Parent-Child Conflict Tactics Scale.
as a stick or wooden spoon) or other around 15 per cent in 1998-1999 to 6 In the survey, nearly one third (30 per
object – within the last year. Nearly per cent in 2009. cent) of mothers reportedly spanked
42 per cent of parents or guardians their child within the last year and,
in the study reported that they used When it came to the use of physi-
of these, one in nine (11 per cent)
physical punishment on their child cal punishment, being smacked on admitted spanking their child more
within the last year. the bottom, hands, arms or legs than 20 times. The use of spanking
was commonly reported by young increased with the child’s age, and
A sample of 1,897 young adults adults in both waves of the study; mothers of boys were slightly more
aged 18 to 24 was also included to however, only a minority in each likely to report using spanking than
collect retrospective information study reported these as regular oc- mothers of girls (32 per cent and 28
on childhood experiences of both currences during childhood. Again, per cent, respectively).23
non-violent and violent discipline. the young adults in 2009 were con-
In comparing the 2009 sample of siderably less likely to report physi- New Zealand
young adults with those who were cal punishment as a regular experi-
interviewed in an earlier wave of ence than those in the earlier study. The 2006/2007 New Zealand Health
the study (n = 2,869) conducted in Survey24 used a multi-stage, stratified,
1998-1999, little or no difference probability-proportionate-to-size
United States of America
was found in the reported rates of sample design, with over-sampling
non-violent discipline.20 Relatively Several studies on violent discipline of some ethnic groups. The sam-
high proportions of respondents in have been conducted in the United pling design ensured that strong
each study said they experienced States to measure the use of violent national estimates for key health be-
non-violent forms of discipline dur- discipline across a variety of set- haviours and outcomes were gener-
ing their childhoods. tings and children’s ages. ated and that the survey included
107
sufficient numbers of population Findings from the 2008-2009 Parent- reported experiences of psycho-
groups of interest (particularly Mao- ing across Cultures Project appear logical aggression from either their
ri, Pacific and Asian populations) to to confirm that corporal punishment mother or father within the last year.
be able to generate accurate esti- is relatively uncommon in Sweden.27 Girls were significantly more likely
mates for all groups. Parents were Less than 10 per cent of children to experience psychological aggres-
asked about the use of various forms aged 7 to 10 (9 per cent of girls and sion than boys (43 per cent versus
of child discipline with children up to 6 per cent of boys) were reportedly 33 per cent, respectively). The rates
age 14 in the preceding four weeks. subjected in the past month to mild of mild physical violence (6 per cent)
corporal punishment (defined as were considerably lower than rates
The survey found that physical pun- spanking, hitting or slapping a child of psychological aggression, with
ishment was one of the least used with a bare hand, hitting or slap- no significant differences due to the
forms of discipline in the country. ping on the hand, arm or leg, shak- sex of the parent. However, girls
One in 10 children under age 14 had ing and/or hitting the child with an were significantly more likely than
experienced physical punishment object), and no parents in Sweden boys to suffer from mild physical
by their primary caregiver in the last reported the use of severe forms of punishment (8 per cent compared to
four weeks. After adjusting for age, physical violence (including hitting 4 per cent, respectively). The rates of
Pacific boys were almost twice as or slapping the child on the face, severe physical violence were simi-
likely to have been physically pun- head or ears and/or beating the child lar for boys and girls: 2 per cent of
ished compared to boys in the total repeatedly with an implement).
adolescents of both sexes reported
population.
experiences of severe physical vio-
Low levels of violent discipline
Overall, children aged 2 to 4 years were confirmed by the 2007 study lence at the hands of a parent.
were the most likely to experience conducted in Western Europe,
physical punishment – 19 per cent discussed above.28 Comparisons Finland
of children in this age group had ex- across the five countries surveyed The Finnish Child Victim Survey30
perienced physical punishment by showed significantly lower rates was conducted in 2008 by the Police
their primary caregiver in the previ- of corporal punishment in Sweden College of Finland. The approach
ous four weeks. One in 14 children than in any of the other four nations.
to sampling used was a stratified
(7 per cent) under 2 years of age ex- About 17 per cent of Swedish par-
cluster sample based on ”county,
perienced such methods. ents reported giving a child a slap
the quality of the municipality, and
on the bottom, 14 per cent reported
the size of the school.” The original
Sweden giving a child a ‘mild’ slap on the
sample included ninth-graders in
face and 4 per cent reported spank-
Sweden became the first country 184 schools; of these, 161 schools
ing their child’s bottom with their
to prohibit corporal punishment of participated (88 per cent) with a to-
hand. The use of more severe forms
children when it added the follow- tal of 5,807 respondents. The rep-
of corporal punishment (including
ing provision to its Parenthood and resentative nature of the final data
severe beating, a ‘resounding’ slap
Guardianship Code in 1979: “Chil- was checked by comparing basic
on the face and beating the child
dren are entitled to care, security social and economic factors with
with an implement) was reported
and a good upbringing. Children other representative studies in the
by less than 5 per cent of parents.
are to be treated with respect for country.
their person and individuality and
Denmark According to this survey, 51 per cent
may not be subjected to corporal
punishment or any other humiliat- The Danish Youth Health Survey, of Finnish girls experienced psycho-
ing treatment.”25 The law – which conducted in 2008 by the National logical aggression by a parent in the
was intended to be primarily educa- Institute of Public Health, University past 12 months compared to 25 per
tional rather than punitive in nature of Southern Denmark, was based on cent of boys. Twelve per cent of girls
– was followed by a large-scale me- a nationally representative sample experienced mild physical violence
dia campaign. A few studies have of ninth-grade students in the coun- and 2 per cent of girls were subject-
tried to assess the effectiveness of try’s mandatory school system.29 ed to severe physical violence by a
these initiatives in reducing the use The study’s findings showed that parent compared to 5 per cent and 1
of violent methods.26 38 per cent of Danish adolescents per cent of boys, respectively.
108
“ ... There is no ambiguity: ‘All forms

of physical or mental violence’ does
not leave room for any level of
legalized violence against children.
Corporal punishment and other cruel
or degrading forms of punishment
are forms of violence and the State
must take all appropriate legislative,
administrative, social and educational
measures to eliminate them.”
United Nations Committee on the Rights of
the Child, General Comment No. 8 on the
Convention on the Rights of the Child
© UNICEF/NYHQ2010-2686/LeMoyne
109
Only 8% of children worldwide
prohibited corporal punishment in all
2 billion children without
Corporal punishment fully prohibited at home

Corporal punishment partially prohibited at home
Albania
Corporal punishment not prohibited at home Corporal punishment is defined as “any form of
punishment resorting to the use of force aimed
to cause pain or suffering, even in the slightest
extent, by parents, siblings, grandparents, legal Turkmenistan
representative, relative or any other person
Sweden legally responsible for the child”. “Humiliation of the dignity
Children are entitled to care of the child, intimidation,
Germany - Article 3(f) of the 2010 Law on the Protection corporal punishment, other
and security, to be treated of the Rights of the Child
“Children have the right to a non-violent upbringing. with respect, and “may not be physical abuse harmful for
Corporal punishment, psychological injuries and other subjected to corporal punish- the child’s mental or physical
humiliating measures are prohibited.” ment or any other humiliating health are inadmissible.”
- Article 1631 of the 2000 amendment to the Civil Code treatment”. - Article 85(2) of the 2012
- Article 6.1 of the 1979 Family Code
amendment to the Parenthood
and Guardianship Code
home
Honduras
Amendment prohibits the use of “physi-
cal punishment or any type of humiliating,
degrading, cruel or inhuman treatment
as a form of correction or discipline of Togo
children or adolescents [...]”. The State
will also ensure awareness education “Physical and psycho-
programmes to promote “positive, logical abuse, corporal
participative and non-violent forms of punishment, deprivation
discipline”. of care or withholding of
food are punished by the
- 2013 amendment to Article 191 of the penalties provided [...]” South Sudan
1984 Family Code in the law.
“Every child has the right
- Article 357 of the 2007 [...] to be free from corporal
Children’s Code punishment and cruel and
inhuman treatment by any
person including parents,
Brazil Israel school administrations and Japan
Uruguay other institutions.”
“Children and adolescents are “It is prohibited [...] to use physical or “Corporal punishment of The use of corporal punish-
entitled to be educated and cared for children [...] is entirely imper- - Article 17(1) of the 2011 ment in the home is only
any other kind of humiliating punish- Transitional Constitution
without the use of physical punishment as a form of correcting or missible, and is a remnant of prohibited in Kawasaki City
ment
Notes: This map or cruel
is stylized andor degrading
not to scale. Ittreatment disciplining
does not reflect a position by children
UNICEF on the or adolescents.”
legal status a societal-educational
of any country or territory or the delimitation of outlook byand
any frontiers. The final boundary between the Republic of Sudan local
theordinance: “Parents
Republic of South Sudan has not y
as forms of correction, discipline, that has lost its validity. Such shall not inflict abuse and
education or any other pretext [...]” - Article 12bis of the 2004 Children punishment injures his or her corporal punishment on
(unofficial translation) and Adolescent Code body, feelings, dignity and their child.”
- Article 18-A of the 2014 amend- proper development.” - Article 19 of the 2000
ment to the 1990 Code on Children - 2000 Supreme Court judge- Kawasaki City Ordinance
and Adolescents ment against a defence for on the Rights of the Child
corporal punishment included
in the 1944 Civil Wrongs
Ordinance
MOST CORPORAL PUNISHMENT OCCURS AT HOME – THE SETTING

WHERE IT IS LEAST LIKELY TO BE PROHIBITED BY LAW
1979 1983 1987 1989 1994 1997 1998 1999 2000 2000 2000 2002 2003 2004 2004 2005 2006 2007 2007
Sweden
Finland
Norway
Austria
Cyprus
Denmark
Latvia
Croatia
Bulgaria
Israel
Germany
Turkmenistan
Iceland
Ukraine
Romania
Hungary
Greece
Netherlands
New Zealand
Only 39 countries have prohibited corporal

110 punishment in all settings, including the home
live in countries that have fully
settings, leaving slightly more than
full legal protection
Corporal punishment fully prohibited in all other settings

Corporal punishment fully or partially prohibited in at least one other setting
Corporal punishment not prohibited in any other settings
Other
settings
(alternative care,
day care, penal
institutions and
sentences for crime)
Corporal punishment fully prohibited at school

Corporal punishment partially prohibited at school
Corporal punishment not prohibited at school
Notes: This map is stylized and not to scale. It does not reflect a position by UNICEF on the legal status of any country or territory or the delimitation of any frontiers. The final boundary between the Republic of Sudan and the Republic of South Sudan has not
school
Notes: These maps are stylized and not to scale. They do not reflect a position by UNICEF on the legal status of any country or territory or the delimitation of any frontiers. The dotted line represents approximately the
Line of Control in Jammu and Kashmir agreed upon by India and Pakistan. The final status of Jammu and Kashmir has not yet been agreed upon by the parties. The final boundary between the Sudan and South Sudan has
not yet been determined. The final status of the Abyei area has not yet been determined.
Source: Global Initiative to End All Corporal Punishment of Children <www.endcorporalpunishment.org>
Notes: This map is stylized and not to scale. It does not reflect a position by UNICEF on the legal status of any country or territory or the delimitation of any frontiers. The final boundary between the Republic of Sudan and the Republic of South Sudan has not y
2007 2007 2007 2007 2007 2008 2008 2008 2008 2010 2010 2010 2010 2010 2011 2013 2013 2014 2014 2014
Portugal
Uruguay
Venezuela (Bolivarian
Republic of)
Spain
Togo
Costa Rica
Republic of Moldova
Luxembourg
Liechtenstein
Poland
Tunisia
Kenya
Congo
Albania
South Sudan
The former
Yugoslav Republic
of Macedonia
Honduras
Malta
Brazil
The Plurinational
State of Bolivia
111
Chapter 6 When
children
hurt one
another
Friendship is an important component of a child’s
social life at every age.1 Adolescents in particular
often rely more heavily on peers than parents for
support and interaction.2 In fact, some researchers
have estimated that teenagers spend about one
third of their waking hours with friends.3 Not
surprisingly, research also indicates that having
friends contributes to a child’s sense of well-being,
greater social competence and self-esteem.4
The quality of peer relations is also important since

it seems that having problematic friendships is
more detrimental to well-being than the absence of
supportive ones.5 In general, research has shown
that having quality friendships that offer support
and intimacy can contribute to better outcomes in
school involvement and achievement and in overall
adjustment during adolescence.6 On the other hand,
friends who are antisocial or who have behavioural
problems might provide each other with support, but
the interactions among them may not be positive.
Studies have shown, for example, that exposure
to depressed peers increases adolescents’ reports
of depressive symptoms in themselves,7 and that
adolescents with antisocial friends show comparable
levels of depression to adolescents without friends.8
Furthermore, several studies have found that
adolescents with antisocial peers are more likely to
engage in delinquent behaviour and are at higher risk
for conduct disorders than those without antisocial

peers.9 Evidence also suggests that friendships with

aggressive peers can heighten children’s aggressive
behaviour towards other children.10 Children who
are victimized by their peers are at heightened risk
for a wide range of emotional problems, including
depression, loneliness, social anxiety, diminished
self-worth and increased risk of suicide.11
112
Although peer interaction takes place in a variety Reports of physical attacks among students aged
of settings, school is the predominant site in which 13 to 15 are relatively common in the 25 countries
most children form friendships and establish peer with comparable data from the Global School-
groups.12 Schools, therefore, are typically the settings based Student Health Survey (GSHS), ranging
for peer violence, along with routes to and from from around 20 per cent in the former Yugoslav
school. Peer violence can take many forms, including Republic of Macedonia and Uruguay to over 50
physical attacks, fighting and bullying. Children can per cent in Botswana, Djibouti, Egypt, Ghana, the
experience such violence directly, as either victims United Republic of Tanzania (Dar es Salaam) and
and/or perpetrators, or indirectly, as witnesses. Yemen (Figure 6.1).13 In most countries, boys are
significantly more likely than girls to report having
Physical attacks been the victim of a physical attack at least once in
Because adolescence tends to be a time of intense the past year. In Libya, Tunisia and Yemen, boys are
emotions, both positive and negative, teenagers’ more than twice as likely as girls to report a physical
feelings of anger, jealousy, frustration or humiliation attack. That said, the proportions of physical attacks
may drive them to physically attack their peers. among both boys and girls are relatively high for
A physical attack is characterized by the use of all countries (Figure 6.2). In Botswana, Egypt,
force against a passive recipient. In some cases, Ghana and the United Republic of Tanzania (Dar
an attack may be provoked by something that was es Salaam), the proportions of students who have
said or done by the victim; in others, it could be been physically attacked are above 50 per cent for
unprompted. Such attacks can be carried out by both sexes.
an individual or a group, with or without the use of
weapons (see Box 6.1).
Figure 6.1
Many adolescents report being victims of physical attacks

Percentage of students aged 13 to 15 years who reported being physically attacked one or more times in the past 12 months
100
90
80
70
CHAPTER 6: When Children Hurt One another

60
50
40
30
20
10
0
Ghana
Egypt
Djibouti
Botswana
(Dar es Salaam)
Yemen
Sri Lanka
Tunisia
Grenada
Trinidad and Tobago
Indonesia
Saint Vincent and
the Grenadines
Jordan
Ecuador
Saint Lucia
Thailand
Kuwait
Libya
Mauritius
Colombia
Argentina
Tajikistan
Myanmar
The former Yugoslav
Uruguay
Notes: Data for Colombia and Ecuador are not national but have been recalculated on the basis of subnational surveys that took place in selected cities in each country. Data for the United Republic
of Tanzania are not national but represent only the city of Dar es Salaam.
Source: GSHS, 2005-2011.
113
with age, peaking at 24 per cent around ages 10 to
Figure 6.2
13. By the time children reached mid-adolescence
In almost all countries, boys are significantly (ages 14 to 17), around 4 in 10 reported having
more likely to report being physically been physically assaulted by a peer during their
attacked than girls lifetime. As was the case in low- and middle-income
countries, male children were more likely to report
Percentage of students aged 13 to 15 years who reported
being physically attacked one or more times in the past 12 being physically assaulted by a peer in the past year
months, by sex than their female counterparts (23 per cent and 13
100 per cent, respectively). There were no statistically
significant changes in either the lifetime or last year
90
exposure to peer assaults between the two waves
80
of the NatSCEV study (2008 and 2011), suggesting
70 that the likelihood of experiencing peer violence has
60 remained relatively unchanged over a three-year
Girls are more likely to be Boys are more likely to period in the United States.
50 physically attacked be physically attacked
40
Fighting
Girls
30
In contrast to physical attacks, fighting generally
20 involves conflict between two or more persons in
10 which the distinction between perpetrators and
victims is not always clear-cut.19 In some instances,
0 10 20 30 40 50 60 70 80 90 100 both parties may have instigated or chosen to
Boys
participate in the fight, while in others, one person
Note: Each dot represents a country. may be fighting back in self-defence.
The literature on the subject suggests that children

In a 2012 nationally representative survey of who are involved in fighting are more likely than
secondary school students in South Africa,14 around those who are not to report a lack of perceived
6 per cent of respondents reported having been parental support in relation to school. They also
physically attacked or hurt at school by someone report greater difficulties establishing close peer
(using any kind of weapon or their hands) in the past relationships, poor emotional health, less parental
year as compared to about 4 per cent who reported supervision, feelings of alienation from school and
a physical attack at school in a comparable 2008 low academic success.20
survey.15 There was no significant difference in the
Available data from a large cross-section of countries

2012 reported rates of physical attacks between
reveals that fighting among adolescents is a common
male and female secondary school students (7 per
occurrence (Map 6.1).21 Anywhere from 14 per cent
cent and 6 per cent, respectively).
of adolescents aged 13 to 15 in Cambodia to 68 per
cent in Samoa reported engaging in a physical fight
Data from the United States also confirm the in the past 12 months. In a majority of countries
widespread exposure of children to physical attacks. with available data, between 30 and 40 per cent of
The second wave of the National Survey of Children’s adolescents aged 13 to 15 reported having been
Exposure to Violence (NatSCEV II),16 conducted in involved in a physical fight. This includes a mixture
2011, revealed that more than one in four children of both low- and middle-income countries, including
(28 per cent) under age 17 reported having been Iraq, Pakistan, Peru and Thailand, and high-income
physically assaulted17 at some point in their lives by countries, such as Canada, France and the United
a non-sibling peer; 18 per cent reported having been States.22 More than half of adolescents reported
victimized in the last year.18 The likelihood of physical involvement in a physical fight in countries as diverse
assaults by peers within the past year increased as Djibouti, Mauritania, Samoa and Yemen.
114
Box 6.1
Carrying weapons
Findings from a few national sur- former Yugoslav Republic of Mac- such as a gun, knife or club at least
veys suggest that it is not at all un- edonia, students aged 13 to 15 were once in the month preceding the
common for adolescents to have ac- asked about the frequency with survey. The rates of weapon car-
cess to weapons and to feel a need which they carried a weapon such rying were higher among males
to arm themselves. as a gun, knife or club within the (28 per cent) than females (8 per
past 30 days. Five per cent of the cent). The same survey found that
In a 2012 survey conducted in New respondents reported weapon car- 5 per cent of students (8 per cent
Zealand, around 3 per cent of sec- rying during the past month, with a of boys and 3 per cent of girls) re-
ondary school students admitted to higher percentage among male chil-
ported having carried a weapon
carrying a weapon within the last 12 dren (9 per cent) than their female
on school property in the previous
months with the thought of harming counterparts (1 per cent).
month. The prevalence of having
someone. Boys were significantly
A nationwide survey conducted in carried a weapon on school prop-
more likely to admit to weapon car-
rying than girls (5 per cent and 2 per the United States in 2013 revealed erty decreased from 12 per cent in
cent, respectively).23 that around one in six students (18 1993 to 6 per cent in 2003 but has
per cent) in grades 9 through 12 not changed significantly since that
In the 2007 GSHS conducted in the reported having carried a weapon time.24
Map 6.1
Physical fighting is a common occurrence among adolescents in many countries

Percentage of adolescents aged 13 to 15 years who reported being in a physical fight one or more times during the past 12
months, by country
Above 50% CHAPTER 6: When Children Hurt One another

41% - 50%
30% - 40%
Less than 30%
Countries with no comparable data
from either the HBSC or GSHS
Notes: This map is stylized and not to scale. It does not reflect a position by UNICEF on the legal status of any country or territory or the delimitation of any frontiers. The dotted line represents approxi-
mately the Line of Control in Jammu and Kashmir agreed upon by India and Pakistan. The final status of Jammu and Kashmir has not yet been agreed upon by the parties. The final boundary between the
Sudan and South Sudan has not yet been determined. The final status of the Abyei area has not yet been determined. Data from the HBSC were recalculated as weighted averages for 13- to 15-year-olds to
allow for comparison with data collected in the GSHS. Data for China, Colombia, Ecuador, State of Palestine, the Bolivarian Republic of Venezuela and Zimbabwe are not national but have been recalculated
on the basis of subnational surveys that took place in selected cities in each country or area. Data for Belgium are a weighted average of the Flemish and French samples. Data for Ghana refer to students in
junior high school only. Data for the United Kingdom are a weighted average of the samples in England, Scotland and Wales. Data for the United Republic of Tanzania are not national but represent only the
city of Dar es Salaam.
Source: HBSC, 2009/2010 and GSHS, 2004-2013.
115
Significant rates of involvement in Across all regions, boys engage in fighting to a far
serious physical fights have also been greater extent than girls
recorded in surveys conducted in
additional countries. According to the Figure 6.3A
2012 New Zealand Youth Survey,25
14 per cent of secondary school Percentage of students aged 13 to 15 years who reported being in a physical
students reported involvement in a fight one or more times during the past 12 months, by sex, in countries or
areas in the Middle East and North Africa
serious physical fight in the 12 months
Tunisia
preceding the survey. This represents
Lebanon
a decline from the rate of physical Yemen
fighting reported in the 2001 (21 per Djibouti Boys Girls
Jordan
cent) and 2007 (20 per cent) surveys.26
United Arab Emirates
Qatar
A number of studies of sex differences Egypt
Algeria
in aggression have led to the well-
Kuwait
documented finding that males tend State of Palestine
to be more physically aggressive Morocco
than females.27 Some researchers Libya
Oman
emphasize biological or evolutionary Iraq
reasons for this,28 while others suggest 0 10 20 30 40 50 60 70 80 90 100
the importance of considering how Note: Data for the State of Palestine are not national but have been recalculated on the basis of subnational surveys
social, cultural and environmental that took place in selected cities.
factors influence aggression.29
While a more traditional viewpoint Figure 6.3B

might expect boys to engage more
Percentage of students aged 13 to 15 years who reported being in a physical
in fighting as they seek to live up to fight one or more times during the past 12 months, by sex, in countries in
male stereotypes of being powerful Latin America and the Caribbean
and strong, the importance of Jamaica
environmental factors in understanding Saint Vincent and the Grenadines
a male propensity for aggression and Antigua and Barbuda
Grenada Boys Girls
violence has also been underscored.
Saint Lucia
Some of the literature suggests, Peru
in fact, that the use of violence by Guyana
Barbados
males may serve both a survival and

Dominica
status function, particularly for young Ecuador
men in certain cultural or low-income Venezuela (Bolivarian Republic of)
settings.30 Bolivia (Plurinational State of)
Colombia
Trinidad and Tobago
When it comes to fighting and Saint Kitts and Nevis
differences between the sexes, Argentina
Belize
available data confirm that adolescent Chile
boys engage more in physical fighting Uruguay
than girls in low-, middle- and high- Honduras
Costa Rica
income countries. In all the countries Suriname
in the Middle East and North Africa and 0 10 20 30 40 50 60 70 80 90 100
Latin America and the Caribbean with
available data from the GSHS, boys Note: Data for Colombia, Ecuador and the Bolivarian Republic of Venezuela are not national but have been recalcu-
lated on the basis of subnational surveys that took place in selected cities in each country.
are significantly more likely to report Source: GSHS, 2003-2013.
116
Figure 6.3C

fight one or more times during the past 12 months, by sex, in countries in
Samoa
Mongolia
Vanuatu
Fiji Boys Girls
Solomon Islands
Nauru
Tonga
Indonesia
Thailand
Cook Islands
Philippines
Kiribati
Malaysia
Viet Nam
China
Myanmar
Cambodia
0 10 20 30 40 50 60 70 80 90 100
Note: Data for China are not national but have been recalculated on the basis of subnational surveys that took place
in selected cities.
Figure 6.3D

fight one or more times during the past 12 months, by sex, in countries in
Namibia
Botswana Boys Girls
Kenya

Zambia
Mauritius
Zimbabwe
United Republic of Tanzania (Dar es Salaam)
Uganda
© UNICEF/BANA2013-01066/Kiron
Swaziland
Malawi
0 10 20 30 40 50 60 70 80 90 100
Notes: Data for Zimbabwe are not national but have been recalculated on the basis of subnational surveys that took
place in selected cities. Data for the United Republic of Tanzania are not national but represent only the city of Dar es
Salaam.
117
fighting than girls, a pattern that generally holds true were reportedly involved in physical fights in the
across other regions (Figures 6.3A-6.3D). That said, past year (Figure 6.3B).
girls still engage in physical fights to a noticeably
high extent. In both Djibouti and Oman, over 40 per Physical fighting is also far more common among
cent of girls aged 13 to 15 reported involvement in adolescent boys than girls in European countries,
a physical fight at least once in the last year (Figure Canada, the Russian Federation and the United States
6.3A). In some Caribbean countries, including (Maps 6.2A and 6.2B). Girls of all ages (11, 13 and
Antigua and Barbuda, Jamaica, and Saint Vincent 15) surveyed in the latest HBSC were significantly
and the Grenadines, close to 4 in 10 adolescent girls less likely than boys to report fighting in nearly all of
Physical fighting is also far more common among boys than girls in European countries,
Canada, the Russian Federation and the United States
Map 6.2A
Percentage of boys aged 13 to 15 years who reported being in a physical fight one or more times during the past 12 months in
European countries, Canada, the Russian Federation and the United States
Above 50% Less than 20%

31% - 50% Countries that did not
20% - 30% participate in the HBSC
Map 6.2B
Percentage of girls aged 13 to 15 years who reported being in a physical fight one or more times during the past 12 months in
European countries, Canada, the Russian Federation and the United States
Above 50% Less than 20%

31% - 50% Countries that did not
20% - 30% participate in the HBSC
Notes: These maps are stylized and not to scale. They do not reflect a position by UNICEF on the legal status of any country or territory or the delimitation of any frontier. Data were recalculated as
weighted averages for 13- to 15-year-olds. Data for Belgium are a weighted average of the Flemish and French samples. Data for the United Kingdom are a weighted average of the samples in England,
Scotland and Wales.
Source: HBSC, 2009/2010.
118
the 37 countries with available data. Among 13- to countries with available data for boys and in a few
15-year-olds, the prevalence of fighting ranges from countries for girls, although differences between
a low of 31 per cent for boys and 11 per cent for the age groups were relatively small (Table 6.1).32
girls in Germany to highs of 80 per cent for boys Among students in the United States, the 2013
in Armenia and 34 per cent for girls in Greece. The Youth Risk Behavior Survey (YRBS) suggested that
highest rates of physical fighting among boys were the proportions of those involved in physical fighting
found in Armenia, the Czech Republic and Greece, were higher among children in the 9th and 10th
where the proportions are above 60 per cent. Rates grades (28 per cent and 26 per cent, respectively)
of involvement in physical fights among girls are than among those in the 11th and 12th grades (24
per cent and 19 per cent, respectively).33 Finally,
highest in Greece, Hungary, Slovakia and the United
in New Zealand, reported rates of engagement in
States, where at least one in four adolescent girls
serious physical fights were found to be significantly
were involved in a fight within the past year.
higher among 14- and 15-year-old students than
older adolescents (those aged 17 and above).34
In the 2012 New Zealand Youth Survey, physical
fights were found to be more common among
Bullying
boys than girls in secondary school, with nearly
one in five boys compared to one in 10 girls Bullying refers to the use of aggression to assert
reporting engagement in serious physical fights in power over another person.35 More specifically, it
the past year.31 has been defined by researchers as actions, either
physical or verbal, that have a hostile intent, are
repeated over time, cause distress for the victim and
In approximately one third of countries with available
involve a power imbalance between the perpetrator
data, the prevalence of fighting appears to decline
and victim.36 As social dynamics have shifted over
with age. Results from the HBSC found that fighting
time, and with the growing use of information and
declined between ages 11 and 15 in most of the 37
communication technologies such as the Internet
and cell phones, children are increasingly exposed
to new forms of bullying (see Box 6.2).37
Table 6.1
In some countries, involvement in physical A growing body of literature examines the

fights tends to decline with age prevalence, risk factors and impact of bullying on
both victims and perpetrators. However, much of
Percentage of adolescents aged 11, 13 and 15 years who the available evidence is derived from research
reported being in a physical fight one or more times during
the past 12 months in countries where there is a difference conducted in the Western world.38
of at least 10 percentage points in the prevalence of physical

fights between 11- and 15-year-olds While research into the individual risk factors that
11-year-olds 13-year-olds 15-year-olds
lead to bullying has highlighted a variety of possible
causes, a few factors have consistently been
Belgium 48 39 36
found to predict the likelihood that an adolescent
Czech Republic 49 48 39
or younger child will bully others: Those who have
Denmark 38 31 22
been maltreated by caregivers are significantly more
Estonia 35 30 25
likely to bully others, particularly those who have
France 40 36 30
experienced physical or sexual abuse.39 Witnessing
Germany 28 23 18
parental physical abuse or domestic violence has
Hungary 48 47 36
also been documented as a strong risk factor for
Iceland 40 33 22
bullying.40 In addition, research has identified
Latvia 53 43 35
hyperactivity-impulsiveness, low self-control and
Poland 44 34 30
attention deficit hyperactivity disorder as strong
Slovenia 46 47 34
predictors of bullying. Children who bully tend to
Source: HBSC, 2009/2010. have weak inhibitions against aggression and are
119
significantly more likely than victims of bullying to On the other hand, numerous studies have also
exhibit anger.41 found a strong relationship between bullying others,
increased depressive symptoms and thoughts
Some research suggests that boys are more likely of suicide.54 Bullying has been linked to future
to bully others than girls42 and are more likely to use engagement in juvenile delinquency, including theft
physical violence and threats.43 Girls, on the other and robberies, vandalism, arson, physical attacks,
hand, seem more prone to psychological/relational gang involvement and the selling of drugs.55
forms of bullying, which involve actions such Children who bully others also report increased
as excluding others or spreading rumours.44 For rates of risky behaviours, including smoking and
drinking,56 fighting, being injured in physical fights
instance, in a national survey in Malta, researchers
and carrying weapons.57
found that 61 per cent of boy bullies reported
bullying others with physical violence, compared to
Bullying is a problem worldwide. It exists at some
30 per cent of girl bullies. In contrast, 43 per cent
level and in some form in every country. Available
of girl bullies reported isolating others (not talking
data from 106 countries collected through the HBSC
to them), compared to 26 per cent of boy bullies.45
and GSHS show that the proportions of adolescents
aged 13 to 15 who say they have recently experienced
Many individual risk factors for being bullied have bullying ranges from 7 per cent in Tajikistan to 74
also been identified. Children who are bullied are per cent in Samoa (Map 6.3). Among the high-
often marginalized by their peers for a wide variety income countries with available data, proportions of
of reasons. Risk factors include not having many reported bullying range from 9 per cent in Italy to
friends (particularly those who can be trusted) and 52 per cent in Lithuania. In 14 of the 67 low- and
loneliness.46 Particular groups of children, such as middle-income countries with available data, more
ethnic minorities and those with disabilities, can be than half of the student population said they recently
especially vulnerable to bullying.47 Teenagers may experienced bullying. These adolescents come from
also be targeted because of their sexual orientation diverse parts of the world, from small Pacific islands
(see Box 6.4). For example, one study in the United such as Vanuatu to large African nations including
Kingdom found that between 30 and 50 per cent Kenya. On the flip side, a significant proportion (31
of adolescents in secondary schools who were per cent) of teens in Europe and North America
attracted to the same sex experienced homophobic admitted to having bullied others, with prevalence
bullying.48 ranging from around one in seven (14 per cent) in
the Czech Republic and Sweden to nearly 6 in 10 (59
Research highlights a wide range of negative long- per cent) in Latvia and Romania (Figure 6.4).
term outcomes of bullying on both victims and
Data from other nationally representative surveys

perpetrators.49 Children who are bullied are likely
confirm that bullying is common in some other
to experience a range of negative psychological
high-income countries as well. For instance,
outcomes, including depression, anxiety, thoughts
the Australian Covert Bullying Prevalence Study
of suicide and low life satisfaction.50 Across
(ACBPS), a 2007 national survey of students in
multiple ethnic groups, being bullied by peers has
grades 4 to 9 (around ages 9 to 15) from primary and
also been connected to a heightened risk of eating secondary high schools across the country, found
disorders and to social and relationship difficulties, high rates of frequent peer bullying. Students were
such as loneliness and being socially withdrawn.51 asked how often during the current school term they
Furthermore, students who are bullied are more bullied another student or group of students “again
likely to experience academic difficulties, including and again” and/or how often they were bullied by
underachievement, lower attendance and dropping students “again and again”. A total of 27 per cent of
out, among others.52 The social, emotional and Australian students reported experiencing frequent
psychological effects of bullying can be severe and bullying while only around 9 per cent admitted to
can persist throughout childhood into adulthood.53 frequently bullying others.58
120
Map 6.3
Bullying is a problem worldwide, experienced by a large proportion of teens in many countries

Percentage of adolescents aged 13 to 15 years who reported being bullied at least once in the past couple of months, by country
Above 50%
41% - 50%
31% - 40%
20% - 30%
Less than 20%
Countries that did not participate in
either the HBSC or GSHS
Sudan and South Sudan has not yet been determined. The final status of the Abyei area has not yet been determined. Data from the HBSC were recalculated as weighted averages for 13- to 15-year-olds to
allow for comparison with data collected in the GSHS. Reference periods for the two surveys differ slightly. Data for China, Colombia, Ecuador, State of Palestine, the Bolivarian Republic of Venezuela and
Zimbabwe are not national but have been recalculated on the basis of subnational surveys that took place in selected cities in each country or area. Data for Belgium are a weighted average of the Flemish
and French samples. Data for Ghana refer to students in junior high school only. Data for the United Kingdom are a weighted average of the samples in England, Scotland and Wales. Data for the United
Republic of Tanzania are not national but represent only the city of Dar es Salaam.
Source: HBSC, 2009/2010 and GSHS, 2003-2013.
Figure 6.4
Significant proportions of teens admit to having bullied others

Percentage of adolescents aged 11 to 15 years who reported bullying others at school at least once in the past couple of months
100
90
80
70

60
50
40
30
20
10
0
Romania
Latvia
Lithuania
Estonia
Ukraine
Switzerland
Austria
Greece
France
Belgium
Slovakia
Luxembourg
Germany
Canada
Russian Federation
Poland
Portugal
Netherlands
Finland
United States
Slovenia
Hungary
The former Yugoslav
Norway
Croatia
Denmark
Spain
United Kingdom
Italy
Ireland
Iceland
Armenia
Czech Republic
Sweden
Notes: Data for Belgium are a weighted average of the Flemish and French samples. Data for the United Kingdom are a weighted average of the samples in England, Scotland and Wales.
121
Box 6.2
A new age of bullying

Although the wide variety of to- A unique aspect of cyber-bullying is Canada65
day’s information and communica- that it allows perpetrators to remain
A secondary analysis of the Cana-
tion technologies and social media anonymous.62 Children have been dian data from the 2006/2007 wave
outlets afford many benefits and found to be the most common per- of the HBSC included 1,972 high
conveniences, they are not without petrators.63 school students. The analysis found
risk. The emergence of cyber- (or
that about 14 per cent of students
online) bullying is one such risk. Most of the available data on the
said they were bullied through
The Cyberbullying Research Center prevalence of cyber-bullying comes
the computer, e-mail messages/
in the United States defines cyber- from surveys conducted in the
pictures or mobile phones within
bullying as “Willful and repeated Western world since access to the
the two months preceding the sur-
harm inflicted through the use of types of technologies used for on-
vey. Rates of cyber-bullying were
computers, cell phones, and other line bullying are widely accessible
especially high for girls, with 18 per
electronic devices”.59 It can include there. A brief overview of data from
cent reporting they were victimized
threats, intimidation or harassment representative surveys in selected
compared to 8 per cent of boys.
via email, online chat rooms, instant countries and regions follows.
On the other hand, about 12 per
messaging, texting and social me-
cent of the students surveyed said
dia websites. It can also encompass Australia64 they bullied another student(s) in
acts such as defamation, exclusion
In the 2007 Australian Covert Bul- the last two months using either a
or peer rejection, impersonation,
lying Prevalence Study, around 7 computer, e-mail or mobile phone.
unauthorized publication of private
per cent of students in grades 4 to Boys and girls were equally likely to
information or images and manipu-
9 (roughly ages 9 to 15) reported admit to cyber-bullying others.
lation.60 As described in Chapter
4, cyber-bulling can have a sexual experiences of cyber-bullying, with
rates for girls and boys being rela- Europe66
connotation, involving for instance
the use of sexually loaded terms to tively similar (8 per cent and 5 per Between 2009 and 2011, the EU Kids
insult someone or the distribution cent, respectively). On the other Online Survey collected data on
of sexually explicit photos and vid- hand, only about 4 per cent of the children’s Internet use and experi-
eos for the purpose of shaming or students admitted having bullied ences online from a representative
causing emotional distress to the others online, with no differences sample of over 25,000 children aged
victim. While cyber-bullying can in the rates reported by boys and 9 to 16 in 25 European countries.67
occur without other forms of peer girls. Both experiences of being The survey included questions on
violence, many children who report victimized and perpetrating cyber- whether or not children had ever
being victimized online also recount bullying increased with grade levels. been treated, or treated others, in
experiences of bullying through Moreover, Internet-based bullying a hurtful or mean way on the Inter-
traditional means.61 Cyber-bullying through, for example, social net- net, whether as a single or repeated
can take place between friends and working sites was reportedly more episode. Across the 25 countries, 6
peers as well as within the context common than bullying via mobile per cent of these Internet users re-
of teen dating relationships, and the phones, especially as students got ported being bullied online, while 3
two can overlap. older. per cent admitted to having bullied
122
others online. However, children South Africa68 petrate cyber-bullying through the
were more likely to report that they distribution of picture or video clips
The 2012 National School Violence
had been bullied in person, with Study collected information on re- (35 per cent), followed by instant
almost one fifth stating that they ported experiences of a number of messaging systems such as Whats-
had experienced bullying offline. forms of cyber-bullying among a App or BlackBerry Messenger (27
There was also considerable over- nationally representative sample of per cent). When it came to students’
lap between those who are bullied secondary school students. In total, self-reported acts of online aggres-
online and offline, with around half around one in five (21 per cent) of sion, 6 per cent admitted to having
of online victims of bullying report- the students said they experienced ever sent a text about someone to
ing that they had also been bullied one of the following acts of cyber- make them angry or make fun of
in person. A similar proportion of bullying within the last year: an them while 4 per cent said they ever
online bullies stated that they had online fight with someone, where posted something hurtful about
also bullied others in person. Ad- rude or angry language was sent others online.
ditionally, those who reported bul- in a chat room; rude or insulting
lying others online were also found messages were sent about them United States of America69,70
to commonly report being bullied via computer or mobile phone;
Around one in six adolescents (15
by others online: Half of children messages were sent or posted that
per cent) in grades 9 through 12
who were cyber-bullies reported were hurtful, with the intention of
were bullied electronically through
damaging their reputation; some-
that they had also been a victim of emails, chat rooms, instant messag-
one shared secrets or embarrass-
cyber-bullying. ing, websites or texting in the past
ing pictures or information online
year according to the 2013 YRBS.
Among those who were bullied on- without their permission; someone
Girls were more than twice as likely
line, children varied considerably in used their account and pretended to
to report having been victims of
the level of harm they reported in be them by sending messages and
trying to damage their reputation; cyber-bullying than boys (21 per
response. Over half of children indi- cent and 9 per cent, respectively).
they were threatened with harm or
cated that they were either “very up- In the 2011 NatSCEV II, the life-
intimidated by someone online; or
set” or “fairly upset” by being bul- time victimization rate for Internet/
someone sent sexually explicit im-
lied online; 15 per cent reported not cell phone harassment was 9 per
ages or messages about them us-
being upset at all. Girls were more cent among children aged 5 to
ing a phone or computer. Of these
likely to report being upset than 17. The highest rate was reported
various forms, online fighting was
boys, with 37 per cent stating they among 14- to 17-year-olds, with
most commonly reported; sexual
were “very upset” compared to 23 cyber-bullying was the least com- one in five (20 per cent) stating
per cent of boys. Slightly more than mon. The most frequently cited that they experienced Internet/cell
three out of four children reported perpetrators of online fights, public phone harassment at some point in

having told someone about their sharing of secrets or information, their lives. Last-year victimization
experiences of being bullied (most account theft, and sending sexually among this age group was similar
often a friend or parent), nearly half explicit images or messages were to that recorded in the 2013 YRBS,
blocked the person from contacting friends of the victim. Students re- with around 14 per cent saying
them online, and about one third ported that mobile phones were the they were cyber-bullied during this
said they tried to fix the problem. most common medium used to per- period.
123
Box 6.3
A closer look at bullying in the Americas

United States of America a nationally representative sample been bullied in the past month
of over 100,000 students in their 9th at school in the following ways:
Several nationally representative
year of elementary school (around robbed, insulted or threatened, or
surveys of crime and risky behav-
ages 13 to 16).72 Students were physically struck or mistreated.
iours in the United States have in-
asked about the frequency with
cluded questions on bullying. The Just over half of the students (51 per
which they felt hurt, bothered, an-
rate recorded by the 2009/2010 HBSC cent) in the region reported experi-
noyed, offended or humiliated be-
of 26 per cent among 13- to 15-year- encing some type of bullying within
cause they were taunted, mocked
olds (see Map 6.3) is consistent with the past month. The rates of bully-
or bullied by their classmates within
these other sources. For example, ing varied across countries from a
the last 30 days preceding the sur-
the 2011 US School Crime Supple- high of 63 per cent in Colombia to
vey. Slightly more than one quarter
ment (SCS) to the National Crime a low of 13 per cent in Cuba. In 8 of
(28 per cent) of students reported
Victimization Survey found that 28 the 16 countries with available data,
that they “rarely or sometimes” felt
per cent of students aged 12 to 18 at least half of the children said they
hurt by things their classmates said
reported being bullied at school at were bullied in the last month.
or did; a further 7 per cent reported
some point during the school year.71
the frequency of such feelings to Across all countries, the most com-
A somewhat lower rate (20 per cent)
be “almost always or always”. On monly reported form of bullying
of having been bullied on school
the other hand, slightly more than was being robbed. Nearly 4 in 10
property within the last year was re-
one in five (21 per cent) of the ado- sixth-graders in the region said they
corded among students in grades 9
lescents reported that they bullied were the victim of robbery at school
to 12 in the 2013 national YRBS.
their classmates in the previous in the past month. Here again, the
Results are somewhat mixed when month, with boys being significantly highest reported rates were found in
it comes to differences between the more likely to report bullying oth- Colombia (55 per cent) and the low-
sexes. Data from the 2011 SCS show ers than girls (26 versus 16 per cent, est in Cuba (11 per cent). After being
that around 30 per cent of female respectively). robbed, reports of being insulted or
students between the ages of 12 and threatened were second most com-
Information on bullying was collect- mon among the students at around
18 reported being bullied at school;
ed in the Second Regional Compara- 27 per cent for the region overall.
the proportion of bullied male stu-
tive and Explanatory Study (SERCE) Children in Argentina reported the
dents of the same age was slightly
conducted by the United Nations highest rate (37 per cent) of this form
lower at 25 per cent. According to
Educational, Scientific and Cultural of bullying, while Cuban children re-
the 2013 YRBS, the prevalence of
Organization (UNESCO) between ported the lowest (7 per cent). Final-
being bullied on school property
2006 and 2007 in 16 countries in ly, experiences of physical bullying
among boys and girls was 16 per
Latin America and the Caribbean.73 were reported by 16 per cent of the
cent and 24 per cent, respectively.
Data were collected from a sample sixth-graders in the region. In Ar-
of over 91,000 sixth-grade students gentina, Costa Rica, the Dominican
(ages 10 to 14), representing ap- Republic, Ecuador and Nicaragua,
Brazil’s 2012 National Survey of proximately 10 million sixth-graders at least one in five children reported
School Health (Pesquisa Nacional in the region.74 Students were asked being a victim of physical bullying
de Saúde do Escolar) was based on about their experiences of having within the past month.
124
Percentage of sixth-grade students who were robbed, insulted or threatened, or physi-

cally bullied in the past month, and percentage of sixth-grade students who reported
having suffered some type of bullying at school in the past month
Insulted or Physically Any bullying

Robbed threatened bullied incident
Argentina 42 37 23 59
Brazil 35 25 13 48
Colombia 55 24 19 63
Chile 33 22 12 43
Cuba 11 7 4 13
Costa Rica 47 33 21 60
Dominican Republic 46 29 22 60
Ecuador 48 29 22 56
El Salvador 33 19 16 43
Guatemala 36 21 15 39
Mexico 40 25 17 44
Nicaragua 48 29 21 51
Panama 37 24 16 57
Paraguay 32 24 17 46
Peru 45 34 19 45
Uruguay 32 31 10 50
Total for Latin America

and the Caribbean 39 27 16 51
Note: Results for Latin America and the Caribbean as a whole were calculated by weighting the results
for each country.
Source: Román, M., and F. J. Murillo, ‘Latin America: School bullying and academic achievement’, Cepal
Review, vol. 104, 2011, pp. 37-53.
125
On the question of victimization by When differences between the sexes are found, boys are
bullying and differences between more likely than girls to be victims of bullying and to admit
the sexes, available data suggest bullying others
that both sexes are at equal risk.
In about half of the countries with Figure 6.5A
available data, girls and boys were
Percentage of adolescents aged 13 to 15 years who reported being bullied at least
equally likely to report being bullied once in the past couple of months, by sex
in recent months (Figure 6.5A). In
the few countries with a recorded 100
difference between the sexes, 90
boys were more likely than girls to
80
be victims of bullying. In countries
70
including Kuwait and Lebanon,
boys were twice as likely as girls to 60
report being bullied. Girls 50 Girls are more likely to Boys are more likely to
be bullied be bullied
40
On the other hand, boys were
30
significantly more likely than girls to
report having bullied others at least 20
once in the past couple of months 10

– a clear difference between
the sexes that emerged in the 0 10 20 30 40 50 60 70 80 90 100
Boys
findings from the HBSC in almost
all countries and among all age Notes: Each dot represents a country. Data from the HBSC were recalculated as weighted averages for 13- to 15-year-olds
to allow for comparison with data collected in the GSHS. Reference periods for the two surveys differ slightly.
groups. In Armenia, for example, Source: HBSC, 2009/2010 and GSHS, 2003-2013.
boys aged 13 to 15 were about
three times more likely to report
bullying others than girls, while in Figure 6.5B
Iceland, Ireland and Norway boys
Percentage of adolescents aged 13 to 15 years who reported bullying others at
were around twice as likely to admit
school at least once in the past couple of months, by sex
bullying others than girls (Figure
6.5B). Some caution is warranted 100
when interpreting these findings, 90
however, since it is not clear how
80
much of the difference might have
been caused by reporting biases 70
(that is, that girls were just less 60
likely to want to confess to bullying
Girls
50 Girls are more likely to Boys are more likely to

behaviour than boys). bully others bully others
40
In Australia, similar results were 30

obtained from the 2007 ACBPS: 20
Sex differences in children’s
10
experience of being bullied were
negligible, but rates of bullying 0 10 20 30 40 50 60 70 80 90 100
others were found to be slightly Boys
higher among male students (11 Note: Each dot represents a country. Data from the HBSC were recalculated as weighted averages for 13- to 15-year-olds.
per cent) in grades 4 to 9 than Source: HBSC, 2009/2010.
126
Victimization by bullying tends to decline with age in some among their female counterparts
countries… (7 per cent).75
Figure 6.6 Do children experience less

bullying as they get older?
Percentage of adolescents aged 11, 13 and 15 years who reported being bullied Because data from the HBSC
at least once in the past couple of months, in countries with a difference of more
than 10 percentage points in prevalence of victimization by bullying between were collected for three separate
11- and 15-year-olds ages (11, 13 and 15), it is possible
to explore whether the prevalence
70 Canada of victimization by bullying
Denmark declines with age. In some of
60 Estonia the countries with available data,
Finland
prevalence is found to decline
50 Hungary
between the ages of 11 and 15
Iceland
40 and in some cases the difference
Latvia
is quite significant. In 12 countries,
Poland
30 Russian Federation the survey found more than a
Switzerland 10-percentage-point difference
20 Ukraine between ages 11 and 15 (Figure
United States 6.6). Among these countries,
10 the differences are greatest in
Estonia, Hungary, Latvia and the
0
11-year-olds 13-year-olds 15-year-olds
Russian Federation. Results found
in the 2007 ACBPS in Australia
also confirmed this pattern. In that
survey, experiences of frequent
…while bullying others increases with age school bullying were highest (32
per cent) among fifth-graders
Figure 6.7
(aged 10 to 11) and lowest (24 per
Percentage of adolescents aged 11, 13 and 15 years who reported bullying others cent) among ninth-graders (aged
at school at least once in the past couple of months, in countries with a difference 14 to 15).
of more than 10 percentage points in prevalence of bullying others between
11- and 15-year-olds
On the other hand, reported
70 prevalence of bullying others

Austria generally increased from age
60 Germany 11 to age 15 in the majority of
Greece countries with available data
50 Lithuania from the HBSC. The increase
Luxembourg is particularly sharp in Greece,
40 Romania
where the prevalence of bullying
Switzerland
behaviour in the last couple of
30
months jumped from 27 per cent
20 among 11-year-olds to 51 per cent
among 15-year-olds (Figure 6.7).
10 In countries including Austria and
Germany, 15-year-olds were more
0 than one and a half times more
11-year-olds 13-year-olds 15-year-olds likely than 11-year-olds to report
Source: HBSC, 2009/2010. having recently bullied others.
127
Some of the GSHS included an additional question than physical intimidation, both in terms of past-
on the ways in which students were bullied most year (37 per cent and 14 per cent, respectively) and
frequently in the past 30 days. A few of the lifetime victimization (52 per cent and 25 per cent,
available country reports include information on respectively). Data from the 2011 US School Crime
the proportion of students aged 13 to 15 who were Supplement confirm the prominence of relational
most often bullied in the last month by being hit, aggression, since being the subject of rumours and
kicked, pushed, shoved around or locked indoors. being made fun of, called names or insulted were
Available data from 14 countries show that more the bullying behaviours with the highest reported
than one in five students in Djibouti, Tajikistan, prevalence among students aged 12 to 18.78
Thailand, Tonga, Uganda and Zambia most often
experienced physical acts of bullying (Figure 6.8). The NatSCEV II also found that 41 per cent of girls
In 6 of the 14 countries, significant differences reported being a victim of relational aggression
between the sexes are evident, with boys more within the last year compared to 32 per cent of
likely than girls to report physical bullying. In Libya, boys. The levels of physical intimidation were
boys were nearly three times more likely to report roughly the same for boys and girls for both
being physically bullied than girls. past-year and lifetime victimization. The rates
of physical intimidation were highest among
The NatSCEV II conducted in the United States children younger than 10 years, but relational
in 2011 includes findings on the proportions of aggression was highest among the 10- to 13-year-
children aged 2 to 17 who experienced physical old age group. By mid-adolescence (ages 14 to 17),
intimidation76 and those who experienced around one in three teens reportedly experienced
relational aggression,77 both within the past year physical intimidation at some point in their lives,
and at any point in their lives. Overall, the reported while nearly three quarters of them were victims
prevalence of relational aggression was higher of relational aggression.79
Figure 6.8
In Tajikistan, one in two students experience physical acts of bullying

Percentage of students aged 13 to 15 years who reported being bullied most often in the past 30 days by being hit, kicked, pushed,
shoved around or locked indoors, by sex
100
Boys Girls
90
80
70
60
50
40
30
20
10
0
Tajikistan
Thailand
Djibouti
Uganda
Zambia
Tonga
Bolivia (Plurinational
State of)
Benin
Maldives
Jordan
Oman
Peru
Libya
Indonesia
128
Box 6.4
Violence against children due to their

sexual orientation
An important aspect of the transition Research has documented that, com- for not having told school personnel
from childhood to adulthood is the pared to same-age heterosexual about their experiences were doubts
stronger emergence of one’s sexual peers, LGBT children are at increased that the staff would effectively address
identity. A key component of this pro- risk of interpersonal violence, includ- the situation and fear that reporting
cess entails the establishment of a ing forms of bullying and harassment, would somehow make the situation
sexual orientation, which has been de- particularly within the school environ- worse.
fined as “a person’s physical, romantic ment. In the 2011 National School Cli-
and/or emotional attraction towards mate Survey conducted in the United Research also confirms that LGBT
other people [...] Homosexual people States, 82 per cent of LGBT children children are at increased risk of self-
are attracted to individuals of the same and adolescents aged 13 to 20 said inflicted forms of violence, including
sex as themselves. Heterosexual peo- they had been verbally harassed, such suicide, often as a consequence of the
ple are attracted to individuals of a dif- as being called names or threatened, isolation and harassment experienced.
ferent sex from themselves. Bisexual at school in the past year because of One such study pooled results from
people may be attracted to individuals their sexual orientation; over one third representative samples of students in
of the same or different sex.”80 Anoth- reported that such abuse occurred fre- grades 9 to 12 collected in the Youth
er element, sometimes but not always quently.84 Almost two out of three stu- Risk Behavior Surveys carried out
linked to sexual orientation, is the way dents said they were verbally harassed from 2001 to 2009 in five major cities
in which a person identifies as, or feels within the last year because of their in the United States. The study found
a sense of being masculine and/or ‘gender expression’, with around one that lesbian, gay and bisexual children
feminine. This is referred to as ‘gen- in four reporting that such harassment were significantly more likely to think
der identity’. While a person’s gender occurred frequently. With regards to about, plan and attempt suicide than
identity is typically consistent with the physical harassment, such as being their heterosexual counterparts, even
sex they were born with (based on the shoved or pushed, 38 per cent of LGBT after controlling for other risk factors
presenting genitals), this is not always students reported such victimization at (such as experiences of interpersonal
the case. Transgender people have school within the last year as a result violence) and demographic variables.85
a sense of their own gender that is of their sexual orientation and 27 per The association between identifying
in contradiction to the sex they were cent due to their gender expression. as LGBT in adolescence and increased
given at birth.81 Relatively fewer students reported risk of attempted suicide has also been
incidents of physical assault, such as found in studies conducted outside the
Children who develop a non-tradition- being punched, kicked or injured with United States.86
al sexual orientation and/or gender a weapon; still, such experiences were
identity can find adolescence challeng- not altogether uncommon among this As with other children, exposure to
ing. Boys, in particular, may have a dif- group. About two out of three LGBT violence has been found to exert a
ficult time both resolving and express- students said they experienced some negative impact on the health and

ing a non-traditional sexual orientation form of sexual harassment (such as well-being of children who identify as
and/or gender identity because of soci- unwanted sexual touching or sexual LGBT. Research has shown that the
etal and cultural ideas of masculinity. remarks) at school in the past year, stresses faced by LGBT children and
While many children worldwide who and nearly one in five reported that adolescents can put them at risk of
are lesbian, gay, bisexual or transgen- such events occurred frequently. Fi- both physical and mental health prob-
der (LGBT)82 thrive in ways no differ- nally, the vast majority (90 per cent) of lems and substance use.87 Changing
ent than their peers, others find them- LGBT students surveyed said they felt attitudes and promoting positive so-
selves the victims of targeted acts of deliberately excluded or ‘left out’ by cial norms that welcome and embrace
violence as a result of their real, or other students; about half experienced diversity in cultures around the world
perceived, sexual orientation. Nega- this feeling frequently. When it came are critical steps in acknowledging and
tive societal attitudes towards those to reporting incidents of assault or har- protecting the rights of all children.
who are LGBT are often a contributing assment in school, 60 per cent of LGBT Schools and parents also play a cru-
factor leading to such victimization. students said they never reported their cial role in creating environments that
Children can also suffer from abuse experiences to school staff and 56 per are accepting and supportive of chil-
and discrimination when their parents cent never told a family member. The dren to fully express themselves with-
are LGBT.83 most commonly cited reasons given out fear.
129
Chapter 7 Violent
unions
among
adolescents
Intimate partner violence includes any physical,
sexual or emotional abuse perpetrated by a current
or former partner within the context of marriage,
cohabitation or any other formal or informal union.
Although both girls and boys can be victims of
intimate partner violence, girls are at greater risk.
In fact, violence directed at girls and women by
an intimate partner is the most common form of
gender-based violence.1 In societies that sanction
male dominance over women, violence between
intimate partners may be perceived as an ordinary
component of interpersonal dynamics between
the sexes, particularly in the context of marriage or
other formal unions.
In many cultures, girls reaching puberty are

expected to assume gender roles associated with
womanhood. These include entering into a union
and becoming a wife and mother. Parents may
pursue marriage for their adolescent daughters
in an attempt to secure a better economic future
for them. In turn, adolescent girls may consent to
the arrangement due to family pressures without
fully understanding – and being prepared for – the
responsibilities, risks and considerable complexity
of navigating the roles of wife, mother and daughter-
in-law.2
Marriage before the age of 18 is a fundamental

violation of human rights. Although child marriage

© UNICEF/NYHQ2012-0096/Asselin
is often against the law (along with other harmful

practices, such as female genital mutilation/cutting),
it persists because it is a deeply embedded social
norm associated with perceptions of femininity and
masculinity and is considered critical to upholding
societal values.3 Child marriage typically results in
the abandonment of school and in early pregnancy.4
130
Globally, nearly one in five adolescent girls aged 15 age of 18 or older.11 Data from a population-based
to 19 are currently married or in union. South Asia survey in seven Ethiopian regions also revealed that
has the highest proportion of married adolescent early marriage was associated with a heightened
girls (29 per cent), followed by West and Central risk of partner violence – including both physical
Africa (25 per cent), Eastern and Southern Africa (21 violence and forced first marital sex.12
per cent) and Latin America and the Caribbean (19
per cent). The rates of early marriage are particularly Adolescents who are involved in informal relationships
high in the Central African Republic and Niger, where can be equally vulnerable to intimate partner violence.
more than half of girls between the ages of 15 and Dating violence refers to a pattern of controlling
19 are currently married or in union.5 Girls who marry or violent behaviours by a current or former dating
or enter into union at an early age often end up with partner. Like intimate partner violence in formal
considerably older men. Available data from more unions, this can include various forms of physical,
than 60 low- and middle-income countries indicate emotional and sexual violence. Dating violence can
that an estimated 20 per cent of married or cohabiting occur in person or online and includes such acts as
girls aged 15 to 19 are with a man who is at least texting or posting sexually explicit photographs on
10 years older. The age gap between spouses is the Internet (see Box 4.5 on page 70).13 Adolescents,
particularly high in Mauritania and Nigeria, where 60 who are often new to romantic liaisons, may have
per cent and 52 per cent, respectively, of married or difficulty coping with feelings associated with
cohabiting adolescent girls are with a man who is at intimate relationships and may lash out at their
least 10 years older than they are. partners when they lack constructive ways to deal
with their frustrations.14 Many adolescents do not tell
A girl who marries early may find herself in a anyone about the abuse they experience because
vulnerable position vis-à-vis her husband and they feel embarrassed or afraid or do not understand
his family. She may also be cut off from her own that the way they are being treated is inappropriate.
family, friends and other sources of social support6 Dating violence can, however, escalate into very
and be more economically dependent than same- serious forms of physical, emotional or sexual abuse.
age peers who are not married.7 Furthermore, in It can also set the stage for lifelong involvement in
societies where girls and women are believed to unhealthy intimate relationships.15
hold a lower status than boys and men, they may
be socialized into thinking that certain forms of Dating violence can affect adolescents of any social
violence against them are justifiable, carrying this and demographic background. Reports of physical
set of expectations into their marriage at a young victimization in dating relationships are common
age. Research confirms that girls who marry in among adolescents of both sexes. However, girls
childhood are at greater risk for intimate partner
CHAPTER 7: violent unions among adolescents

are more likely to engage in physical force in self-
violence than same-age peers who marry later.8 defence and are more likely to be seriously injured
For example, a research study in India and Nepal
in dating violence than boys.16 They are also far
found that women who married before age 18
more likely to be victims of sexual violence in
experienced increased risk of both current and
dating relationships.17 Adolescents who witness or
lifetime physical and sexual violence by a partner
experience violence in the home, who have friends
compared to those who married after age 18.9
involved in violent intimate relationships or who
Similarly, a study in Bangladesh found that women
have been socialized into believing that violence
who married before age 18 were more likely than
is acceptable behaviour are at heightened risk for
those who married later to be subjected to physical
perpetrating dating violence.18
violence by their partners.10 A study involving a
nationally representative sample of youths aged 14
Lifelong – sometimes
to 25 in Viet Nam found that marriage before age
18 placed young women at nearly twice the lifetime
intergenerational – consequences
risk of exposure to intimate partner violence in Partner violence can have devastating consequences
comparison to young women who married at the for adolescents’ health, well-being and overall
131
development. It can lead to homicide, suicide, are also found in other low- and middle-income
physical injuries, disability and reduced physical countries with available data (see Box 7.2).
functioning.19 It can also damage the mental health
of adolescents, resulting in depression, anxiety Rates of partner violence are particularly high
and post-traumatic stress disorder, among other in many areas of Africa. In Eastern and Southern
conditions.20 Furthermore, research has shown that Africa, more than one third of adolescent girls
the experience of intimate partner violence is often who have ever been married or in union reported
associated with high-risk and antisocial behaviours, experiences of partner violence in each of the nine
including substance use and aggression.21 Partner countries for which data are available, with the
violence can also be detrimental to the reproductive exception of Comoros. In this region, the prevalence
and sexual health of adolescents, with consequences of partner violence approaches or exceeds 50 per
including unwanted pregnancies, abortions and cent in Uganda, the United Republic of Tanzania
sexually transmitted infections, including HIV.22 and Zimbabwe. In West and Central Africa, the
prevalence of partner violence is more varied –
Since many girls who enter unions early become ranging from 8 per cent in Burkina Faso to 70 per
mothers at a young age, partner violence holds cent or more in the Democratic Republic of the
significant intergenerational consequences. Expo- Congo and Equatorial Guinea.
sure to domestic violence among children is highly
detrimental to their emotional, psychological, Partner violence is also pervasive in South Asia,
cognitive and social development.23 And children where at least one in five girls who have ever been
who witness domestic violence are themselves married or in union experienced partner violence
at heightened risk of experiencing abuse within in each of the four countries with available data.
the home.24 Moreover, these children are likely to In this region, the prevalence of partner violence
develop aggressive behaviours and become the per- is particularly high in Bangladesh (47 per cent)
petrators of violence against their siblings or peers.25 and India (34 per cent). Latin America and the
Finally, children who grow up witnessing domestic Caribbean shows similarly high levels of partner
violence are prone to carry violence into adulthood, violence among adolescents, with more than one
as either victims or perpetrators.26 For instance, girls in four girls who have ever been in a formal union
and boys who have experienced violence within the
reporting such violence in all six countries with data.
family are more likely to encounter partner violence
The prevalence is especially high in the Plurinational
in their own unions.27 Furthermore, exposure to do-
State of Bolivia and Haiti, exceeding 40 per cent.
mestic violence between their parents may influ-
ence children’s own attitudes about the acceptabil-
In East Asia and the Pacific, more than one in
ity of violence, which may be passed down to their
six ever-married or partnered girls experienced
own children in the future.

violence at the hands of their partners in all of
the four countries with available data. Overall, the
Prevalence of partner prevalence of partner violence is lowest in Central
violence against ever- and Eastern Europe and the Commonwealth of
married adolescent girls Independent States (CEE/CIS), ranging from 2
Comparable data from 43 low- and middle-income per cent in Ukraine to 14 per cent in Azerbaijan
countries (see Box 7.1) show that prevalence rates and 16 per cent in Tajikistan. In this region, the
for partner violence against ever-married adolescent prevalence of partner violence exceeds 20 per cent
girls range from 2 per cent in Ukraine to 73 per cent in two countries – Kazakhstan and the Republic of
in Equatorial Guinea (Figure 7.1).28 More than one Moldova. DHS data on this topic from the Middle
in three ever-married adolescent girls between the East and North Africa are extremely limited. In both
ages of 15 and 19 have experienced some form countries represented in this region (Egypt and
of physical, sexual or emotional violence at the Jordan), more than one in five girls who have ever
hands of their partners in half of these countries. been married or in a formal union reported one or
Widespread levels of intimate partner violence more incidents of partner violence.
132
Figure 7.1
More than half of ever-married girls have experienced partner violence in Cameroon, the
Democratic Republic of the Congo, Equatorial Guinea, Gabon and Zimbabwe
Percentage of ever-married girls aged 15 to 19 years who ever experienced any physical, sexual or emotional violence committed
by their husbands or partners
Equatorial Guinea
Gabon
Zimbabwe
Cameroon
Bangladesh
Uganda
Liberia
Rwanda
Haiti
Zambia
Marshall Islands
Malawi
Mozambique
Ghana
India
Colombia
Kenya
Dominican Republic
Philippines
Timor-Leste
Honduras
Pakistan
Republic of Moldova
Peru
Jordan
Côte d’Ivoire
Kazakhstan

Nepal
Egypt
Cambodia
Cabo Verde
Tajikistan
Nigeria
Mali
Azerbaijan
Comoros
Burkina Faso
Kyrgyzstan
Ukraine
0 10 20 30 40 50 60 70 80 90 100
Notes: Data for the Plurinational State of Bolivia and Honduras refer to girls who experienced any physical, sexual or emotional partner violence in the last 12 months. Data for Côte d’Ivoire refer only to
adolescent girls who are currently married or in union. Data for Kazakhstan are from MICS 2010-2011, which used an adapted version of the DHS module on domestic violence. Data for Pakistan refer only
to physical or emotional violence. Data for the Philippines refer to physical, sexual, emotional or other forms of violence, including economic violence. Data for Bangladesh, Colombia, Peru and Rwanda
refer to physical or sexual violence only. Data for Cabo Verde, Equatorial Guinea, Marshall Islands and Ukraine are based on 25-49 unweighted cases.
133
Box 7.1
Data sources on intimate partner violence

Definitions of partner or spousal vio- experiences of specific acts rather him or someone close to her/him
lence vary considerably across data than leaving it up to participants to • Insulted her/him or made her/him
sources. For instance, some surveys define what constitutes violence. The feel bad about herself/himself.
incorporate questions on physi- modified version of the CTS used in
cal, sexual and emotional violence the DHS standard module includes Partner sexual violence: Sexual
whereas others include only one or questions on specific acts of physi- violence is measured by asking re-
two of these forms. Some surveys cal, emotional and sexual violence spondents whether their spouse/
capture additional forms of partner perpetrated by a respondent’s current partner ever:
violence, such as economic violence. or former husband or partner. The
• Physically forced her/him to have
Diverse age groups of adolescents DHS module asks all girls and women
sexual intercourse with him/her
are also included within different aged 15 to 49 whether (1) they are cur-
even when she/he did not want to
surveys, further complicating com- rently married or living with a man as
• Physically forced her/him to per-
parisons across sources. if married, or (2) they have ever been
form any other sexual acts she/he
married or ever lived with a man as
While intimate partner violence has did not want to
if married. The questions on intimate
been widely researched in many • Forced her/him with threats or in
partner violence are then asked only
high-income countries – for example, any other way to perform sexual
to respondents who answer ‘yes’ to
Australia, Canada, the United King- acts when she/he did not want to.
either scenario. Some DHS also use
dom and the United States – the data the domestic violence module with
In the DHS, adolescents are classified
have largely been collected from the ever-married29 boys and men aged 15
as having experienced partner vio-
adult population (that is, women and to 49 (or an expanded age range, typi-
lence if they were subjected to any of
men over the age of 18). This is most- cally to ages 54, 59 or 64).
the above-mentioned acts.
ly due to the fact that relatively few
adolescents in such countries can be The three forms of violence meas-
The DHS domestic violence module
found in marriages or other formal ured are as follows:
also collects information on the de-
unions before that age. The focus gree of marital control exercised by
Partner physical violence: To meas-
of this chapter is therefore on low- husbands or wives by asking wheth-
ure this form of violence, respond-
and middle-income countries. How- er he/she exhibits any of the follow-
ents are asked whether their spouse/
ever, data on dating violence among ing behaviours: is jealous or angry if
partner ever:
adolescents in a number of selected the respondent talks to other men/
high-income countries are presented • Pushed her/him, shook her/him or women; frequently accuses her/him
in the final section of the chapter. threw something at her/him of being unfaithful; does not permit
• Twisted her/his arm, pulled her/his meetings with female/male friends;
This chapter relies primarily on na- tries to limit her/his contact with her/
hair or slapped her/him
tionally representative data from his family; and insists on knowing
• Punched her/him with his/her fist
Demographic and Health Surveys where she/he is at all times (older
or with something that could hurt
(DHS). The use of DHS data confers DHS also asked whether he/she does
her/him
numerous advantages since stand- not trust her/him with any money).
• Kicked her/him, dragged her/him
ardized definitions and measures
or beat her/him up
allow for comparisons across coun- Respondents are also asked when
• Tried to choke her/him or burn her/
tries. Questions on intimate partner partner violence first began to oc-
him
violence included in the DHS module cur in the relationship, any physical
• Threatened or attacked her/him
on domestic violence are based on a injuries that resulted from physical
with a knife, gun or other type of
modified version of the Conflict Tac- or sexual violence, and whether their
weapon.
tics Scale (CTS). Originally developed partner consumes alcohol and, if
by sociologist Murray Straus in the Partner emotional violence: Partner so, the frequency with which he/she
1970s, the CTS is the most widely emotional violence is assessed by gets drunk (“often”, “sometimes”
used research tool for measuring asking respondents whether their or “never”). Finally, they are asked
intimate partner violence. There are spouse/partner ever: whether they have ever hit, slapped,
numerous benefits associated with kicked or done anything to physi-
its use, particularly when conduct- • Said or did something to humiliate cally hurt their partner when he/she
ing cross-cultural research, because her/him in front of others was not already beating or physically
it questions participants about their • Threatened to hurt or harm her/ hurting her/him.
134
© UNICEF/ZAMA2011-0260/Nesbitt
Partner physical violence available data, at least one in five adolescent girls
Available data indicate that physical violence and reported incidents of physical violence by a partner.
emotional violence are the most commonly reported Nepal has the lowest prevalence (16 per cent) and
forms of partner violence perpetrated against Bangladesh the highest (40 per cent). Rates of
adolescent girls. In 33 out of 43 countries, at least partner physical violence are lower in East Asia and
1 in 10 adolescent girls (aged 15 to 19) who have the Pacific, with the Marshall Islands exhibiting the
ever been married or in union reported incidents highest rate (30 per cent) and Cambodia the lowest
of physical violence against them at the hands of (6 per cent).
their partners. The rates of physical violence among
these girls vary considerably across countries – In Latin America and the Caribbean, at least 1 in
ranging from 2 per cent in Ukraine to 71 per cent in 10 girls reported partner physical violence in all six
Equatorial Guinea. countries in the region with comparable data. In the
Plurinational State of Bolivia, Colombia, Haiti and
As shown in Figures 7.2A-7.2F, the prevalence of Peru, more than one quarter of girls who were ever
partner physical violence also varies across regions. married or partnered reported incidents of physical

It is most widespread in West and Central Africa, abuse by a husband or live-in partner. Among
Eastern and Southern Africa, and South Asia. In all regions, the lowest reported rates of partner
West and Central Africa, at least one in five girls physical violence among adolescent girls are found
reported partner physical violence in around half of in CEE/CIS. In that region, rates are below 10 per
the countries with available data. Rates exceed 30 cent in Kyrgyzstan, Tajikistan and Ukraine. The
per cent in Cameroon and Liberia, 40 per cent in highest prevalence level is found in the Republic
the Democratic Republic of the Congo and 50 per of Moldova, where roughly one in five adolescent
cent in Equatorial Guinea and Gabon. In Eastern girls reported ever experiencing this kind of abuse.
and Southern Africa, at least one in five adolescent In the two countries with comparable data from the
girls reported partner physical violence in each Middle East and North Africa, close to one in five
of the nine countries with available data (except ever-married adolescent girls in Egypt and around
Comoros), with rates reaching as high as 41 per 1 in 10 in Jordan said they experienced some
cent in Rwanda. form of physical violence at the hands of their
husbands or partners at some point in their lives
In three out of four South Asian countries with (results not shown).
135
The prevalence of partner physical violence among adolescent girls is particularly widespread in
sub-Saharan Africa and South Asia
Percentage of ever-married girls aged 15 to 19 years who ever experienced any physical violence committed by their husbands
or partners
Figure 7.2A West and Central Africa Figure 7.2B Eastern and Southern Africa
Equatorial Guinea Rwanda

Gabon Zambia
Cameroon United Republic of Tanzania
Liberia Zimbabwe
Sao Tome and Principe Uganda
Côte d’Ivoire
Ghana Kenya
Mali Mozambique
Nigeria Malawi
Cabo Verde
Burkina Faso Comoros
0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100
Source: UNICEF global databases, 2014, based on DHS, 2006-2012. Source: UNICEF global databases, 2014, based on DHS, 2007-2011.
Figure 7.2C South Asia Figure 7.2D East Asia and the Pacific
Bangladesh Marshall Islands

India Timor-Leste
Pakistan Philippines
Nepal Cambodia
0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100
Central and Eastern Europe and the

Figure 7.2E Latin America and the Caribbean Figure 7.2F Commonwealth of Independent States
Colombia Republic of Moldova

Haiti Kazakhstan
Bolivia (Plurinational State of) Azerbaijan
Peru Tajikistan
Dominican Republic Kyrgyzstan
Honduras Ukraine
0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100
Source: UNICEF global databases, 2014, based on DHS, 2007-2012. Source: UNICEF global databases, 2014, based on DHS and MICS, 2005-2012.
Notes for all figures on this page: In the Plurinational State of Bolivia, physical violence includes being pushed, beaten with hands or hard objects, and choked or burned and refers to experiences
in the last 12 months. In Colombia, physical violence also includes being bitten. Data for Côte d’Ivoire refer to adolescent girls who are currently married or in union. Data for Honduras refer to girls
who experienced any partner physical violence in the last 12 months. Data for Kazakhstan are from MICS 2010-2011, which used an adapted version of the DHS module on domestic violence. Data for
Cabo Verde, Equatorial Guinea, the Marshall Islands and Ukraine are based on 25-49 unweighted cases.
136
Partner emotional violence Partner sexual violence
In 33 out of 42 countries with comparable data, In general, sexual violence by a spouse or partner is
overall rates of emotional violence inflicted by less commonly reported than physical or emotional
a partner are similar to those of partner physical violence in all 42 countries with comparable data.31
violence, affecting more than 1 in 10 adolescent One notable exception to this pattern is Zimbabwe,
girls who have ever been married or in union.30 where the prevalence of partner sexual violence
Rates of emotional violence vary substantially exceeds that of partner physical violence. As was
across regions and countries, ranging from 0 per the case for physical abuse, the three regions
cent in Ukraine to 57 per cent in Equatorial Guinea that reported the highest rates of sexual violence
(Figures 7.3A-7.3F). among ever-partnered adolescent girls are Eastern
and Southern Africa, West and Central Africa, and
While rates of partner physical violence are South Asia (Figures 7.4A-7.4F).
generally lower than average in most Latin
American and Caribbean countries, that region In Eastern and Southern Africa, at least 1 in 15 girls
has the highest levels of emotional violence: At reported experiences of partner sexual violence in
least one quarter of ever-partnered girls in all six all nine countries (except Comoros) for which data
countries with available data recounted incidents of are available. Zimbabwe has the highest rate in this
emotional abuse by their husbands or partners. The region (over one third). In West and Central Africa,
Plurinational State of Bolivia has the highest rate at prevalence rates are more than 10 per cent in 5 of
33 per cent. the 12 countries with available data. The prevalence
of partner sexual violence is particularly high in the
Rates of partner emotional violence are particularly Democratic Republic of the Congo (36 per cent) and
high in West and Central Africa, as was the case for Cameroon (24 per cent). In South Asia, more than
physical violence. In this region, over one quarter 1 in 10 adolescent girls surveyed in Bangladesh,
of adolescent girls who have ever been married or India and Nepal reported partner sexual violence.
in union reported experiences of partner emotional In this region, this type of abuse is most common
violence in half of the 12 countries with comparable in Bangladesh, where it was reported by about one
data. Prevalence is also very high in Eastern and in five ever-partnered girls between the ages of 15
Southern Africa, as with physical violence. In this and 19.
region, at least one in nine adolescent girls reported

emotional violence by a partner in each of the nine East Asia and the Pacific shows the highest level
countries except Comoros. Rates approach 30 per of variability among countries in a region, ranging
cent in Mozambique, Uganda and Zimbabwe. from about 2 per cent in Cambodia to 13 per cent
in the Marshall Islands. CEE/CIS and Latin America
While the prevalence of partner emotional and the Caribbean reported the lowest rates
violence is lower on average in CEE/CIS, this of partner sexual violence overall. In these two
region demonstrates the highest level of variability regions, prevalence is under 10 per cent in all of the
among countries, ranging from about 0 per cent in countries with available data except Haiti, where
Ukraine to 19 per cent in the Republic of Moldova. roughly one in four girls reported experiences of
Comparable DHS data from South Asia, East Asia sexual violence in their relationships with their
and the Pacific, and the Middle East and North partners during their lifetimes. In the Middle East
Africa are very limited. Within these three regions, and North Africa, rates of partner sexual violence
Jordan, Pakistan and the Philippines exhibit the are higher in Jordan (13 per cent) than in Egypt (3
highest rates of partner emotional violence. per cent) (results not shown).
137
Partner emotional violence among adolescent girls varies across regions but appears to be most
prevalent in Latin America and the Caribbean
Percentage of ever-married girls aged 15 to 19 years who ever experienced any emotional violence committed by their husbands
or partners
Figure 7.3A Latin America and the Caribbean Figure 7.3B West and Central Africa
Equatorial Guinea
Cameroon
Peru
Liberia
Gabon
Dominican Republic
Ghana
Haiti
Côte d’Ivoire
Colombia Cabo Verde
Nigeria
Honduras Mali
Burkina Faso
0 10 20 30 40 50 60 0 10 20 30 40 50 60

Figure 7.3C Eastern and Southern Africa Figure 7.3D Commonwealth of Independent States
Zimbabwe Republic of Moldova

Mozambique
Uganda Tajikistan
United Republic of Tanzania Kazakhstan
Malawi
Kenya Azerbaijan
Zambia Kyrgyzstan
Rwanda
Comoros Ukraine 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60
Figure 7.3E East Asia and the Pacific Figure 7.3F South Asia
Philippines Pakistan
Marshall Islands
India
Cambodia
Timor-Leste Nepal
0 10 20 30 40 50 60 0 10 20 30 40 50 60
Notes for all figures on this page: In the Plurinational State of Bolivia, emotional violence includes the partner humiliating or insulting her, accusing her of being unfaithful, attempting to limit con-
tact with her family, threatening to leave her, and threatening to take away the child(ren) or financial support and refers to experiences in the last 12 months. In Colombia, emotional violence includes
the partner threatening to abandon her, take away the child(ren) and take away financial support. Data for Côte d’Ivoire refer to adolescent girls who are currently married or in union. Data for Hondu-
ras refer to girls who experienced any partner emotional violence in the last 12 months. Data for Kazakhstan are from MICS 2010-2011, which used an adapted version of the DHS module on domestic
violence. In Peru, emotional violence includes humiliating her in front of others, threatening to harm her or those close to her, and threatening to leave the house and take away financial support or the
child(ren). In the Philippines, emotional violence also includes not allowing her to engage in legitimate work or practise a profession; controlling her money or property or forcing her to work; destroying
her personal property, pets or belongings or threatening to harm her pets; and having other intimate relationships. Data for Rwanda are based on 25-49 unweighted cases and are from the 2005 DHS
(questions on emotional violence were not asked in the 2010 DHS). Data for Cabo Verde, Equatorial Guinea, the Marshall Islands and Ukraine are based on 25-49 unweighted cases.
138
In Zimbabwe and the Democratic Republic of the Congo, more than one in three adolescent girls
reported incidents of sexual violence at the hands of an intimate partner
Percentage of ever-married girls aged 15 to 19 years who ever experienced any sexual violence committed by their husbands
or partners
Figure 7.4A Eastern and Southern Africa Figure 7.4B West and Central Africa
Zimbabwe Democratic Republic of the Congo

Uganda Cameroon
Gabon
Malawi Liberia
United Republic of Tanzania Sao Tome and Principe
Ghana
Zambia
Equatorial Guinea
Kenya Côte d’Ivoire
Rwanda Nigeria
Mali
Mozambique
Burkina Faso
Comoros Cabo Verde 0
0 10 20 30 40 50 0 10 20 30 40 50
Figure 7.4C East Asia and the Pacific Figure 7.4D South Asia
Marshall Islands
Bangladesh
Philippines
India
Timor-Leste
Cambodia Nepal
0 10 20 30 40 50 0 10 20 30 40 50

Figure 7.4E Latin America and the Caribbean Figure 7.4F Commonwealth of Independent States

Haiti Kazakhstan
Peru Azerbaijan
Bolivia (Plurinational State of) Tajikistan
Colombia Republic of Moldova
Honduras Kyrgyzstan 0
Dominican Republic Ukraine 0

0 10 20 30 40 50 0 10 20 30 40 50
Notes for all figures on this page: Data for the Plurinational State of Bolivia refer to girls who experienced forced sexual intercourse in the last 12 months. Data for Côte d’Ivoire refer to adolescent
girls who are currently married or in union. Data for Honduras refer to girls who experienced any partner sexual violence in the last 12 months. Data for Kazakhstan are from MICS 2010-2011, which
used an adapted version of the DHS module on domestic violence. In the Philippines, sexual violence includes the partner trying or attempting to force sexual intercourse or any other sexual acts
against her will. In Comoros, Gabon, Haiti, Kyrgyzstan, Nigeria, Tajikistan and Uganda, sexual violence includes being forced with threats or in some other way to perform unwanted sexual acts. Data
for Cabo Verde, Equatorial Guinea, the Marshall Islands and Ukraine are based on 25-49 unweighted cases.
139
Box 7.2
Additional data on intimate partner

violence from five regions
Additional statistics on the preva- physical, sexual or emotional vio- physical and sexual violence by a
lence of intimate partner violence lence at the hands of a husband or partner, respectively. Reported rates
in low- and middle-income coun- partner. In both countries, the most of emotional, physical and sexual
tries or areas are provided here be- commonly reported form of partner violence ever committed by a hus-
cause the sources of data are not violence was emotional abuse, fol- band or partner in Paraguay among
fully comparable with those cited lowed by physical and then sexual girls aged 15 to 19 were 30 per cent,
in the main text. The information violence. In Nicaragua, 17 per cent 12 per cent and 4 per cent, respec-
presented is not exhaustive but is of adolescent girls reported expe- tively.
intended to illustrate what is known riencing either physical or sexual
about intimate partner violence in a violence committed by a partner, Central and Eastern Europe and
selection of other countries with na- with physical abuse being more the Commonwealth of Independent
tionally representative data. Many commonly reported. An additional States
of the surveys relied on a modified 29 per cent of girls aged 15 to 19
In CEE/CIS, information on violence
version of the Conflict Tactics Scale. said they experienced emotional
committed by a partner of ever-mar-
abuse by a partner at some point in
ried girls (aged 15 to 19) is available
Latin America and the Caribbean their lives. In Jamaica, 30 per cent of
from Reproductive Health Surveys
girls and women aged 15 to 24 said in Albania (2004),40 Georgia (2010)41
In Latin America and the Caribbean,
additional data on intimate part- they were a victim of emotional, and Romania (2004).42,43 The rates of
ner violence among ever-married physical or sexual partner violence partner physical violence and emo-
adolescent girls aged 15 to 19 are at some point during their lives. In tional violence among adolescent
available for Ecuador (2004),32 El Mexico, 30 per cent of girls aged 15 girls were 11 per cent and 25 per
Salvador (2008),33 Guatemala (2008- to 17 reported experiences of physi- cent in Albania, 5 per cent and 8 per
2009),34 Nicaragua (2011-2012)35 and cal, emotional, economic or sexual cent in Georgia, and 16 per cent and
Paraguay (2008);36 for ever-married violence in the last 12 months com- 32 per cent in Romania, respective-
girls and women aged 15 to 24 in mitted by a partner or ex-partner, ly. Less than 5 per cent of adolescent
Jamaica (2008);37 and for all girls including 28 per cent who reported girls in all three countries reported
aged 15 to 17 in Mexico (2011).38,39 emotional violence, 4 per cent eco- physically forced sexual intercourse
nomic violence, 3 per cent physi- by a husband or partner.
In El Salvador and Guatemala, 33 cal violence and 2 per cent sexual
per cent and 38 per cent of ado- violence. In Ecuador, 27 per cent, A nationally representative survey on
lescent girls aged 15 to 19, respec- 22 per cent and 4 per cent of girls domestic violence conducted in Tur-
tively, reported ever experiencing aged 15 to 19 reported emotional, key in 2008 measured experiences
140
of physical, sexual and emotional around half said they were a victim violence committed by a husband or
violence committed by partners of of physical, sexual or emotional partner at some point in their lives.
ever-married girls and women aged violence by a partner. In Vanuatu, Further, 5 per cent of girls said they
15 to 59.44 Among girls and women 62 per cent of adolescent girls re- were physically forced by a hus-
aged 15 to 24, 35 per cent reported ported ever having experienced band or partner either to have sex-
experiences of physical or sexual vi- physical and/or sexual violence by a ual intercourse or to perform other
olence committed by a husband or partner, with 44 per cent reporting sexual acts when they did not want
partner at some point in their lives, physical violence and 52 per cent to. Psychological violence included
including 32 per cent who said they reporting sexual violence.50 Finally, humiliating her in front of others or
were physically victimized and 14 in Viet Nam, around one in four threatening her or someone close
per cent who said they were sexu- married girls and women aged 14 to her; 6 per cent of adolescent girls
ally violated. An additional 37 per to 25 reported some form of physi- reported experiencing one of these
cent of girls and women in this age cal, sexual or emotional violence at forms of partner psychological vio-
group said they experienced some the hands of their spouse: 6 per cent lence. In Swaziland, 24 per cent of
form of partner emotional violence said they were hit, 3 per cent were currently married or partnered girls
during their lifetimes. forced to have sexual intercourse aged 15 to 19 said their husbands
and 23 per cent were yelled at or or partners ever hit or beat them for
Asia and the Pacific cursed. having annoyed or angered them;
10 per cent said this happened in
In Asia and the Pacific, nationally
Sub-Saharan Africa the last 12 months.
representative surveys that includ-
ed questions on girls’ and women’s Two recent Multiple Indicator Clus-
experiences of domestic violence ter Surveys (MICS) from Chad
by an intimate partner were con- (2010) and Swaziland (2010) includ- A 2011 survey conducted in the State
ducted in the Maldives (2006),45 ed selected questions on domestic of Palestine included questions on
Samoa (2000),46 Vanuatu (2009)47 violence among girls and women. several forms of domestic violence
In Chad, physical violence includ- (physical, sexual, psychological, so-
CHAPTER 7: Intimate partner violence among adolescents

and Viet Nam (2009-2010).48,49 In
the Maldives, Samoa and Vanuatu, ed the following acts: bumps into cial and economic) experienced by
data are available on experiences you, shakes you or throws some- ever-married girls and women aged
of physical, sexual and emotional thing against you; slaps or twists 15 to 24 at some point in their lives
violence committed by partners of your arm; kicks you or pushes you and also within the last 12 months.51
ever-married girls aged 15 to 19. to ground; tries to strangle or burn In total, about one third said they
you; threatens you with a knife or ever experienced one of the five
Among ever-married adolescent gun; and attacks you with a knife, forms of spousal violence, including
girls in the Maldives, about 1 in pistol or other weapon. Around 12 physical (30 per cent), sexual (16 per
10 reported some form of partner per cent of ever-married girls aged cent), psychological (58 per cent), © UNICEF/NYHQ2012-1462/Dormino
physical, sexual or emotional vio- 15 to 19 said they experienced at social (30 per cent) and economic
lence in their lifetimes. In Samoa, least one of these forms of physical (31 per cent) violence.
141
Partner violence Figure 7.5
initiated by
adolescent girls In a few countries, a significant share of adolescent girls
admit to having initiated physical violence against their
Girls and women are also
partners
perpetrators of violence against
their husbands and partners. Percentage of ever-married girls aged 15 to 19 years who reported ever commit-
ting physical violence against their husbands or partners when the men were not
To explore this issue, the DHS already beating or physically hurting them
included a question on whether
ever-married females had ever
hit, slapped, kicked or done
Equatorial Guinea
anything else to physically hurt
their husbands or partners when Marshall Islands
the men were not already beating Philippines
or physically hurting them.
Adolescent girls’ self-reported Gabon
rates of perpetrating physical Liberia
violence against their partners
are generally low across the Timor-Leste
24 countries for which data are Haiti

available (Figure 7.5).52 However,
Comoros
more than 1 in 10 ever-married
girls aged 15 to 19 admitted Zambia
to having initiated physical
Cameroon
violence against their husbands
or partners in five countries: Uganda
Equatorial Guinea, Gabon, Liberia,
Peru
the Marshall Islands and the
Philippines. It is important to note Kenya
that such questioning is likely to
Zimbabwe
result in some underreporting,
since many respondents may United Republic of Tanzania
have been reluctant to admit Ghana
such acts of violence. The
Malawi
available data also reveal that

the likelihood of a girl or woman Mozambique
initiating violence against her
partner is directly associated with Nigeria
her partner’s violent behaviour Côte d’Ivoire

towards her. In all 24 countries,
India 0.4
girls and women aged 15 to 49
who experienced partner physical Tajikistan 0.4
violence themselves were much
Nepal 0.3
more likely to say they were
physically violent towards their Kyrgyzstan 0
partner than girls and women
0 10 20 30 40 50
who were never the victims of
spousal physical abuse (results
Notes: Data for Côte d’Ivoire refer to adolescent girls who are currently married or in union. Data for Equatorial Guinea and the
not shown).53 Marshall Islands are based on 25-49 unweighted cases.
142
Prevalence of partner Figure 7.6

violence against ever-
married adolescent boys In the Plurinational State of Bolivia, about
Comparable data on intimate partner violence 60 per cent of adolescent boys report
experienced by adolescent boys who were experiencing some form of intimate partner
ever married or in union were collected in five violence
countries that conducted a DHS. Data from three Percentage of ever-married boys aged 15 to 19 years who
of these countries are shown in Figure 7.6.54 In ever experienced any physical, sexual or emotional violence
committed by their wives or partners in the Plurinational
the Plurinational State of Bolivia and Mozambique, State of Bolivia and Mozambique and percentage of ever-
emotional abuse is the most commonly reported married boys and men aged 15 to 24 years who ever
form of intimate partner violence experienced experienced any physical violence committed by their wives
or partners in Ukraine
by boys. In the former, a higher proportion of
adolescent boys than girls (aged 15 to 19) reported Any physical, sexual or emotional violence Any physical violence
Any emotional violence Any sexual violence
experiences of emotional, physical or sexual
violence perpetrated by their partner in the 12 100
months preceding the survey (62 per cent versus 90
48 per cent, respectively). While rates of recent 80
emotional and sexual violence by a partner among 70
adolescent boys were higher than those reported 60
by adolescent girls, rates of partner physical 50
abuse were roughly the same for both sexes. In 40
Mozambique, adolescent boys are only slightly less 30
likely than girls to experience any of the three types 20
of partner violence (30 per cent and 37 per cent, 10
respectively). Reported rates of partner emotional 0
Bolivia (Plurinational Mozambique Ukraine
violence there are about the same for both sexes State of)
(30 per cent for boys and 29 per cent for girls), but
Notes: Data for the Plurinational State of Bolivia are based on 25-49 unweighted cases
adolescent girls are far more likely than adolescent and refer to boys who experienced any physical, sexual or emotional partner violence in the
boys to report experiences of physical or sexual last 12 months; emotional violence includes the wife or partner humiliating or insulting him,
accusing him of being unfaithful, attempting to limit contact with his family and threatening
violence committed by their partners. Finally, in to leave him or to take away the child(ren); physical violence includes being pushed, beaten
with hands or hard objects and choked or burned; sexual violence includes only forced sexual
Ukraine, 2 per cent of boys and men aged 15 to intercourse. Data for Ukraine refer to boys and men who were ever hit, slapped, kicked or
physically hurt in another way by their wives or partners when they were not already beating
24 reported being the victims of physical violence or physically hurting them; data on partner emotional and sexual violence are not available
committed by a partner; this proportion was 7 per for Ukraine.

cent among girls and women of the same age.
A national survey conducted in urban areas of Brazil per cent and 5 per cent, respectively). Around 2 per
in 2005 (the Sexual Behaviour and Perceptions of cent of adolescent boys reported being physically
the Brazilian Population Concerning HIV/AIDS)55 forced by a partner to have sex, 2 per cent said
also collected information on experiences of partner they had sexual intercourse when they did not want
sexual violence among adolescents (aged 16 to to because they were afraid of what their partner
19). In that survey, sexual violence was defined as might do, and less than 1 per cent said they were
physically forced sexual intercourse, having sexual forced to do something sexual that they found
intercourse when a person did not want to because humiliating or degrading.
he or she was afraid of what the partner might
do, and being forced to do something sexual that In a 2011-2012 national survey carried out in
was degrading or humiliating. The rates of partner Nicaragua,56 experiences of partner violence were
sexual violence reported by adolescent boys were collected from ever-married boys and men aged
only slightly lower than those reported by girls (3 15 to 49. Physical violence included the same acts
143
as those included in the DHS module on domestic or sexual violence against their wife or partner at
violence, while emotional violence also included some point (results not shown).58
doing things on purpose to frighten or intimidate
him in addition to the forms of emotional violence Dating violence
covered in the DHS module (see Box 7.1). The
In some Global School-based Student Health
definition of sexual violence used was: being
Surveys (GSHS), students aged 13 to 15 were
physically forced to have sexual intercourse, forced
asked if they had a boyfriend or girlfriend in the
to perform other sexual acts, and agreeing to have
past 12 months and, if so, whether their partner hit,
sexual intercourse for fear of what the partner might
slapped or physically hurt them on purpose during
do. Around one in six (16 per cent) adolescent boys
aged 15 to 19 reported experiencing physical and/or that time (Figure 7.7). In the 2004 survey in Zambia,
sexual violence at some point by a partner, including 21 per cent of students reported dating violence
15 per cent who said they were physically abused of a physical nature in the previous 12 months,
and 2 per cent who said they were sexually abused. including 23 per cent of boys and 19 per cent of
An additional 28 per cent of boys said they were girls. According to the 2004 GSHS, dating violence
subjected to emotional abuse by their partners. is also widespread in Jordan, where 29 per cent
of boys and 15 per cent of girls reported physical
In a few DHS, ever-married boys and men were also violence in the context of dating. Comparable data
asked whether they ever initiated partner violence from the 2003 survey in Swaziland and 2004 survey
against their wife or partner.57 In Mozambique, in Namibia show lower prevalence of physical dating
13 per cent of adolescent boys aged 15 to 19 violence: In those countries, 8 per cent of boys and
admitted to having ever hit, slapped, kicked or done 6 per cent of girls in Swaziland and 16 per cent of
something else to physically hurt their wives or boys and 9 per cent of girls in Namibia reported
partners when they were not already being beaten recent physical abuse at the hands of a boyfriend or
or physically hurt by them. In Cameroon, fully half girlfriend. Of those teens who reported dating in the
(52 per cent) of ever-married adolescent boys aged 2007 GSHS survey in the former Yugoslav Republic
15 to 19 reported committing emotional, physical of Macedonia, 10 per cent of boys and 6 per cent
Figure 7.7
In some countries, a significant proportion of adolescents report incidents of violence in their

dating relationships
Among those students aged 13 to 15 years who had a boyfriend or girlfriend, the percentage who were hit, slapped or physically
hurt by their boyfriend or girlfriend in the past 12 months, by sex
50
Boys Girls
40
30
20
10
0
Jordan Zambia Namibia The former Yugoslav Swaziland
144
of girls said they were hit, slapped or physically hurt A representative survey of around 1,600 high
in the context of a relationship during the previous school students aged 14 to 16 from one city in
year. Italy (Lucca) and two cities in Canada (Kingston
and Toronto) included items to assess perpetration
More recent data from the United States show a of physical dating violence.62 Data were collected
similar incidence of dating violence. In 2013, 10 from the Canadian sample in 2001 and from the
per cent of high school students in grades 9 to 12 Italian sample in 2002. Physical violence included
who had dated or went out with someone in the the following acts (taken from the Conflict Tactics
12 months preceding the survey reported being hit, Scale): pushing, grabbing or shoving; spitting on;
slammed into something or injured with an object pulling hair or scratching; slapping, kicking or biting;
or weapon on purpose one or more times by their physically twisting their arm; throwing, smashing,
boyfriend or girlfriend in the past year. Girls were hitting or kicking an object; slamming or holding
more likely to report physical dating violence than against a wall; hitting or trying to hit with an object;
boys (13 per cent and 7 per cent, respectively). and choking, punching or beating. In both countries,
Reported dating violence of a physical nature was around 30 per cent of adolescents reported
slightly higher among students in grade 12 (12 per perpetrating at least one act of physical partner
cent) than among those in grade 9 (9 per cent). violence “rarely, sometimes, often or always”
An additional 10 per cent of the students reported within their current or past dating relationship(s).
being kissed, touched or physically forced to have However, certain differences between the sexes
sexual intercourse when they did not want to one or were found, with boys in Canada significantly more
more times in the previous 12 months by someone likely than girls to report perpetrating physical dating
they were dating. Again, the reported prevalence of violence. Pushing, grabbing or shoving were the
sexual dating violence was higher among female most commonly reported forms of physical dating
than male students (14 per cent and 6 per cent, violence in Canada (19 per cent), while throwing,
respectively).59 smashing, hitting or kicking an object were most
often cited among adolescents in Italy (14 per cent).
According to 2008 administrative data from Canada, Slapping, kicking and biting were also common,
police-reported rates of dating violence (including with 11 per cent of adolescents in both countries
physical assault, sexual assault and threats as well reporting that they perpetrated this form of dating
as harassment) among unmarried adolescent teens violence.
aged 15 to 19 were higher for girls than boys at a
margin of nearly 10 to 1 (395 per 100,000 unmarried
girls and 42 per 100,000 unmarried boys).60 Younger

adolescents aged 12 to 14 accounted for less than
2 per cent of victims of dating violence reported
to the police, with nearly all of these victims being
female. The most common acts of dating violence
perpetrated against adolescent victims aged 12 to
14 were sexual assault and related offences such
as sexual interference.61 These forms of violence
comprised 45 per cent of all dating violence
incidents reported to the police for the 12- to
14-year-old age group, in comparison to 3 per cent
for those above the age of 15. It is important to
keep in mind, however, that these data only reflect
incidents of dating violence reported to the police
and are therefore likely to be an underestimate of
the true extent of the problem.
145
Chapter 8 Exploring
attitudes
and social
norms
Although violence against children is found
worldwide, the reasons it occurs and persists may
vary in different cultures. Understanding the norms
that govern a society can provide clues to the
underlying causes of such violence and how it can
be prevented.
Social norms are the standards of conduct that

regulate a society. They are the unspoken rules
that govern what is and what is not acceptable
behaviour and how individuals and groups should
interact. When social norms are internalized,
they influence individual attitudes and beliefs
as well as the ways in which people behave. For
example, norms that support violence can be used
to justify violent behaviour and practices, excuse
perpetrators’ actions and blame victims for events
while trivializing or minimizing their suffering.
Cultural factors, therefore, can play a key role both
in the perpetration of violent behaviour at individual
and community levels and in shaping the responses
of both victims and institutions.
Different social norms can help explain the
widespread use of violence against children. In
certain cultures, for example, violence may be
perceived as a normal and acceptable way to resolve
conflict. Children may be considered to have a low
status in society and within the family, which might
be viewed as a justification for adults to use power

to exercise control and coercion. Girls, in particular,
© UNICEF/NYHQ2000-1004/Noorani
may maintain this low status as they grow older and

enter into relationships and early marriage. Boys, on
the other hand, may feel constrained to live up to a
male stereotype of being powerful and strong and/
or the pressure of having to be the main provider
for the family, and this may play a role in their
propensity for aggression and violence.
146
While societal-level beliefs and norms certainly exist 40 to 50 per cent of men the same age. Among
for every form of violence, this chapter focuses on younger respondents (aged 13 to 17), around one
perceptions and attitudes towards three specific third of girls and 40 per cent of boys also agreed
types for which comparable data for a large cross- with the statement. In Zimbabwe, the opinion that
section of countries are available: wife-beating, women should tolerate violence for the sake of
corporal punishment of children and child sexual their family was widespread among both adults and
abuse. Data on attitudes towards wife-beating adolescents of both sexes. Women aged 18 to 24
offer clues on how girls and women are perceived were more likely than men to support this notion
within a given society. Such information may also (78 per cent and 69 per cent, respectively). Among
help explain why intimate partner violence against adolescents aged 13 to 17, girls were also slightly
adolescent girls persists in many countries. The more likely to endorse the statement than boys (78
fact that some girls and women can justify violence per cent and 74 per cent, respectively).
as a means to ‘correcting’ or controlling socially
undesirable behaviour may also impact their own Globally, nearly half (44 per cent) of adolescent girls
children’s exposure to violence within the home. aged 15 to 19 think a husband or partner is justified
Examining attitudes towards corporal punishment in hitting or beating his wife or partner under certain
sheds light on cultural views regarding child-rearing circumstances – if the wife argues with her husband,
and offers relevant insights for the development of goes out without telling him, neglects the children,
strategies aimed at promoting positive parenting refuses to have sexual relations with him or burns
practices. In a similar vein, perceptions and opinions the food (Figure 8.1).4 In sub-Saharan Africa and the
about child sexual abuse are an indication of both the Middle East and North Africa, this proportion rises
status afforded to children within a given society, to more than half. In Central and Eastern Europe
and the Commonwealth of Independent States
the level of knowledge commonly held about the
(CEE/CIS), it drops down to 28 per cent.
issue and the public’s willingness and motivation to
address it.
Figure 8.1
Attitudes towards wife-beating
Violence against girls and women persists for Nearly half of adolescent girls worldwide say
many reasons. One contributing factor may be the wife-beating can be justified under certain
widely held view that girls and women have low circumstances
status in society and are expected to comply with, Percentage of girls aged 15 to 19 years who think that a
and conform to, certain defined gender roles of husband/partner is justified in hitting or beating his wife or
partner under certain circumstances, by region
devoted mothers and wives. Several studies have
CHAPTER 8: Exploring attitudes and social norms

demonstrated that rates of violence against girls Eastern and Southern Africa
and women are higher in societies characterized
by unequal gender roles, where ‘manhood’ is
defined in terms of dominance and ‘womanhood’ is West and Central Africa
constrained by the fulfilment of certain rigid codes
South Asia
of conduct.1 When such roles are not fulfilled,
partner violence may be seen as a justified form of CEE/CIS
punishment in certain contexts.
One illustration of this can be found in the Violence World

Against Children Surveys (VACS) in Kenya2 and 0 10 20 30 40 50 60 70 80 90 100
Zimbabwe,3 which included a question on whether Notes: The world estimate is based on a subset of 102 countries covering 59 per cent of
respondents felt that women should tolerate the population of girls aged 15 to 19 years worldwide. Regional estimates represent data
covering at least 50 per cent of the regional population. Data coverage was insufficient
violence in order to keep the family together. In to calculate regional estimates for East Asia and the Pacific and Latin America and the
Caribbean.
Kenya, between 30 and 40 per cent of women Source: UNICEF global databases, 2014, based on DHS, MICS and other nationally repre-
aged 18 to 24 agreed with this statement, as did sentative surveys, 2002-2013.
147
Map 8.1
In a majority of countries, more than 4 in 10 girls think wife-beating is sometimes justifiable

Percentage of girls aged 15 to 19 years who think that a husband/partner is justified in hitting or beating his wife or partner under
certain circumstances, by country
Above 60%
41% - 60%
20% - 40%
Less than 20%
data in UNICEF global database
Sudan and South Sudan has not yet been determined. The final status of the Abyei area has not yet been determined. For Argentina, the sample was national and urban (municipalities with a population of
more than 5,000), since the country’s rural population is scattered and accounts for less than 10 per cent of the total. Data for Lebanon refer to currently married girls. Data for Bangladesh, Egypt, Jordan,
Maldives, Pakistan, Somalia and Sri Lanka refer to ever-married girls. Data for the Congo, Guinea-Bissau, Jordan, Nicaragua and Turkey differ from the standard definition.
At the country level, more than half of girls aged beating is sometimes justified in 21 of 63 countries
15 to 19 believe that wife-beating is sometimes with data; in over half of the countries, more than
justified in 40 of 102 countries; in an additional 23 one third of them can justify wife-beating (results
countries, more than one third of girls agree with the not shown). Agreement with wife-beating among
statement. Overall, agreement with wife-beating boys is highest in Tuvalu (83 per cent) and lowest in
under certain circumstances ranges from over 80 Ukraine (2 per cent).
per cent of girls in Afghanistan, Guinea, Jordan, Mali
and Timor-Leste to less than 5 per cent in Argentina, Comparing attitudes between the sexes
Barbados, Belarus, Bosnia and Herzegovina, Costa

As noted above, the justification of violence against
Rica, Georgia, Serbia and Ukraine (Map 8.1). The
girls and women can be viewed as a manifestation
largest variations can be found among countries in
of patriarchal oppression when girls and women
CEE/CIS, where the percentage of girls who think do not fulfil expected gender roles. If this is true,
a husband is sometimes justified in hitting/beating it might be expected that stated support for wife-
his wife ranges from 1 per cent in Bosnia and beating would be higher among males than females.
Herzegovina to 63 per cent in Uzbekistan. However, the available data show that this is not
the case in most countries, which may suggest that
Supportive attitudes towards wife-beating are girls are more thoroughly socialized than boys to
also widespread among adolescent boys. In both gender norms that assign wives a lower status than
Eastern and Southern Africa and South Asia, their husbands.6 In 28 of 60 countries with data on
close to 50 per cent of boys aged 15 to 19 think both sexes, a larger proportion of girls than boys
a husband is justified in hitting his wife under believe that wife-beating is sometimes justified;
certain circumstances; in West and Central Africa, in 14 of these countries, the gap between the
the share is slightly more than one third.5 At the two sexes exceeds 10 percentage points (Figure
country level, more than half of boys believe wife- 8.2). This pattern is more common in South Asia,
148
sub-Saharan Africa, and East Asia

Figure 8.2
and the Pacific than in other regions.
In most countries, girls are more likely to justify wife- In Cambodia, Mongolia, Pakistan,
beating than boys Rwanda and Senegal, girls are
Percentage of girls aged 15 to 19 years and boys aged 15 to 19 years who around twice as likely as boys to
think that a husband/partner is justified in hitting or beating his wife or partner think a husband is sometimes
under certain circumstances
justified in hitting his wife. The
Guinea opposite is true in Armenia, Bosnia
Timor-Leste and Herzegovina, and Serbia. It is
Kiribati interesting to note that justification
Burundi
Congo of wife-beating has been found to
Solomon Islands be more common among female
Tuvalu
Ethiopia victims of partner violence.7 It is
Uzbekistan
Uganda
reasonable to assume that this
Zambia may be due to girls’ and women’s
Senegal
Gabon efforts to rationalize and cope
Samoa
Equatorial Guinea
with the violence against them
Kenya and against other females in their
Vanuatu families or communities.8
Rwanda
Niger
Ghana
Pakistan
Variations by socio-
United Republic of Tanzania demographic characteristics
Eritrea
Côte d’Ivoire When examining data by age
Cameroon
Liberia groups, it is reasonable to expect
Zimbabwe
Lesotho that older generations might hold
Marshall Islands more traditional views of the
Madagascar
India roles of women and men. If this
Indonesia
Comoros
is true, it is likely that they would
Swaziland be more inclined to think that wife-
Cambodia
Azerbaijan beating is justifiable if women do
Burkina Faso
Namibia
not live up to certain gender role

Nigeria expectations. In fact, available data
Kyrgyzstan
Nepal show that adolescents are equally
Mozambique likely as older women and men to
Albania
Haiti justify wife-beating under certain
Cabo Verde circumstances (Figures 8.3A and
Girls
Guyana 8.3B). They show, for instance, that
Malawi Boys
Honduras the percentage of females who
Benin
Mongolia
hold this view remains virtually the
Republic of Moldova same across all regions regardless
Kazakhstan
Armenia of whether those surveyed are
Belarus
Serbia
older or younger. In other words,
Ukraine girls aged 15 to 19 have the same
opinion as women aged 45 to
0 10 20 30 40 50 60 70 80 90 100
49. The same pattern can also be
Notes: Data for the Congo differ from the standard definition. Data for Indonesia refer to currently married boys. Data observed among boys and men,
for Pakistan refer to ever-married girls and boys and are based on 25-49 unweighted cases for boys.
Source: UNICEF global databases, 2014, based on DHS, MICS and other nationally representative surveys, 2002-2013. although it is less pronounced.
149
Adolescents are as likely to justify wife-beating as older Agreement that wife-beating
women and men is sometimes justifiable varies
significantly by level of education.
Figure 8.3A In fact, more than any other variable,
little or no education appears to be
Percentage of girls aged 15 to 19 years and women aged 45 to 49 years who
think that a husband/partner is justified in hitting or beating his wife or partner strongly associated with justifi-
under certain circumstances, by region cation for wife-beating among
Girls aged 15 to 19 years Women aged 45 to 49 years both females and males. The data
illustrated in Figure 8.4 support this
point. Significant differences are
found in the level of justification
for wife-beating between girls and
West and Central Africa boys with little or no education and
those with a secondary or higher
South Asia education. Less educated girls and
boys are much more likely overall
CEE/CIS
to report that a husband is justified
in hitting or beating his wife for at
least one of the reasons previously
World mentioned. Among girls, this gap
is most pronounced in the Middle
0 10 20 30 40 50 60 70 80 90 100
East and North Africa, where 67 per
cent of girls with no education think
Notes: The world estimate is based on a subset of 102 countries covering 59 per cent of the population of girls
aged 15 to 19 years and 51 per cent of the population of women aged 45 to 49 years worldwide. Regional estimates a husband is sometimes justified
represent data covering at least 50 per cent of the regional population. Data coverage was insufficient to calculate
regional estimates for East Asia and the Pacific, and Latin America and the Caribbean. in beating his wife compared to
35 per cent of those with higher
levels of education. Among boys,
differences are most marked in
Figure 8.3B
Eastern and Southern Africa, where
Percentage of boys aged 15 to 19 years and men aged 45 to 49 years who those with no education are more
think that a husband/partner is justified in hitting or beating his wife or partner
under certain circumstances, by region
than one and a half times as likely
to justify wife-beating under certain

Boys aged 15 to 19 years Men aged 45 to 49 years circumstances as those with
secondary or higher education.
While this general pattern tends to
hold true at the country level among
South Asia
both adolescent girls and boys,
there are also large variations in
the level of stated justification even
West and Central Africa within the same educational level.
Exceptions to this pattern can also
be found in some countries where
0 10 20 30 40 50 60 70 80 90 100
a girl’s or boy’s level of education
Notes: Regional estimates represent data covering at least 50 per cent of the regional population. Data coverage does not appear to influence their
was insufficient to calculate a global estimate and regional estimates for CEE/CIS, East Asia and the Pacific, Latin
America and the Caribbean, and the Middle East and North Africa.
views on the acceptability of wife-
Source: UNICEF global databases, 2014, based on DHS, MICS and other nationally representative surveys, 2002-2013. beating.
150
Figure 8.4
Uneducated girls and boys are more likely to justify wife-beating than their more educated peers
Percentage of girls aged 15 to 19 years and boys aged 15 to 19 years who think that a husband/partner is justified in hitting or beat-
ing his wife or partner under certain circumstances, by level of education and by region
Boys Girls
n/a Middle East and North Africa
No education
Primary education
Secondary or higher
education West and Central Africa
South Asia
100 90 80 70 60 50 40 30 20 10 0 0 10 20 30 40 50 60 70 80 90 100
Notes: Regional estimates represent data covering at least 50 per cent of the regional population. Data coverage was insufficient to calculate global and regional estimates for girls and boys for CEE/
CIS, East Asia and the Pacific, and Latin America and the Caribbean and for boys for the Middle East and North Africa.
Figure 8.5
Across all regions, wealthier girls and boys are less likely to justify wife-beating under certain
circumstances than poorer girls and boys
Percentage of girls aged 15 to 19 years and boys aged 15 to 19 years who think that a husband/partner is justified in hitting or beat-
ing his wife or partner under certain circumstances, by wealth quintiles and by region
Boys Girls
n/a Middle East and North Africa

Poorest
Richest West and Central Africa
South Asia
n/a CEE/CIS

100 90 80 70 60 50 40 30 20 10 0 0 10 20 30 40 50 60 70 80 90 100
Notes: Regional estimates represent data covering at least 50 per cent of the regional population. Data coverage was insufficient to calculate global and regional estimates for girls and boys for
East Asia and the Pacific, and Latin America and the Caribbean and for boys for the Middle East and North Africa and CEE/CIS.
As with education, household wealth appears to beating is sometimes justifiable. For adolescent
be strongly associated with the justification of boys, the difference is most pronounced in South
wife-beating among both girls and boys. Across all Asia, where 51 per cent of the poorest boys justify
regions, and for both sexes, those from the poorest wife-beating under certain conditions compared to
20 per cent (quintile) of the population are more 24 per cent of the richest boys.
likely than those from the wealthiest quintile to
agree that husbands are justified in hitting or beating While variations in opinion range widely across
their wives for at least one of the reasons specified countries, in almost all of the countries with
(Figure 8.5). This gap is particularly striking in CEE/ available data both girls and boys from the poorest
CIS for girls, where the poorest girls are nearly four quintile are more likely than those from the richest
times as likely as the richest girls to think that wife- quintile to accept justifications for wife-beating.
151
Although males and females may differ with cultures, occasional slaps or spankings are viewed
respect to the extent of their support for wife- as a ‘normal’ method of controlling an ‘unruly’ child.
beating, opinions seem to converge on the main Understanding caregivers’ beliefs about the best
circumstances under which this form of violence way to bring up a child provides valuable insights
can be justified. Neglecting the children is the to inform the development of appropriate policy
most commonly cited reason for justifying wife- responses to violence in the home.
beating among both female and male adolescent
respondents (Tables 8.1A and 8.1B). Available data from 59 countries or areas suggest
that physical punishment is not commonly
considered a necessary disciplinary practice (Figure
Attitudes towards corporal
8.6).9 The proportion of adults who think that
punishment of children physical punishment is necessary to properly raise
In some contexts, any form of physical aggression or educate children ranges from a low of 3 per cent
against children is perceived as a form of abuse and in Armenia and the former Yugoslav Republic of
laws have been established to prevent it. In other Macedonia to a high of 82 per cent in Swaziland.
Neglecting the children is the most commonly cited reason for justifying wife-beating among
girls and boys across almost all regions
Table 8.1A
Percentage of girls aged 15 to 19 years who think that a husband/partner is justified in hitting or beating his wife or partner under
certain circumstances, by reason and by region
If she goes out If she neglects If she argues If she refuses If she burns
without telling him the children with him sex with him the food
CEE/CIS 20 19 17 6 7
Eastern and Southern Africa 32 40 31 24 24
Middle East and North Africa 40 40 29 28 17
South Asia 27 32 30 15 18
West and Central Africa 33 35 33 26 20
World 27 32 26 16 16
Notes: The world estimate is based on a subset of 102 countries covering 59 per cent of the population of girls aged 15 to 19 years worldwide. Regional estimates represent data covering at least 50
per cent of the regional population. Data coverage was insufficient to calculate regional estimates for East Asia and the Pacific, and Latin America and the Caribbean.
Table 8.1B
Percentage of boys aged 15 to 19 years who think that a husband/partner is justified in hitting or beating his wife or partner under
certain circumstances, by reason and by region
If she goes out If she neglects If she argues If she refuses If she burns
without telling him the children with him sex with him the food
Eastern and Southern Africa 25 31 24 17 15
South Asia 26 31 30 12 15
West and Central Africa 19 22 21 14 12
Notes: Regional estimates represent data covering at least 50 per cent of the regional population. Data coverage was insufficient to calculate a global estimate and regional estimates for CEE/CIS,
East Asia and the Pacific, Latin America and the Caribbean, and the Middle East and North Africa.
152
Only in four countries – Liberia,

Figure 8.6
Nigeria, Swaziland and Vanuatu –
Only a minority of adults in most countries or areas think do a majority of adults believe that
physical punishment is a necessary form of discipline physical punishment is a necessary
Percentage of adults who think that physical punishment is necessary to part of child discipline. Countries
raise/educate children in sub-Saharan Africa show the
highest levels of support for
Swaziland corporal punishment. In 16 of the 18
Nigeria
Liberia countries with data in that region,
Vanuatu
Ghana more than one in four respondents
Niger
Tunisia
think that physical punishment is
Yemen needed to raise a child. Countries in
Sierra Leone
Cameroon CEE/CIS are on the opposite side of
Egypt the spectrum: In all countries there
Afghanistan with available data, fewer than
Morocco
Gambia one in four respondents think that
Côte d’Ivoire
Chad physical punishment is needed to
Burkina Faso educate a child.
Barbados
Mauritania
Togo
Djibouti
Discrepancies between
Central African Republic attitudes and practice
Haiti
Jamaica
Belize
Although physical punishment
Guinea-Bissau is not usually regarded as an
Trinidad and Tobago
Lebanon appropriate form of discipline, data
Jordan
Guyana reveal that it is widely practised,
Azerbaijan as was shown in Chapter 5. In all
Congo
Iraq but one country (Swaziland), the
State of Palestine
Saint Lucia percentage of adults who think
Viet Nam
Mongolia
physical punishment is necessary
Tajikistan is consistently lower than the
Republic of Moldova
Algeria percentage of children aged 2 to
14 who are subjected to this form

Albania
Georgia of discipline (Figure 8.7). This
Suriname suggests that many households
Costa Rica
Ukraine practise physical punishment
Dominican Republic
Kyrgyzstan
even when they do not consider
Belarus it necessary. The reasons behind
Panama
Kazakhstan this are likely to be interlinked and
Serbia
Montenegro complex. Part of the reason may
Argentina be that the respondent (mother/
Armenia primary caregiver or other adult
0 10 20 30 40 50 60 70 80 90 100 household member) may not be
the only person in the household
Notes: For Argentina, the sample was national and urban (municipalities with a population of more than 5,000), responsible for disciplining
since the country’s rural population is scattered and accounts for less than 10 per cent of the total. Data for Algeria,
Azerbaijan, Burkina Faso, Cameroon, Côte d’Ivoire, Djibouti, Egypt, Georgia, Guinea-Bissau, Guyana, Jordan, children. In fact, she or he may not
Kyrgyzstan, Montenegro, the Syrian Arab Republic, Tajikistan, Trinidad and Tobago, Vanuatu and Yemen refer to
mothers/primary caregivers. Data for the Dominican Republic, Lebanon and Morocco refer to children aged 2 to 14 be responsible for discipline at all.
years whose mother/primary caregiver thinks that physical punishment is necessary to raise/educate children. Data In other words, children may be
for all other countries refer to any adult household member who responded to questions about child discipline.
Source: UNICEF global databases, 2014, based on DHS, MICS and other nationally representative surveys, 2005-2013. subjected to physical punishment
153
by others living in the household
Figure 8.7
even if the respondent does not
support the practice. Even in Many children are subjected to physical punishment
situations where the respondent even when adults in the household do not think it is a
does use physical force, individual necessary form of discipline
preferences or beliefs may not
Percentage of adults who think that physical punishment is necessary to
be enough to influence practices raise/educate children and the percentage of children aged 2 to 14 years who
if societal norms on child-rearing experienced any physical punishment in the previous month
encourage the use of physical Egypt
Yemen
punishment as a necessary or Jamaica
Vanuatu
prevailing disciplinary method. Central African Republic
Haiti
Nigeria
Variations by socio- Cameroon
demographic characteristics Syrian Arab Republic
Mauritania
Chad
Another possible reason why Togo
many children experience physical State of Palestine
Liberia
punishment at home (whether or Algeria
Guinea-Bissau
not adults in the household believe Tunisia
Gambia
it is necessary) may be a perceived Ghana
absence – or lack of knowledge – Côte d’Ivoire
Afghanistan
of alternative, non-violent methods. Congo
Djibouti
Available data confirm that adults Jordan
with no or little education are more Niger
Swaziland
likely to find physical punishment a Sierra Leone
Guyana
necessary method of educating or Iraq
Albania
raising children in most countries Tajikistan
(Figure 8.8). In Burkina Faso, for Suriname
Burkina Faso
example, 39 per cent of mothers/ Belize
Barbados
primary caregivers with no formal Viet Nam
education believe physically Trinidad and Tobago
punishing a child is necessary Azerbaijan
Georgia
compared to 24 per cent of those Republic of Moldova
Argentina
with secondary or higher education. Montenegro
In Yemen, this association is Saint Lucia
even more pronounced: In that Armenia
country, 51 per cent of mothers/ Kyrgyzstan
Serbia
primary caregivers with no formal Belarus Children aged 2 to 14 years who
education feel it is necessary to use Ukraine experienced any physical punishment
Costa Rica Adults think that physical punishment
physical punishment to educate a Kazakhstan is necessary
Panama
child compared to 21 per cent of Mongolia
those with a secondary or higher 0 10 20 30 40 50 60 70 80 90 100
education. Such differences persist
even in countries with relatively Notes: For Argentina, the sample was national and urban (municipalities with a population of more than 5,000), since
low support at the national level. the country’s rural population is scattered and accounts for less than 10 per cent of the total. Data for Belarus pertain-
ing to the percentage of children aged 2 to 14 years who experienced any physical punishment in the previous month
In Bosnia and Herzegovina, where differ from the standard definition. Data for Kyrgyzstan refer to children aged 3 to 14 years and for Panama to children
aged 1 to 14 years. Data for Algeria, Azerbaijan, Burkina Faso, Cameroon, Côte d’Ivoire, Djibouti, Egypt, Georgia,
around 14 per cent of adults Guinea-Bissau, Guyana, Jordan, Kyrgyzstan, Montenegro, the Syrian Arab Republic, Tajikistan, Trinidad and Tobago,
overall are supportive of physical Vanuatu and Yemen refer to mothers/primary caregivers. Data for all other countries refer to any adult household
member who responded to questions about child discipline.
punishment of children, those who Source: UNICEF global databases, 2014, based on DHS, MICS and other nationally representative surveys, 2005-2013.
154
Figure 8.8
Overall, adults with little or no education are more likely than their more educated peers to think
that physical punishment is necessary in disciplining children
Percentage of adults who think that physical punishment is necessary to raise/educate children, by level of education
Swaziland
Liberia
Vanuatu
Nigeria No education
Yemen Primary education
Ghana Secondary or higher
education
Cameroon
Niger
Morocco
Tunisia
Sierra Leone
Côte d’Ivoire
Gambia
Afghanistan
Burkina Faso
Mauritania
Chad
Togo
Haiti
Djibouti
Guyana
Guinea-Bissau
Congo
Lebanon
Saint Lucia
Republic of Moldova
Belize
Mongolia

Iraq
Azerbaijan
Jordan
Algeria
Suriname
Panama
Tajikistan
Viet Nam
Costa Rica
Kazakhstan
Argentina
0
Kyrgyzstan
0 10 20 30 40 50 60 70 80 90 100
Notes: Only countries with available data for all three levels of education are included in the figure. For some countries, results for some education levels are based on less than 25 unweighted cases
and therefore are not included here. For Argentina, the sample was national and urban (municipalities with a population of more than 5,000), since the country’s rural population is scattered and
accounts for less than 10 per cent of the total. Data for Algeria, Azerbaijan, Burkina Faso, Cameroon, Côte d’Ivoire, Djibouti, Guinea-Bissau, Guyana, Jordan, Kyrgyzstan, the Syrian Arab Republic,
Tajikistan, Vanuatu and Yemen refer to mothers/primary caregivers. Data for Lebanon and Morocco refer to children aged 2 to 14 years whose mother/primary caregiver thinks that physical punish-
ment is necessary to raise/educate children. Data for all other countries refer to any adult household member who responded to questions about child discipline. Data on the proportions of adults with
no education for Bosnia and Herzegovina and Kyrgyzstan, and adults with no education and those with primary education for the Republic of Moldova are based on 25-49 unweighted cases.
155
have no education are twice as likely to think that The relationship between attitudes towards
it is necessary to raise or educate children properly corporal punishment and wealth is also complex
than those with a secondary or higher education. and is linked to other socio-demographic factors,
including education. Adults with low economic
A possible explanation could be that educational status are found to be more supportive of corporal
settings create opportunities for open discourse punishment than their wealthier counterparts in
and exposure to information and attitudes that approximately three quarters of countries with
might discourage the use of force with children. available data (Figure 8.9). In Albania, for instance,
It is also possible that more educated people are one in five respondents from the poorest quintile
of the population believe in the need for physical
more aware of social norms that do not condone
punishment compared to one in 20 respondents
violence, including the use of violent disciplinary
from the richest quintile. Suriname is a similar
practices, thus influencing their responses. Here
case. There, adults from the poorest quintile are
it is important to keep in mind that associations
three times more likely than those from the richest
with socio-demographic factors should always be
quintile to say that physical punishment is necessary
interpreted with caution since they can be due
in the context of raising children. Exceptions to this
to the confounding influence of other, possibly pattern are found in Barbados and Kyrgyzstan, where
overlapping, variables. For instance, more educated the highest levels of support for physical discipline
adults are also more likely to be residing in wealthier can be seen among the richest members of
households. society.
Variations by relationship to the child

Figure 8.9
A few studies have explored the differences
In most countries, wealthier segments of between mothers and fathers in the use of violent
the population are less likely to believe in punishment and in their opinions of its effectiveness.
the necessity of corporal punishment than In general, research indicates that mothers are
their poorer counterparts more engaged than fathers in disciplining children,
mostly due to the fact that they typically spend
Percentage of adults who think that physical punishment is
necessary to raise/educate children, by wealth quintiles more time with them.10 However, the literature is
inconsistent as to whether mothers’ use of violent
100 punishment and their support of it is similar to that
of fathers. Some studies suggest that fathers use
90
corporal punishment more frequently than mothers
80
and tend to justify its use to a greater extent; other
70 studies suggest more similarities than differences
Adults in the richest quintile
60 Adults in the poorest

in mothers’ and fathers’ use of and opinions on
Adults in the richest quintile are
more likely to think
quintile are more likely corporal punishment. Rydstrøm, for instance, found
50 to think
physical punishment is
physical punishment is that in Viet Nam men tend to interfere less often in
necessary
40 necessary
shaping children’s behaviour than women; however,
30 when they do get involved, their involvement usually
implies the use of force.11 She also pointed out that,
20
in most cases, corporal punishment was elicited not
10
because of ‘anger’ or ‘impulses’; rather, many fathers
and grandfathers expressed the view that physical
0 10 20 30 40 50 60 70 80 90 100
Adults in the poorest quintile punishment to correct a child’s misbehaviour should
be a planned action. Other studies indicate that
Note: Each dot represents a country.
Source: UNICEF global databases, 2014, based on DHS, MICS and other nationally repre-
mothers tend to use corporal punishment more
sentative surveys, 2005-2013. frequently than fathers while sharing similar opinions
156
Figure 8.10
Mothers and fathers are equally likely to support the use of physical punishment of children
Percentage of adults who think that physical punishment is necessary to raise/educate children, by relationship to the child
100
Mother Father
90
80
70
60
50
40
30
20
10
0
Swaziland
Nigeria
Ghana
Niger
Tunisia
Sierra Leone
Afghanistan
Chad
Mauritania
Gambia
Togo
Barbados
Haiti
Jamaica
Belize
Jordan
Congo
Saint Lucia
State of Palestine
Iraq
Viet Nam
Mongolia
Republic of Moldova
Costa Rica
Suriname
Ukraine
Belarus
Panama
Serbia
Kazakhstan
Argentina
Armenia

Notes: Data refer to any adult household member who responded to questions about child discipline. For Argentina, the sample was national and urban (municipalities with a population of more than
5,000), since the country’s rural population is scattered and accounts for less than 10 per cent of the total.
on its effectiveness.12 Interviews with parents Comparable data for 37 countries or areas confirm
in China, Colombia, Italy, Kenya, the Philippines, that mothers and fathers are equally likely to express
Sweden, Thailand and the United States indicate that favourable views of physical discipline (Figure 8.10).
mothers use corporal punishment more frequently In almost all countries, an equal percentage of
than fathers but find similar opinions on the need for mothers and fathers report that physical punishment
physical discipline: Overall, 17 per cent of parents is necessary. When differences are observed they
believed that the use of corporal punishment was tend to be very small, with inconsistent patterns.
necessary in child-rearing (Box 8.1).13
157
Box 8.1
Views on corporal punishment across

cultures
The following are selected research Chile mothers and fathers representing a
findings on attitudes towards corpo- total of 1,417 families were recruited
In 2012, UNICEF conducted the
ral punishment in other countries. from schools and asked whether or
Fourth Study on Child Maltreatment
not they considered corporal pun-
and Sexual Abuse in Chile. The
Australia survey was administered to chil- ishment to be necessary in child-
dren and adolescents in the eighth rearing.16
In 2005, the Australian Childhood
Foundation and the National Re- grade, who ranged in age from 12
Across the entire sample, 17 per
to 17 years.15
search Centre for the Prevention of cent of both mothers and fathers
Child Abuse commissioned a study felt that corporal punishment was
Children were asked: “Do you be-
on community attitudes towards lieve that physical punishment is necessary in raising sons, while 15
the physical punishment of children useful in some child-rearing situa- per cent of mothers and 16 per cent
in that country. The survey revealed tions?” Not surprisingly, children’s of fathers believed the same for
high levels of support among sam- belief in the usefulness of physical daughters.
pled parents aged 18 and older.14 A punishment increased in tandem
total of 45 per cent of respondents Parents in Italy and Sweden were
with the level of violence they ex-
indicated that it was reasonable to least likely to consider physical force
perienced from their parents. One
leave a mark on a child as a result necessary in raising a child. In Italy,
in four children who experienced
of physical punishment; 69 per cent the proportions of mothers and fa-
no violence at the hands of their
agreed that it is sometimes neces- thers who believed in its necessity
parents indicated that physical pun-
sary to ‘smack’ a naughty child. Of were at or below 5 per cent, while
ishment is useful in child-rearing.
the latter, only two in five (41 per none of the parents from Sweden
In contrast, the proportions rose to
cent) believed in its effectiveness in expressed the view that corporal
more than one third of children who
experienced psychological aggres- punishment was necessary.
achieving the desired outcome. One
in 10 of them said it was appropriate sion, 42 per cent who experienced
Parents in Colombia, Kenya, the
to use implements such as canes, mild physical punishment and 53
Philippines, Thailand and the United
sticks, belts or slippers to punish a per cent who experienced serious
States expressed stronger support
child. However, it was also evident physical punishment. These find-
for the use of physical discipline
that many parents viewed physi- ings suggest that early experiences
with children. In Colombia, roughly
cal punishment as a last resort. All of violence may increase a child’s
the same proportions of mothers
of the parents believed non-violent acceptance of its use in certain situ-
and fathers (14 per cent and 13 per
strategies, such as making children ations, and that the level of accept-
cent, respectively) thought physical
feel loved, spending time with them ance may increase relative to the
punishment was necessary in rais-
and setting a good example were level of violence the child directly
ing daughters, while 19 per cent of
important in teaching children right experiences. Such findings speak to
mothers and 8 per cent of fathers
from wrong. In contrast, 52 per cent the ways in which violence can be
thought it was a necessity in rais-
of parents perceived smacking as transmitted intergenerationally and
ing sons. In the Philippines, the
how children’s direct exposure to vi-
important in teaching children right proportion of mothers who thought
olence can influence their attitudes
from wrong. Support for physical physical punishment was necessary
towards it – and possibly even their
punishment appears to have de- with sons was slightly, although not
future behaviour.
clined slightly over time, however, significantly, higher than the pro-
dropping from 75 per cent in a com- portion of fathers (20 per cent ver-
parable 2002 survey to 69 per cent China, Colombia, Italy, Kenya, the
sus 15 per cent). The opposite was
in 2005, with more participants ex- Philippines, Sweden, Thailand and
true when it came to daughters:
pressing opposition or uncertainty the United States of America
16 per cent of Filipino fathers saw
on the issue in the more recent As part of the larger 2008-2009 corporal punishment as necessary,
survey. Parenting Across Cultures Project, compared to 13 per cent of mothers.
158
Roughly one in six mothers and one New Zealand Ireland, conducted in 2000, revealed
in five fathers in Thailand saw cor- that a majority of parents did not
A 2013 survey found public support
poral punishment as needed with condone the use of physical disci-
of physical punishment of children
their daughters; around the same pline.22 Furthermore, a majority of
to be declining in New Zealand.18
percentage of mothers and fathers parents reported that physical pun-
According to the study, 40 per
(11 per cent and 10 per cent, re- ishment rarely leads to respect for
cent of respondents thought it was
spectively) believed in its necessity parents and/or behavioural change
‘sometimes alright’ for parents to
with sons. In the United States, 17 in the child. Most parents thought
physically punish children – com-
per cent of mothers saw corporal that physical punishment led to
pared to 58 per cent in 2008, over
punishment as necessary with their parental guilt, one third thought it
80 per cent in 1993 and over 90 per
daughters compared to 11 per cent increased child aggression and one
cent in 1981. The proportion of par-
of fathers; when it came to sons, quarter thought that it frequently or
ents with children under 18 who
13 per cent of US mothers and 16
thought it was acceptable to use always caused long-term emotional
per cent of fathers thought physical
corporal punishment also fell from distress in children.
punishment was required.
62 per cent in 2008 to 35 per cent
Belief in the necessity of physical in 2013. Central and Eastern Europe
punishment was highest in Kenya. A study conducted in 2005-2006 in
More than half of both mothers and The United Kingdom of Great Brit-
some Central and Eastern European
fathers (56 per cent and 54 per cent, ain and Northern Ireland
countries (Bulgaria, Lithuania, Lat-
respectively) saw corporal punish-
Over the past decade, a range of via and Poland)23 included a ques-
ment as required for raising sons;
surveys in the United Kingdom tion on whether adults perceived
rates were slightly less than half (44
have explored parental and public corporal punishment to be an ac-
per cent of mothers and 48 per cent
attitudes towards corporal punish- ceptable form of discipline.24 Re-
of fathers) when it came to daugh-
ment.19 In 2003, the first nationally spondents were asked: “Do you be-
ters.
representative British survey of dis- lieve beating a child by a parent as a
In all of the countries except China, ciplinary practices used on children ‘punishment’ is a disciplinary meas-
no significant difference was found aged 0 to 12 revealed that half of ure that…”. Response categories
in mothers’ and fathers’ views about parents believed it was ‘sometimes’ included “should never be used”,
the necessity to physically punish acceptable to use physical force in “shouldn’t be used but is justified
boys versus girls. In China, parents disciplining a child.20 Roughly 10 in some situations” and “can be
saw corporal punishment as more per cent of parents believed that used whenever the parent believes
necessary in raising sons. Among physical discipline was ‘always’ ac- it will be effective”. In all of the
mothers in China, 36 per cent be- ceptable, while 40 per cent reported
countries surveyed, more than half
lieved corporal punishment was that they did not believe in the use
of respondents thought corporal
required with boys compared to 14 of physical discipline. A more re-
punishment could always or some-
per cent for girls, while one third of cent nationally representative sur-
times be used to discipline children.

fathers (33 per cent) believed corpo- vey in 2007 found similar results: 52
Of the three possible responses, the
ral punishment was necessary with per cent of parents in England and
dominant view in most countries
boys versus 20 per cent for girls. Wales reported that it was some-
(except Bulgaria) was that corporal
times necessary to use physical dis-
punishment should not be used but
cipline.21 While rates remain high,
Finland that it can sometimes be justified. In
there has been a decline: A com-
A series of six nationally represent- parable survey in 1998 found that Bulgaria, the highest proportion of
ative surveys conducted among 88 per cent of respondents thought respondents (47 per cent) thought
adults in Finland between 1981 and that it was sometimes necessary to it should never be used. Only a mi-
2012 demonstrate a consistent de- use physical discipline. nority of adults in each country (less
crease in adult acceptance of cor- than 15 per cent) were of the opin-
poral punishment, dropping from A nationally representative survey ion that parents should use physical
47 per cent in 1981 to 17 per cent of the disciplinary practices and at- punishment whenever they think it
in 2012.17 titudes of 1,000 parents in Northern is needed.
Notes: The above research should be carefully considered in light of study limitations. All the studies draw on self-reported data, which may bias the results
towards underreporting. Moreover, samples in some of the studies were not nationally representative. For this reason, caution should be exercised when attempt-
ing to generalize findings. Furthermore, varying measures used to assess attitudes towards child discipline were used, limiting the possibility of cross-country
comparisons.
159
Comparison of attitudes towards few regions and countries illustrates the variations
different forms of violence and complexity of commonly held beliefs about
this particular form of violence against children.
Data on attitudes towards corporal punishment of
These sources can begin to shed light on the many
children and wife-beating can also be compared;
community norms and societal values that interact
however, it is more relevant to use data on attitudes
to shape people’s perceptions of and opinions on
towards wife-beating from the entire sample of
the issue.25
females and males aged 15 to 49 rather than focusing
on the adolescent population since information
on attitudes towards corporal punishment are For example, a study carried out in six countries in
available on the overall sample of respondents 2008-2009 in the Eastern Caribbean explored the
of all ages. The comparison reveals that in 19 of perceptions of women and men aged 18 and older
the 41 countries with data for both indicators, a about child sexual abuse.26 It found that the vast
much larger proportion of respondents justify wife- majority of respondents (76 per cent) agreed that
beating compared to those who condone physical sexual activity between adults and children was
punishment of children (Table 8.2). This holds true never acceptable.27
for countries in all regions. In Afghanistan, for
instance, the proportion of girls and women who Closely related to ideas of what constitutes child
justify wife-beating under certain circumstances sexual abuse are opinions on who is most likely to
is twice the proportion of respondents of both commit it. A survey of parents of children under
sexes (aged 15 and above) who support the use of age 15 in an urban community in southwest Nigeria
corporal punishment. The opposite is found in 18 revealed that around 70 per cent of respondents
countries concentrated largely in CEE/CIS and Latin believed that child sexual abuse was most often
America and the Caribbean. In 6 of the 11 CEE/ committed by someone known to the child.28 In
CIS countries and all 7 countries in Latin America contrast, slightly more than half of parents of third-
and the Caribbean with available data, a larger grade students in central China held the opposite
percentage of adults think that physical punishment view.29 Interestingly, in the same study in China,
is needed to educate children than the percentage around one third of respondents said they did not
who think that a husband is justified in hitting or think that females could sexually abuse children;
beating his wife under certain circumstances. Equal slightly more than half (53 per cent) of parents in
proportions of respondents justify wife-beating the Nigerian study agreed with a similar statement.
and corporal punishment in Egypt, Mauritania and
Suriname. The nature of the Eastern Caribbean study allows
for a more substantial exploration of attitudes and

Attitudes towards child perceptions around the issue of family incest. Table
sexual abuse 8.3 presents responses to some of the individual
In attempting to develop holistic strategies to items included in the survey questionnaire. Around
address the sexual abuse of children, it is important 25 per cent of respondents felt that sex between
to identify and consider social norms and shared adults and children is considered ‘normal’ in some
beliefs around the issue. Attitudes towards child families, while another 20 per cent were unsure.
sexual abuse are one indicator of the level of Similarly, 22 per cent of respondents said that sex
knowledge commonly held about the issue and between siblings is considered ‘normal’ in some
the public’s willingness and motivation to confront families, with about 17 per cent being unsure. The
it. Perceptions of how ‘childhood’ should be responses to the latter question showed significant
interpreted and what constitutes sexual abuse are differences according to the educational level of
likely to vary at the individual, societal and cultural the respondent. Women were also more likely than
levels. No large-scale, internationally comparable men to agree that incest is considered ‘normal’
data on attitudes towards child sexual abuse in some families (26 per cent and 17 per cent,
currently exist. However, research conducted in a respectively).
160
Table 8.2
Countries are divided on the level of social acceptability of different forms of violence
Percentage of girls and women aged 15 to 49 years and boys and men aged 15 to 49 years who think that a husband/partner is
justified in hitting or beating his wife or partner under certain circumstances and percentage of adults who think that physical pun-
ishment is necessary to raise/educate children
Girls and women aged 15 to 49 Boys and men aged 15 to 49 Adults who think physical pun-
years who justify wife-beating years who justify wife-beating ishment of children is necessary
Afghanistan 90 - 41
Albania 30 36 13
Algeria 68 - 15
Argentina 2 - 4
Armenia 9 20 3
Barbados 3 - 36
Belarus 4 4 8
Belize 9 - 26
Bosnia and Herzegovina 5 6 14
Central African Republic 80 75 31
Chad 62 - 38
Congo 73 62 22
Costa Rica 4 - 12
Democratic Republic of the Congo 76 - 32
Egypt 39 - 42
Gambia 75 - 39
Georgia 7 - 13
Ghana 44 26 50
Haiti 17 15 30
Iraq 51 - 22
Jordan 70 - 23
Kazakhstan 12 17 7
Lao People's Democratic Republic 58 49 42
Lebanon 10 - 24
Liberia 59 30 61
Mauritania 38 - 36
Mongolia 10 9 16
Montenegro 11 - 5

Niger 60 27 45
Republic of Moldova 11 13 15
Saint Lucia 7 - 21
Serbia 3 7 7
Suriname 13 - 13
Swaziland 28 23 82
The former Yugoslav Republic of Macedonia 15 - 3
Togo 43 - 35
Trinidad and Tobago 8 - 25
Tunisia 30 - 44
Ukraine 3 9 11
Vanuatu 60 60 51
Viet Nam 36 - 17
Notes: Only those countries for which data were available on both indicators from the same most recently available source were included in the figure. Data on attitudes towards wife-beating among
boys and men are only available for a selection of countries. For Argentina, the sample was national and urban (municipalities with a population of more than 5,000), since the country’s rural population
is scattered and accounts for less than 10 per cent of the total. Data on the justification of wife-beating for the Congo differ from the standard definition. Data for Egypt refer to ever-married girls and
women only. Data for Jordan refer to ever-married girls and women only; data on the justification of wife-beating differ from the standard definition. Data for Lebanon refer to currently married girls and
women only. Data for Serbia refer to boys and men aged 15 to 29 years.
161
Table 8.3
Percentage of women and men who agreed, disagreed or were unsure about selected statements regarding child sexual abuse in
the Eastern Caribbean
Agree Disagree Not sure Missing
Sexual activity between adults and children is never okay, no matter what 76 17 2 5
In some families, sex between adults and children is considered normal 25 50 20 4
In some families, sex between brothers and sisters is considered normal 22 57 17 4
Families in which sex between children and adults has occurred for generations
6 84 7 3
without any apparent harm is okay, because it has become a family pattern
Some men who have sexual activity with children have
53 17 25 5
had the same thing done to them as children
Men’s negative attitudes towards women is a reason for child sexual abuse 14 60 27 0
Girls draw men’s sexual attention by the way they dress 77 13 5 5
As it is a man’s role to provide for his children, it should be up to him
13 75 8 4
to decide when it’s okay for his children to be involved in sex
It is okay for mothers who need money to support their family to
5 90 5
allow their daughters to have sex with adults for money
Women sometimes turn a blind eye when their partners
70 14 11 5
have sex with children in their families
Children don’t tell because they are scared of the consequences 82 8 5 5
Children don’t tell because they don’t mind that this happened to them 9 80 11
Adults who have sex with children cause long-term emotional harm to children 85 8 3 4
Adults who have sex with children can be helped to change 70 12 14 4
Children should be helped to speak out if an adult has taken sexual advantage of them 93 1 1 6
Police officers generally know how to deal with unlawful sex between adults and children 22 48 23 8
Any incident or suspicion of an adult taking sexual advantage of
82 8 4 6
a child should always be reported to the authorities
Note: Due to rounding, figures might not add up to 100.

Source: Jones, A. D., and E. Trotman Jemmott, Child Sexual Abuse in the Eastern Caribbean, United Nations Children’s Fund, Action for Children and University of Huddersfield, Huddersfield, 2009.
Historically, intergenerational violence has been a who sexually abuse children had been sexually
subject of deep concern among many development abused themselves, while 25 per cent were unsure.
practitioners. To explore this issue, participants in
the Eastern Caribbean study were asked whether In attempting to understand some of the factors that
they agreed with the following statement: may contribute to the perpetuation of child sexual
“Families in which sex between children and adults abuse, respondents in the Eastern Caribbean study
has occurred for generations without any apparent were asked whether they felt that men’s negative
harm is okay because it has become a family attitude towards women was one of the reasons
pattern.” Data revealed that most respondents (84 behind child sexual abuse. Most respondents either
per cent) disagreed with this statement. However, disagreed (60 per cent) or were unsure (27 per
another question that aimed to elicit opinions cent). However, 77 per cent of respondents said
on the cycle of abuse revealed that a majority of that the way girls dress draws sexual attention from
respondents (53 per cent) thought that some men men. Casting some blame on the victim appears to
162
© UNICEF/NYHQ2010-1336/Estey
be a common response in the Eastern Caribbean Regarding the perceived consequences of child
region. However, it has not been found to be the sexual abuse, a majority of respondents (85 per
case everywhere in the world. For instance, most cent) in the Eastern Caribbean considered it to be
Chinese parents (84 per cent) placed the blame emotionally damaging to children in the long term.
for child sexual abuse on perpetrators and not Research in other parts of the world, however,
children.30 Similarly, findings from the parental has found that opinions on this issue diverge.
survey in Nigeria found that the overwhelming For instance, the Norway study of opinions in
majority of respondents (90 per cent) believed that the general population found that respondents
children who are sexually abused are not to blame.31 expressed uncertainty about whether sexual contact
In a study conducted in 2004 among adults aged 18 was damaging to a child or not.34 In Nigeria, 61 per
cent of parents surveyed believed that abuse could
to 64 in Norway, respondents tended to perceive
only have a serious health impact when it involves
adults as being responsible for any sexual contact
intercourse.35
with children.32
In exploring perceptions of the potential rehabilitation

In terms of the reasons why sexual abuse of children
of perpetrators of child sexual abuse, 70 per cent
remains largely hidden and unreported, 82 per cent
of respondents in the Eastern Caribbean believed
of respondents in the Eastern Caribbean said that
that people could be helped to change. On the other
children did not tell for fear of the consequences.
hand, nearly all parents (91 per cent) in the China
Far fewer respondents (9 per cent) said children do study said they thought a person who had sexually
not report sexual abuse because they do not mind abused a child was likely to repeat the offense.36
that it has happened to them. Close to three in four
(73 per cent) parents of young children in central Finally, almost all (93 per cent) of the adults surveyed
China were of the opinion that children who have in the Eastern Caribbean felt that children should be
been sexually abused are often intimidated by the empowered to speak out if they have been sexually
perpetrator not to disclose the abuse.33 Participants abused. Most participants (82 per cent) also felt that
in the Eastern Caribbean study were also asked any incident in which a child is taken advantage of
whether they thought women sometimes know sexually should always be reported to authorities,
that their partners are having sex with children in although nearly half (48 per cent) did not think that
their family but do not report it, and the majority of police officers know how to deal with unlawful sex
respondents (70 per cent) believed this to be true. between minors and adults.
163
Chapter 9 Key
findings
and
CONCLUSIONS
All children have the right to protection from
violence, regardless of the nature or severity of
the act: a slap by a parent, emotional humiliation
inflicted by a peer, the unwanted sexual advances
of a boyfriend, physical assault by a stranger. All are
forms of violence that can cause harm to children,
reduce their sense of self-worth, affront their dignity
and hinder their development.
Research has shown that violence is detrimental

to all aspects of a child’s growth, including
physical, psychological and social development and
functioning, with sometimes lifelong repercussions.
While repeated exposure to moderate or severe
acts of violence can alter brain development and
compromise a child’s potential, other forms of
violence have more subtle effects that are difficult
to measure. Beyond the unnecessary hurt and
damage such violence inflicts is the fact that it
becomes a learned and accepted behaviour that
is often replicated and passed on to successive
generations. Children who experience or witness
violence are more likely to develop aggressive
tendencies and become perpetrators themselves,
continuing the cycle of violence.
One of the reasons violence against children is

so difficult to confront is because it encompasses
such a broad range of actions and is so widely
practised. Corporal punishment, for example, is

used by parents and other caregivers everywhere,

on a regular basis. Yet few recognize its potential
harm, its degrading nature or its ineffectiveness as
a parenting method. Research shows that repeated
exposure to violent discipline, as with other forms
of violence, can have long-term consequences,
including cognitive and physical impairments, lower
164
self-esteem and educational achievement, and life around the world. Interpersonal violence takes
increased risk for delinquency, substance use and many forms – physical, sexual and emotional –
adult criminal behaviour. Research also suggests and occurs in many settings, including the home,
that non-violent forms of discipline are suitable even school, community and over the Internet. Similarly,
with young children and can be used effectively a wide range of perpetrators commit violence
to achieve desired behaviours and teach children against children, such as family members, intimate
lifelong skills. partners, teachers, neighbours, strangers and other
children. Such violence not only inflicts harm, pain
Ensuring that all forms of violence are recognized as and humiliation on children; it also kills.
a fundamental violation of children’s human rights is
a first step in moving towards their elimination. Homicide
In 2012 alone, almost 95,000 children and
Challenges, however, abound, including the fact that
adolescents under age 20 were victims of homicide,
violence against children remains largely hidden. It
making it a leading cause of preventable injury and
often goes unrecognized due to social acceptance,
death among children. The vast majority of victims
shame, fear of reprisal and the voicelessness of
(85,000) lived in low- and middle-income countries.
children, among many other reasons. One of the From 0 to 9 years of age, 85 per cent of deaths are
keys to uncovering its covert nature and eroding its the result of communicable and non-communicable
acceptance is hard evidence about its prevalence diseases, with little differentiation by sex. As children
and impact. Over the last two decades, the quality enter the second decade of their lives, however, the
and quantity of data on the subject has grown share of deaths due to intentional injuries, including
considerably. Nevertheless, research remains patchy homicide, becomes greater, particularly among boys.
both in coverage and scope – largely due to a range
of methodological and ethical issues along with a Globally, Latin America and the Caribbean has the
lack of systematic investments in the generation of largest share of homicide victims under age 20
comprehensive statistics. (25,400). West and Central Africa has the second
largest share (23,400), followed by Eastern and
Despite countless gaps in the current knowledge Southern Africa (15,000). The lowest number of
base, this report is testimony to the improvements homicides in this age group is found in Central
in data generation that have been made in recent and Eastern Europe and the Commonwealth of
years. It is the largest compilation to date of statistics Independent States (CEE/CIS), followed by the
on violence against children, drawing on data from Middle East and North Africa. The three countries
190 countries. By examining global patterns of with the highest homicide rates in the world among
violence against children as well as attitudes and children and adolescents under age 20 are El
social norms, it sheds light on an issue that has Salvador, Guatemala and the Bolivarian Republic of
CHAPTER 9: Key findings and CONCLUSIONS

remained largely undocumented. Its objective Venezuela. In each of these countries, homicide is
is to use data to make violence against children the leading cause of death among adolescent boys.
and its many ramifications more visible, bringing Nigeria has the largest number of young homicide
about a fuller understanding of its magnitude and victims, with almost 13,000 deaths in 2012, followed
nature and offering clues to its prevention. While by Brazil with approximately 11,000.
intensified efforts are needed to strengthen the
availability of reliable and comprehensive statistics Violent discipline
on the issue, the findings presented here are a clear
Though staggering in themselves, the above
call for action.
statistics represent an extreme form of violence
that is relatively uncommon in the lives of children.
Key findings The data show that far more children experience
This report provides evidence that violence is ever- violence in the form of discipline – usually in their
present in the lives of children from all walks of own homes and from a very young age. On average,
165
about 6 in 10 children worldwide (almost 1 billion) Uruguay to over 50 per cent in Botswana, Djibouti,
between the ages of 2 and 14 are subjected to Egypt, Ghana, the United Republic of Tanzania (Dar
physical (corporal) punishment by their caregivers es Salaam) and Yemen. Worldwide, more than one
on a regular basis. For the most part, children are in three students between the ages of 13 and 15
exposed to a combination of physical punishment experience bullying on a regular basis. Among 106
and psychological aggression. The most severe countries with comparable data on adolescents who
forms of corporal punishment – hitting a child on were recently bullied, rates range from 7 per cent
the head, ears or face or hitting a child hard and in Tajikistan to 74 per cent in Samoa. On the flip
repeatedly – are less common overall: On average, side, nearly a third (31 per cent) of teens in Europe
about 17 per cent of children in 58 countries and North America admitted to bullying others, with
experience these harsh practices. In 23 countries, prevalence ranging from around one in seven (14
severe physical punishment is widespread, with per cent) in the Czech Republic and Sweden to
more than one in five children affected.
nearly 6 in 10 (59 per cent) in Latvia and Romania.
Attitudes towards corporal punishment The experience of violence continues into late
of children adolescence. Among girls aged 15 to 19 worldwide,
The large share of children who are exposed to almost one quarter (around 70 million) said they
violent discipline appears to contradict the findings were the victims of some form of physical violence
on attitudes towards it: Only about 3 in 10 adults since age 15. In Eastern and Southern Africa, at
worldwide believe that physical punishment is least 12 per cent of girls in this age group reported
necessary to properly raise or educate a child. incidents of physical violence in the last year in all
In fact, in all countries but one – Swaziland – the nine countries for which data are available except
percentage of adult respondents who think physical Comoros. In West and Central Africa, the proportion
punishment is necessary is consistently lower than is at least 1 in 14 girls in each of the 11 countries
the percentage of children aged 2 to 14 who are with available data; the share reaches as high as
subjected to violent discipline. In most countries, 26 per cent in Cameroon and 42 per cent in the
adults with no or little education are more likely Democratic Republic of the Congo.
to find physical punishment a necessity in raising
children than their more educated peers. In Yemen, Never-married girls are most likely to report
for instance, 51 per cent of mothers or primary physical violence at the hands of family members,
caregivers with no formal education feel it is friends or acquaintances and teachers. However,
necessary to use physical punishment to educate a among ever-married girls, current and/or former
child, compared to 21 per cent of mothers or primary intimate partners are the most commonly reported
caregivers with a secondary or higher education. perpetrators of physical violence in all the countries
Adults with low economic status are also found to with available data. In India, Mozambique, Nepal,
be more supportive of corporal punishment than

Pakistan, the United Republic of Tanzania and
their wealthier counterparts in about three quarters
Zambia, for instance, over 70 per cent of girls
of countries with available data.
named their current or former husbands or partners
as the perpetrators of physical violence against
Violence among peers and by intimate partners them. Indeed, intimate partner violence is the most
As children grow up, they become more vulnerable common form of gender-based violence against
to other forms of aggression, including violence girls. Globally, nearly one in three adolescent girls
inflicted by their peers and intimate partners. Physical aged 15 to 19 (84 million) in formal unions have
attacks among students are common, especially been the victims of emotional, physical and/or
among boys. In 25 countries with comparable data, sexual violence perpetrated by their husbands or
the prevalence of physical attacks against students partners. Rates of partner violence are particularly
aged 13 to 15 ranges from around 20 per cent in high in sub-Saharan Africa, South Asia, and Latin
the former Yugoslav Republic of Macedonia and America and the Caribbean.
166
Data regarding adolescent boys’ exposure to lesser extent than girls, according to data from four
physical violence, including by intimate partners, are countries. In Uganda, adolescent boys are nearly
much more limited. However, at least one in four two times less likely than their female counterparts
adolescent boys aged 15 to 19 said they experienced to report incidents of forced intercourse or other
physical violence since age 15 in each of five low- forced sexual acts. Similarly, in Mozambique, a
and middle-income countries with comparable data; much smaller proportion of adolescent boys than
prevalence exceeds 40 per cent in all but two of girls said they were the victims of sexual violence
these countries (Ghana and Mozambique). The (3 per cent versus 9 per cent, respectively). As with
most commonly reported perpetrators vary across girls, incidents of sexual violence among boys most
the five countries and include family members, often occur for the first time between the ages of 15
friends or acquaintances, and teachers. and 19 and the most commonly named perpetrators
are current or former intimate partners.
Sexual violence
Adolescents living in high-income countries are
Around 120 million girls worldwide (slightly more also at risk of sexual violence. In Switzerland, for
than 1 in 10) have experienced forced intercourse or instance, a 2009 national survey of girls and boys
other forced sexual acts at some point in their lives. aged 15 to 17 found that 22 per cent and 8 per cent,
However, girls living in certain parts of the world respectively, experienced at least one incident of
seem to be at greater risk than others. Prevalence sexual violence involving physical contact in their
rates of 10 per cent or more for forced sex are found lifetimes. The most common form of sexual violence
in 13 of the 18 countries in sub-Saharan Africa with for both sexes in that country is cyber-victimization.
available data. In contrast, in all countries of CEE/ In the United States, the second National Survey
CIS with comparable data (except the Republic of of Children’s Exposure to Violence (NatSCEV II),
Moldova), less than 1 per cent of adolescent girls carried out in 2011, recorded lifetime rates of sexual
reported instances of sexual violence. victimization among girls and boys aged 14 to 17 at
35 per cent and 20 per cent, respectively.
The age at which sexual violence first occurs has
also been measured. In all but 3 of 21 countries with
comparable data, most adolescent girls said they Reporting incidents of violence
were sexually victimized for the first time between Regardless of the type of violence experienced or
the ages of 15 and 19. However, a substantial the circumstances surrounding it, most victims keep
share experienced sexual violence for the first their abuse secret and never seek help. The data
time at younger ages. In all 21 countries except presented in this report confirm that nearly half of
India, Liberia, the Republic of Moldova, Sao Tome all adolescent girls aged 15 to 19 who reported ever
and Principe, and Zimbabwe, at least one in five having experienced physical and/or sexual violence
girls who reported at least one incident of sexual said they never told anyone about it. In Jordan,

violence said it occurred for the first time between Kyrgyzstan, Nepal, Nigeria, Pakistan, Tajikistan and
the ages of 10 and 14. Uganda, the percentage rises to more than half.
Female victims of sexual violence only are least
By far the most common perpetrators of sexual likely to disclose the abuse, as opposed to those
violence against girls are current or former who experienced either physical violence only or
husbands, partners or boyfriends. A significant both physical and sexual violence. A comparison of
share of girls in the Plurinational State of Bolivia, data from countries with information on both sexes
the Dominican Republic, Guatemala, Kenya, the indicates that boys are about as likely as girls to
Republic of Moldova, the United Republic of keep their victimization secret.
Tanzania and Uganda also reported being victimized
by a friend or acquaintance. Among adolescent girls aged 15 to 19 who have
ever been victims of physical and/or sexual violence,
Boys experience sexual violence too, but to a far about 7 in 10 said they never sought help to end
167
it. While the reasons vary, many girls said they did reason among both female and male respondents.
not realize that what they experienced was a form
of violence or did not see the abuse as a problem. Attitudes towards child sexual abuse
Limited data show that, in some countries, boys
are even less likely than girls to seek help. And, like Attitudes concerning the sexual abuse of children
girls, most boys remain silent about their experience have also been explored. While no large-scale,
of violence, regardless of the type, because they internationally comparable data on this issue
do not view it as a problem. When female victims currently exist, research conducted in a few
do seek help, most look to their own families for regions and countries illustrates the variations
and complexity in commonly held beliefs about
assistance. They are much more likely to turn to
this particular form of violence against children.
individuals they know personally for support rather
For example, a study carried out in six countries
than to institutions such as the police department,
in the Eastern Caribbean in 2008-2009 analysed
medical centres, legal aid establishments, religious
the perceptions of women and men aged 18 and
groups and/or social services, even when they
older about child sexual abuse. In attempting
know that help can be found there.
to understand some of the factors that may
contribute to this form of violence, the study asked
Attitudes towards wife-beating respondents whether they felt that men’s negative
Exploring attitudes and social norms related to attitude towards women was one of the causes.
violence can provide insights into why it occurs, Most respondents either disagreed (60 per cent) or
persists and goes unreported. The evidence in this were unsure (27 per cent). However, 77 per cent of
report suggests that close to half of all girls aged 15 respondents said that the way a girl dresses draws
to 19 worldwide (about 126 million) think a husband sexual attention from men.
or partner is sometimes justified in hitting or beating
his wife (or partner). In sub-Saharan Africa and In terms of the perceived consequences of child
the Middle East and North Africa, this proportion sexual abuse, a majority of respondents in the
rises to more than half. In CEE/CIS, it drops to 28 Eastern Caribbean (85 per cent) considered it to be
per cent. emotionally damaging to children in the long term.
Research in other parts of the world, however,
Supportive attitudes towards wife-beating are reveals a range of opinions. In Norway, for instance,
also widespread among adolescent boys. In both a study among the general population found that
Eastern and Southern Africa and South Asia, respondents expressed uncertainty about whether
close to 50 per cent of boys aged 15 to 19 think sexual contact was damaging to a child or not. In
a husband is justified in hitting his wife under Nigeria, 61 per cent of parents surveyed believed
certain circumstances; in West and Central Africa, that abuse could only have a serious health impact
the share is slightly more than one third. Perhaps when it involves intercourse.
surprisingly, however, in 28 of 60 countries with

data on both sexes, a larger proportion of girls than In our hands
boys believe wife-beating is sometimes justified; in Violence against children is widespread but
14 of these countries, the gender gap exceeds 10 not inevitable. Bringing it to an end is a shared
percentage points. This pattern is found more often responsibility.
in South Asia, sub-Saharan Africa, and East Asia and
the Pacific than in other regions. While often regarded as an individual problem,
violence against children is, in fact, a societal
While males and females may differ with respect to problem, driven by economic and social inequities
the extent of their support for wife-beating, opinions and poor education standards. It is fuelled by social
seem to converge on the main circumstances norms that condone violence as an acceptable
under which this form of violence can be justified: way to resolve conflicts, sanction adult domination
Neglecting the children is the most commonly cited over children and encourage discrimination. It is
168
enabled by systems that lack adequate policies and One of the limitations inherent in any attempt to
legislation, effective governance and a strong rule of document violence against children is what it leaves
law to prevent violence, investigate and prosecute out: the presumably large numbers of children
perpetrators, and provide follow-up services and unable or unwilling to report their experiences.
treatment for victims. And it is allowed to persist While this report also suffers from that constraint,
when it is undocumented and unmeasured as a the story it does tell is motivation enough to spur
result of inadequate investments in data collection
action. For if even one child is harmed through an
and poor dissemination of findings.
act of violence, it is one child too many.
The evidence provided in this report clearly shows

The process of understanding and addressing
that too many children do not receive adequate
violence against children will continue to be
protection from violence. Most violence against
children occurs at the hands of the people charged fraught with difficulties. Nevertheless, as additional
with their care or with whom they interact daily – strategies to end violence are formulated and
caregivers, peers and intimate partners. Children carried out (see Box 9.1), it is also clear that
are also frequently deprived of the protection they systematic investments in data generation are vital.
need and deserve from the State. Only 39 countries The evidence that results is essential to monitoring
worldwide protect children legally from all forms of commitments, informing the development of new
corporal punishment, including at home. Moreover, programmes, policies and laws and assessing their
large discrepancies exist between the protections effectiveness. Future research should focus on not
afforded to adult and child victims of violence. For only documenting the prevalence of violence but
example, if an adult is hit by a relative or peer, this also understanding the underlying factors that fuel
is generally regarded as unacceptable behaviour
it and evaluating interventions aimed at preventing
and legal provisions are usually in place to protect
and responding to it. Broad dissemination of data
the victim’s rights. However, if children are violently
in accessible formats will continue to be needed
punished by their parents or other caregivers, such
to raise awareness and to foster the political will
acts are generally seen as inconsequential, and the
same type and level of legal protection is usually not required to develop and implement effective
available. This lack of protection – combined with strategies and action – at all levels of society.
attitudes and social norms that justify certain acts of
violence against children – creates an environment Ending violence against children is in our hands.
in which many forms of violence are considered With reliable data, we will know when this human
normal and treated with impunity. rights imperative is finally achieved.
“A child rights-based approach to child caregiving and protection requires a paradigm

shift towards respecting and promoting the human dignity and the physical and
psychological integrity of children as rights-bearing individuals.”
United Nations Committee on the Rights of the Child, General Comment No. 13 on the
Convention on the Rights of the Child
169
Box 9.1
Strategies to prevent and respond to

violence against children
UNICEF’s approach to preventing children protect themselves and cope programmes; and help in finding
and responding to violence against with violence when it does occur by temporary accommodations, if re-
children is guided by its 2008 Child developing their critical thinking, quired. The success of these services
Protection Strategy, at the heart of building their self-esteem, and teach- depends, among other things, on
which is strengthening legal and ing them to communicate effectively their quality, accessibility and confi-
policy frameworks along with related and to solve problems in a coopera- dentiality. Another key factor is public
services that seek to protect children. tive way. Life-skills education can be awareness of such services, which
The strategy also addresses societal incorporated into a variety of formal can be generated through advocacy
factors and social norms that per- and informal activities targeting chil- and mass media campaigns as well
petuate violence, while promoting dren of different ages, ranging from as educational and sports institutions
positive social change. UNICEF has pre-school programmes to recreation and after-school programmes.
embraced six key actions to prevent and sports activities to vocational
and respond to violence: training. Peer-support programmes 5. Implementing laws and policies
can be especially effective, particu- that protect children. Without a sup-
1. Supporting parents, caregivers larly when children exhibit antisocial portive policy and legislative environ-
and families. A significant body of behaviour. ment, programmes addressing vio-
evidence suggests that providing lence against children can never fully
parents with child-rearing strate- 3. Changing attitudes and social succeed. This means that appropriate
gies and techniques, as well as eco- norms that encourage violence and policies and legislation need to be
nomic support, can help mitigate discrimination. Preventing violence not only instituted but also enforced
children’s risk of physical abuse. before it occurs requires a major shift and monitored. On a smaller scale,
This area of action seeks to prevent in what society regards as acceptable schools and other institutions, work-
violence from the outset by promot- behaviour. Evidence suggests that places and sports programmes can
ing positive parent-child interactions, deeply engrained attitudes and social develop codes of conduct that com-
including non-violent discipline. The norms that feed violence and intol- mit to providing a healthy, safe and
approach is to integrate support into erance can shift through school and secure environment for children. A
the everyday lives of families, either community programmes that engage real shift in public attitudes towards
through regular visits at home or a influential, trusted individuals as violence must happen at two levels:
community centre or through other agents of change. Such programmes On the government side, a solid legal
local venues staffed by nurses, social are most effective when they are framework must be instituted, imple-
workers or other trained profession- complemented by mass media and mented and monitored; on the part of
als. Home visits and parent education social mobilization campaigns along each citizen, an effort must be made
classes combine health care, parent- with supportive services that encour- to drive change on a daily basis.
ing education, child abuse prevention age the reporting of abuse and the
and early intervention services for enactment of new laws and policies 6. Carrying out data collection and
infants and toddlers. Programmes for that make forms of violence punish- research. Effective advocacy and
parents of school-aged children seek able offences. programming are grounded in sol-
to promote effective communication id evidence. The last two decades
around issues including sexuality, 4. Providing and promoting support have seen progress in the availabil-
HIV prevention and sexual abuse. services for children. Child-focused ity of statistics on violence against
services and mechanisms are cru- children, primarily through national
2. Helping children and adolescents cial to enabling children and adoles- household surveys aimed at provid-
manage risks and challenges. Pov- cents to cope with violence and break ing information on the magnitude of
erty, difficulty in school, poor self- the cycle of abuse. Services include the problem. Investments in sustain-
worth and self-discipline, and lack of counselling; information and refer- able data collection mechanisms and
information on where to get help can rals for other protection services of- robust research tools are needed to
increase a child’s vulnerability to vio- fered through the police, doctors, continue to expand the knowledge
lence. Life-skills education can help social welfare workers and assistance base on this important issue.
Adapted from: United Nations Children’s Fund, Strategies for Preventing and Responding to Violence against Children, UNICEF, New York, 2014.
170
© UNICEF/MLWB2009-00032/Pirozzi
In 2009, during a play-therapy session, a male counsellor interacts with a young child in a youth club in Lilongwe, Malawi.
This type of therapy involves giving a wide variety of toys to children who have suffered abuse to see how they play
with them. “We don’t guide the children; we just let them play randomly. It is the manner in which the children play and
the toys that they select that is significant for us. It is very difficult to get children to talk about such things. Instead, they
communicate through playing,” says Sergeant Mwasinga.
171
Dominican Republic
Belize
STEPS TO
Eastern Caribbean
03
01 04
ENDING A
GLOBAL
PROBLEM
Plurinational State of Bolivia
02
Uruguay
Highlighted here are
selected efforts to 05
prevent and respond

to violence against
children. These activities,
implemented by national
governments and civil
society organizations LATIN AMERICA AND THE CARIBBEAN
with support from 01 Inpassed

Belize in 2013, the House of Representatives
amendments to the Criminal Code, strengthen-
UNICEF, involve legal ing the protection of child victims of sexual abuse and
increasing penalties for perpetrators.
and policy reforms;
strengthening of 02 Iningthe Plurinational State of Bolivia, start-
in 2013, the Prosecutor’s Office established child-
friendly legal services for child victims of sexual abuse
institutional and and their families. The initiative included the revision
of an institutional ‘road map’ of the justice system and
community-based the development of protocols that outline the roles and
responsibilities of individual institutions in dealing with
capacity to detect, such cases.
investigate, prosecute
03 InthetheNational
Dominican Republic, the Ministry of Tourism,
and refer cases; delivery Association of Hotels and Restaurants,
ECPAT International (a non-governmental organiza-
of comprehensive tion [NGO]/child rights network) and UNICEF signed
an agreement in 2013 to promote a national policy for
services for child the prevention of sexual exploitation, and the tourism
police committed to re-train its entire staff.
victims; and social
mobilization and
04 Inagainst
the Eastern Caribbean, a Break the Silence
Child Sexual Abuse campaign is being rolled
out in all 12 countries of the sub-region to raise aware-
awareness-raising. ness about the issue and the importance of report-
ing suspected cases. The campaign is also providing
information on available services and has mobilized
Notes for all maps on pages 172 to 175: These maps are stylized and not to scale.
They do not reflect a position by UNICEF on the legal status of any country or ter- a wide range of government and other partners to
ritory or the delimitation of any frontiers. The dotted line represents approximate- accelerate action to protect children against sexual
ly the Line of Control in Jammu and Kashmir agreed upon by India and Pakistan.
The final status of Jammu and Kashmir has not yet been agreed upon by the
parties. The final boundary between the Sudan and South Sudan has not yet
been determined. The final status of the Abyei area has not yet been determined.
172
abuse. In accordance with the 2012 Bridgetown

Declaration and Agenda for Action to Combat
07 Inpaign
Côte d’Ivoire, the global End Violence cam-
is being taken nationwide: Country-specific
Child Sexual Abuse in the Caribbean, the following messages developed in five languages are being
countries have developed laws mandating report- broadcast on local radio and child protection
ing: Anguilla, Antigua and Barbuda, Grenada, Saint networks are being expanded. The Ministry of
Kitts and Nevis, Saint Vincent and the Grenadines, Education adopted a strategy to address sexual vio-
Trinidad and Tobago, Turks and Caicos, and the lence against girls in schools and Codes of Conduct
Virgin Islands (UK). Anguilla, Grenada, Saint Vincent were developed for teachers.
and the Grenadines, Trinidad and Tobago, and Turks
and Caicos have also established related reporting 08 In the Democratic Republic of the Congo
protocols, and inter-agency reporting protocols have in 2013, UNICEF and partners assisted 12,247 sur-
been developed in Barbados and Montserrat. The vivors of sexual and other forms of gender-based
Governments of Grenada and Trinidad and Tobago violence, nearly a third of whom were children – the
are launching new national child protection agen- vast majority girls.
cies to better investigate and respond to reported
cases of child abuse, while Barbados is assessing 09 In Sierra Leone, a national referral protocol for
recommendations for modernization of its half- victims of sexual and other forms of gender-based
century-old child protection agency. violence was launched in 2012, defining the respon-
sibilities of formal and informal service providers.
05 Inteachers
Uruguay, more than 500 primary school
were trained to detect, report and provide
assistance to children in cases of sexual abuse and EASTERN AND SOUTHERN AFRICA
other forms of violence. This is one initiative in a
campaign to address violence against children in 10 In Malawi, child victims of sexual and physical
multiple cities across the country. violence are now receiving services through over
100 victim support units within law enforcement,
300 within communities, and additional ‘one-stop’
WEST AND CENTRAL AFRICA centres established with UNICEF support.
06 Inchild
Burkina Faso, action plans, training and local
protection networks are beginning to yield
In Swaziland, the Sexual Offences and Domestic
11 Violence Bill of 2009 was passed by Parliament in
result for victims of violence and trafficking. In 2013, October 2011, following a debate that spanned two
nearly 27,000 victims or children at risk received sessions. The bill extends the definition of rape to
support – well beyond the initial target of 18,000. include male victims and also introduces a register
of sexual offenders.
In 2013, the United Republic of Tanzania

12 developed Codes of Conduct for teachers based
on zero tolerance of sexual violence. This is one
element of a national programming framework to
Sierra Leone address sexual violence, unintended pregnancies
Burkina Faso and HIV among adolescent girls.
06 Côte d’Ivoire
Zimbabwe, a national Girls’ Empowerment
09 Democratic Republic
of the Congo
13 InFramework will be launched in the second half Steps to ending a global problem
07 of 2014 as part of a larger programme to address
08 sexual violence, unintended pregnancies and HIV
United Republic
of Tanzania among adolescent girls. The framework, the first
12 of its kind, outlines how empowerment efforts can
Malawi be maximized when girls are deliberately targeted
before they drop out of school, marry or even enter
into their first romantic relationships. It is anchored
10
Zimbabwe in five areas of intervention – education, economic
empowerment, safety and protection, reproductive
Swaziland health, and leadership and decision-making skills –
13 around which strategic objectives, priority actions
and key actors have been identified.
11
173
Montenegro
Serbia Islamic
1415 16 Turkey
Republic
17 of Iran
State of Palestine
Jordan 20
22 21
Gulf Area Sub-regional
Programme
Egypt
18 19
CENTRAL AND EASTERN EUROPE AND THE

COMMONWEALTH OF INDEPENDENT STATES
14 Insocial
Bosnia and Herzegovina, protocols for
workers, teachers, health professionals and MIDDLE EAST AND NORTH AFRICA
police on how to detect, report and refer cases of
violence have been launched at the national level
and tested in 10 municipalities. The first Network
18 In Egypt, a campaign to address social norms
on violence against children, including physical
of NGOs on Violence against Children has also and psychological violence, continued in schools,
been created to call attention to violence against homes and communities in 2013. In one governor-
children among both authorities and the public and ate alone, 118,000 people were reached via initia-
is referring cases to relevant professionals. tives that included media campaigns, activities in
schools, and work with religious leaders.
15 InLawMontenegro, the Government adopted a
on Protection from Family Violence and a cor-
responding strategy and protocol. It also formal-
19 Within the Gulf Area Sub-regional Pro-
gramme, a network of 140 school teachers
ized the work of local multidisciplinary teams for are serving as trainers in early detection of child
the protection of children from violence, abuse abuse, potentially reaching an additional 10,500
and neglect. A related study conducted in 2013 teachers in 2014.
revealed some level of social tolerance for child
sexual abuse, indicating that awareness-raising on
the issue is also urgently needed.
20 In 2013, the Islamic Republic of Iran’s
Ministry of Health scaled up two successful pilot
projects to promote the prevention, early detection
16 Inlence,
Serbia, following a 2012 study on digital vio-
UNICEF and a corporate partner initiated
and management of child maltreatment within the
health system. The project includes communica-
a social media campaign called ‘Choose Words, tions strategies to promote positive parenting and
Prevent Hate’. Youth-led blogs and a Facebook a national broadcasting campaign to raise public
page are reaching more than 11,000 people. awareness of key child protection issues, includ-
UNICEF National Ambassadors have joined the ing violence against children.
campaign by blogging, tweeting and participating
21
Steps to ending a global problem
in selected events. A school-based programme In Jordan in 2013, UNICEF renewed its part-
on preventing digital violence is also under way in nership with the Ministry of Education to expand
10 schools. the Ma’An (Together) campaign to end violence in
schools. Over 4,000 teachers and school counsel-
17 Inistries
Turkey, children, NGOs and relevant min-
all took part in developing a draft Violence
lors were trained in methodologies including class-
room management skills, positive discipline and
against Children National Strategy. In addition, 425 community mobilization. Online tracking systems
professionals from state services were trained are operational in most government schools to
to work together at the local level when interven- encourage the reporting of incidents and monitor
ing in violence against children cases. Through a steps taken in response.
series of 10 workshops, 182 NGO representatives
from 30 cities were also equipped to advocate
for and monitor systematic responses to violence
22 In the State of Palestine, a policy on the use
of non-violent discipline in schools was launched
against children. by the Ministry of Education in September 2013.
174
Mongolia
27
China
India Bhutan 26 conducting home visits. A services directory will
also be completed at the local level to facilitate
23 referrals. ACWF for the first time publicly recog-
24 nized its mandate to prevent and protect children
from violence and abuse.
Sri Lanka Viet Nam Philippines 27 Inrevision

2014, the Government of Mongolia finalized
of a law on Combating Domestic Violence
29 that incorporates clear provisions for the protection
28 of children at home. In addition, the Government
25
is developing a stand-alone law on child protec-
tion to address procedural gaps in existing laws to
guarantee children are protected in all settings. In
May 2014, Mongolia launched its first nationwide,
toll-free, 24/7 Child Helpline, which is a child’s first
point of contact with child protection services.
Children only need to dial 108 on any phone, be it
in a remote herder community or other area out-
SOUTH ASIA side existing formal systems, to get linked to an
operator standing by to offer assistance.
23 Innewtheearly
most remote communities of Bhutan, 31
childhood care and development cen-
tres were built in 2013. The education curriculum 28 Inof 2012 in the Philippines, the Department
Education adopted ‘Guidelines on Protecting
on parenting has also been strengthened to ensure Children in School from Abuse, Violence, Exploi-
holistic, responsive care, including the prevention tation, Discrimination, Bullying and other Forms of
of abuse, violence and neglect. Abuse’, which reiterates a zero tolerance policy for
acts of violence in schools. The policy was devel-
24 Incalled
2013, India launched a social media campaign
‘Time to Sound the Red Siren’. It addresses
oped and issued following a major advocacy effort
and the completion of a regional study of six East
sexual violence against children, particularly girls, Asian and Pacific countries on existing systems for
by challenging social norms that sanction such vio- child protection in educational settings.
lence and encouraging the reporting of suspected
cases.
29 The Government of Viet Nam has introduced a
National Plan on Child Protection (2011-2015) to
25 Inworking
2013 in Sri Lanka, 345 senior police officers
in the criminal justice system received
address violence against children. With support
from UNICEF, the Government then developed a
training on their role in processing cases of vio- national project to respond to cases of neglect,
lence against women and children, through a child sexual abuse and violence, child labour, and
continued partnership between UNICEF and the juveniles in conflict with the law. To support imple-
Government. UNICEF also partnered with Police mentation of this country-wide effort, Community-
Training Colleges in three districts to conduct based Child Protection Systems were established
trainings for an additional 856 police officers on in eight provinces in 2011 and are piloting special-
aspects of gender-based violence and the protec- ized services for children. The protection systems
tion of child victims. involve coordinated efforts on the part of local
Steps to ending a global problem
authorities, relevant agencies, social organizations,
NGOs, communities and families to both prevent
EAST ASIA AND THE PACIFIC violence and protect children. The service mod-
els being demonstrated include alternative care
26 InCommunity
China, a Violence against Children Hotline and
Response Toolkit was launched in
for children without family support; child-friendly
interview rooms; community-based diversion for
December 2013 through the All-China Women’s juveniles in conflict with the law; and Social Work
Federation (ACWF). The toolkit consists of two Service Centres, which provide psycho-social
sets of guidelines: one outlines the roles and care, legal protection and support to child victims
responsibilities of ACWF staff running the com- of violence and abuse as well as other vulnerable
munity hotline to respond effectively to cases of children. In recognition of the positive impact of
family violence, and the other is aimed at building the new initiative, the Government expanded its
basic social work skills for women’s cadres when operation to cover 40 of the country’s 63 provinces.
175
references
Chapter 1 achievement, NCAC, Huntsville, 2013.
11 Macmillan, R., and J. Hagan, ‘Violence in the Transition to
1 Krug, E.G., et al., editors, World Report on Violence and Adulthood: Adolescent victimization, education, and socio-
Health, World Health Organization, Geneva, 2002. economic attainment in later life’, Journal of Research on
2 See: United Nations Children’s Fund, Female Genital Muti- Adolescence, vol. 14, no. 2, 2004, pp. 127-158; Paolucci, E.
lation/Cutting: A statistical overview and exploration of the O., M. L. Genuis and C. Violato, ‘A Meta-analysis of the Pub-
dynamics of change, UNICEF, New York, 2013. lished Research on the Effects of Child Sexual Abuse’, The
Journal of Psychology, vol. 135, no. 1, 2001, pp. 17-36; Velt-
3 Krug et al., op. cit. man, M. W. M., and K. D. Browne, ‘Three Decades of Child
4 Additional data sources used throughout the report include a Maltreatment Research’, Trauma Violence & Abuse, vol. 2,
survey conducted in Swaziland in 2007 by the United States no. 3, 2001, pp. 215-239; Gilbert, R., et al., ‘Burden and Con-
Centers for Disease Control and Prevention (CDC) and UNI- sequences of Child Maltreatment in High-income Countries’,
CEF, in collaboration with the Government of Swaziland, and The Lancet, vol. 373, no. 9657, 2009, pp. 68-81; Slade, E.
the surveys conducted under the Together for Girls initia- P., and L. S. Wissow, ‘The Influence of Childhood Maltreat-
tive. Together for Girls is a global partnership of international ment on Adolescents’ Academic Performance’, Econom-
organizations from the public, private and non-profit sectors ics of Education Review, vol. 26, no. 5, 2007, pp. 604-614;
focused on ending violence against children. Building on the Leiter, J., ‘School Performance Trajectories After the Advent
experience in Swaziland, it brings together 10 public and pri- of Reported Maltreatment’, Children and Youth Services
vate sector organizations including the United States Depart- Review, vol. 29, no. 3, 2007, pp. 363-382.
ment of State-President’s Emergency Plan for AIDS Relief,
12 Widom, C. S., ‘Childhood Victimization: Early adversity, later
the Office of Global Women’s Health Issues, CDC, UNICEF,
psychopathology’, National Institute of Justice Journal, no.
the United Nations Population Fund, the Joint United Nations
242, 2000, pp. 3-9.
Programme on HIV/AIDS, the United Nations Development
Fund For Women (UNIFEM, now merged into UN Women), 13 Duncan, R. D., ‘Childhood Maltreatment and College Drop-
the World Health Organization, BD (Becton, Dickinson and Out Rates: Implications for child abuse researchers’, Journal
Company), the CDC Foundation, Grupo ABC and the Nduna of Interpersonal Violence, vol. 15, no. 9, 2000, pp. 987-995.
Foundation. The Together for Girls initiative focuses on three
14 Lansford, J. E., et al., ‘A 12-year Prospective Study of the
core activities: conducting national surveys and collecting
Long-term Effects of Early Child Physical Maltreatment on
data to document the magnitude, nature and effects of vio-
Psychological, Behavioral, and Academic Problems in Adoles-
lence (including sexual violence); informing government lead-
cence’, Archives of Pediatrics and Adolescent Medicine, vol.
ers, communities and donors and supporting a plan of action
156, no. 8, 2002, pp. 824-830.
at the country level to address violence; and building public
awareness to motivate change in societal and gender norms 15 Huang, L., and S. Mossige, ‘Academic Achievement in Nor-
and behaviours that are harmful to children and women. wegian Secondary Schools: The impact of violence during
Under this initiative, six surveys had been completed at the childhood’, Social Psychology of Education, vol. 15, no. 2,
time of this report (in Cambodia, Haiti, Indonesia, Kenya, the 2012, pp. 147-164.
United Republic of Tanzania and Zimbabwe) and others were
in the planning, pilot or implementation stages. Data from 16 Currie, J., and C. S. Widom, ‘Long-term Consequences of
the surveys conducted in Cambodia (2012), Haiti (2012) and Child Abuse and Neglect on Adult Economic Well-being’,
Indonesia (2012) could not be included since there were no Child Maltreatment, vol. 15, no. 2, 2010, pp. 111-120.
published reports or publicly available documentation at the 17 Macmillan and Hagan, op. cit.
time this report was compiled. Information on the initiative
can be found at: <www.togetherforgirls.org>, accessed 15 18 Veltman and Browne, op. cit.; Gilbert et al., op. cit.
May 2014. These surveys are referred to throughout the pub- 19 Johnson-Reid, M., et al., ‘A Prospective Analysis of the Rela-
lication as the Violence Against Children Surveys (VACS). tionship between Reported Child Maltreatment and Special
5 Leiner, M., et al., ‘Children’s Mental Health and Collective Education Eligibility among Poor Children’, Child Maltreat-
Violence: A binational study on the United States-Mexico bor- ment, vol. 9, no. 4, 2004, pp. 382-394.
der’, Pan American Journal of Public Health, vol. 31, no. 5, 20 Fantuzzo, J. W., S. M. Perlman and E. K. Dobbins, ‘Types and
2012, pp. 411-416. Timing of Child Maltreatment and Early School Success: A
6 Evans, E., K. Hawton and K. Rodham, ‘Suicidal Phenomena population-based investigation’, Children and Youth Services
and Abuse in Adolescents: A review of epidemiological stud- Review, vol. 33, no. 8, 2011, pp. 1404-1411; Leiter, op. cit.
ies’, Child Abuse & Neglect, vol. 29, no. 1, 2005, pp. 45-58. 21 Lansford et al., op. cit.
7 Dardis, C. M., et al., ‘An Examination of the Factors Related to 22 Pereznieto, P., et al., The Economic Impact of School Vio-
Dating Violence Perpetration among Young Men and Women lence: A report for Plan International, Plan International and
and Associated Theoretical Explanations: A review of the lit- Overseas Development Institute, London, 2010.
erature’, Trauma Violence & Abuse, published online 10 Janu-
ary 2014. 23 Save the Children Denmark, Ministry of Education and Min-
istry of Women’s Affairs, A Study on Violence against Girls in
8 Ellonen, N., et al., ‘Exposure to Parental Violence and Out- Primary Schools and Its Impacts on Girls’ Education in Ethio-
comes of Child Psychosocial Adjustment’, Violence and Vic- pia, Save the Children, Addis Ababa, 2008.
tims, vol. 28, no. 1, 2013, pp. 3-15.
24 Schwartz, D., et al., ‘The Link Between Harsh Home Environ-
9 Literature on the social and economic impact of violence
ments and Negative Academic Trajectories is Exacerbated by
References
remains mostly limited to high-income countries. The review

Victimization in the Elementary School Peer Group’, Develop-
presented here is not meant to be exhaustive but to be illus-
mental Psychology, vol. 49, no. 2, 2013, pp. 305-316; Holt, M.
trative of what is known from selected studies.
K., D. Finkelhor and G. K. Kantor, ‘Multiple Victimization Expe-
10 National Children’s Advocacy Center, As a Matter of Fact: riences of Urban Elementary School Students: Associations
The relationship between child maltreatment and academic with psychosocial functioning and academic performance’,
176
Child Abuse & Neglect, vol. 31, no. 5, 2007, pp. 503-515. Obstetrics and Gynecology, vol. 198, no. 2, 2008, pp. 171.
e1-e6.
25 Holt, Finkelhor and Kantor, op. cit.
41 Taillieu, T. L., and D. A. Brownridge, ‘Violence against Preg-
26 Currie and Widom, op. cit.; Zielinski, D. S., ‘Child Maltreat-
nant Women: Prevalence, patterns, risk factors, theories, and
ment and Adult Socioeconomic Well-being’, Child Abuse &
directions for future research’, Aggression and Violent Behav-
Neglect, vol. 33, no. 10, 2009, pp. 666-678; Macmillan and
ior, vol. 15, no. 1, 2010, pp. 14-35.
Hagan, op. cit.
42 See, for example, Pinheiro, P. S., World Report on Violence
27 It is possible that some of the differences observed in eco-
against Children, UN Secretary-General’s Study on Violence
nomic outcomes may be accounted for by differences in edu-
against Children, United Nations, Geneva, 2006, p. 50; Bou-
cational achievement which are also found to be associated
dreaux, M. C., and W. D. Lord, ‘Combating Child Homicide:
with the experience of violence. See: Currie and Widom, op.
Preventive policing for the new millennium’, Journal of Inter-
cit.
personal Violence, vol. 20, no. 4, 2005, pp. 380-387; Bernard
28 Zielinski, op. cit. Van Leer Foundation, Setting Our Agenda on Early Learning,
Violence and Physical Development: Early childhood matters,
29 Zielinski, op. cit.
Bernard Van Leer Foundation, The Hague, 2010, p. 1.
30 Currie and Widom, op. cit.
43 Dutton-Scott, E., ‘Domestic Violence and its Impact on Chil-
31 Gelles, R. J., and S. Perlman, Estimated Annual Cost of Child dren’s Development’, edited version of presentation deliv-
Abuse and Neglect, Prevent Child Abuse America, Chicago, ered at the Department of Community Services’ Fourth
2012. Domestic Violence Forum, Glebe, 2002, <www.community.
nsw.gov.au/docswr/_assets/main/documents/dv_paper.pdf>,
32 Fang, X., et al., ‘The Economic Burden of Child Maltreatment
accessed 18 June 2014.
in the United States and Implications for Prevention’, Child
Abuse & Neglect, vol. 36, no. 2, 2012, pp. 156-165. 44 Gunnar, M. R., and R. Barr, ‘Stress, Early Brain Development,
and Behaviour’, Infants and Young Children, vol. 11, no. 1,
33 Fang, X., D. Brown and P. Corso, Analytical Report for the 1998, pp. 1-14; Osofsky, J. D., ‘The Effects of Exposure to
Project “Development of Regional Costing Model to Estimate Violence on Young Children’, American Psychologist, vol. 50,
the Economic Burden of Child Maltreatment in the East Asia no. 9, 1995, pp. 782-788; Holt, S., H. Buckley and S. Whelan,
and Pacific Region”, Bangkok, 2013 (unpublished). ‘The Impact of Exposure to Domestic Violence on Children
34 Fang et al., op. cit.; Gelles and Perlman, op. cit.; Currie and and Young People: A review of the literature’, Child Abuse
Widom, op. cit.; Noor, I., R. A. Caldwell and D. Strong, The & Neglect, vol. 32, no. 8, 2008, pp. 797-810; Lupien, S. J., et
Costs of Child Abuse vs. Child Abuse Prevention: A decade al., ‘Effects of Stress Throughout the Lifespan on the Brain,
of Michigan’s experience, Michigan Children’s Trust Fund, Behaviour and Cognition’, Nature Reviews Neuroscience, vol.
Lansing, 2004. 10, no. 6, 2009, pp. 434-445.
35 Vogler, P., G. Crivello and M. Woodhead, Early Childhood Tran- 45 Experiences of extreme stress activate the brain’s stress
sitions Research: A review of concepts, theory and practice, management systems and lead to neurobiological responses
Bernard Van Leer Foundation, The Hague, 2008. that are generally protective under normal circumstances.
However, when the brain’s stress management systems
36 Cunningham, A., and L. Baker, What about Me! Seeking to are activated persistently under experiences of extreme
understand a child’s view of violence in the family, Centre adversity, these responses can become pathogenic. See:
for Children and Families in the Justice System, London, ON, Shonkoff, J. P., W. T. Boyce and B. S. McEwen, ‘Neurosci-
2004. ence, Molecular Biology, and the Childhood Roots of Health
37 Schwarzer, R., and C. Schwarzer, ‘A Critical Survey of Cop- Disparities: Building a new framework for health promotion
ing Instruments’, in Handbook of Coping: Theory, research, and disease prevention’, JAMA, vol. 301, no. 21, 2009, pp.
applications, edited by M. Zeidner and N. S. Endler, Wiley, 2252-2259.
New York, 1996; Pynoos, R. S., ‘Traumatic Stress and Devel- 46 Dietz, T. L., ‘Disciplining Children: Characteristics associ-
opmental Psychopathology in Children and Adolescents’, in ated with the use of corporal punishment and non-violent
Review of Psychiatry, edited by J. Oldham, M. Riba and A. discipline’, Child Abuse & Neglect, vol. 24, no. 12, 2002, pp.
Tasman, American Psychiatric Press, Washington, DC, 1993. 1529-1542.
38 Mulder, E. J. H., et al., ‘Prenatal Maternal Stress: Effects 47 Pereznieto, et al., op. cit.
on pregnancy and the (unborn) child,’ Early Human
Development, vol. 70, no. 1, 2002, pp. 3-14; Quinlivan, J., 48 Mayall, B., and H. Zeiher, editors, Childhood in Generational
‘Study of Adolescent Pregnancy in Western Australia,’ paper Perspective, Institute of Education, University of London,
presented at ‘The Way Forward: Children, young people and London, 2003; Lancy, D. F., The Anthropology of Childhood:
domestic violence’, Melbourne, 2000. Cherubs, chattel, changelings, Cambridge University Press,
Cambridge, 2008; Vogler, Crivello and Woodhead, op. cit.
39 Margolin, G., ‘Effects of Witnessing Violence on Children’, in
Violence against Children in the Family and the Community, 49 United Nations Children’s Fund, Eliminating Discrimination
edited by P. K. Trickett and C. J. Schellenbach, American Psy- against Children and Parents Based on Sexual Orientation
chological Association, Washington, DC, 1998; Murphy, C. C., and/or Gender Identity, UNICEF Position Paper, no. 9, May
2014 (unpublished).
et al., ‘Abuse: A risk factor for low birth weight? A system-
atic review and meta-analysis’, Canadian Medical Association 50 García-Moreno, C. et al., WHO Multi-country Study on Wom-
Journal, vol. 164, no. 11, 2001, pp. 1567-1572; Valladares, E., en’s Health and Domestic Violence against Women, World
References
et al., ‘Physical Partner Abuse During Pregnancy: A risk factor Health Organization, Geneva, 2005; United Nations Children’s
for low birth weight in Nicaragua’, Obstetrics and Gynecol- Fund, A Study on Violence against Girls, UNICEF Innocenti
ogy, vol. 100, no. 4, 2002, pp. 700-705. Research Institute, Florence, 2009.
40 Rodrigues, T., L. Rocha and H. Barros, ‘Physical Abuse dur- 51 Le, M. T. H., et al., ‘Early Marriage and Intimate Partner Vio-
ing Pregnancy and Preterm Delivery’, American Journal of lence Among Adolescents and Young Adults in Viet Nam’,
177
Journal of Interpersonal Violence, vol. 29, no. 5, 2014, pp. 14 Krug et al., op. cit.
889-910; Speizer, I. S., and E. Pearson, ‘Association between
15 United Nations Children’s Fund, Manual for the Measurement
Early Marriage and Intimate Partner Violence in India: A focus
of Indicators of Violence against Children, UNICEF, New York,
on youth from Bihar and Rajasthan’, Journal of Interpersonal
2006 (unpublished).
Violence, vol. 26, no. 10, 2011, pp. 1963-1981.
16 ChildONEurope, op. cit.
52 United Nations Children’s Fund, 2009, op. cit.
17 Krug et al., op. cit.
53 Harris, J. R., The Nurture Assumption: Why children turn out
the way they do, Free Press, New York, 1998. 18 Radford, L., et al., Child Abuse and Neglect in the UK Today,
National Society for the Prevention of Cruelty to Children,
54 Hart, J. Children’s Participation in Humanitarian Action: Learn- London, 2011.
ing from zones of armed conflict, synthesis report prepared
for the Canadian International Development Agency (CIDA), 19 Gilbert, R., et al., ‘Burden and Consequences of Child Mal-
Refugee Studies Centre, Oxford, 2004. treatment in High-income Countries’, The Lancet, vol. 373,
no. 9657, 2009, pp. 68-81.
20 Havercamp, B. E., and R. A. Young, ‘Paradigms, Purpose, and
Chapter 2 the Role of the Literature: Formulating a rationale for qualita-
tive investigations’, The Counseling Psychologist, vol. 35, no.
1 ChildONEurope, Guidelines on Data Collection and Monitor- 2, 2007, pp. 265-294.
ing Systems on Child Abuse, Florence, 2009; Krug, E. G., 21 Sofaer, S., ‘Qualitative Methods: What are they and why use
et al., editors, World Report on Violence and Health, World them?’, Health Services Research, vol. 34, no. 5, 1999, pp.
Health Organization, Geneva, 2002. 1101-1118.
2 ChildONEurope, op. cit. 22 Tyldum, G., ‘Limitations in Research on Human Trafficking’,
3 Smith, K., et al., Using Administrative Records to Evaluate International Migration, vol. 48, no. 5, 2010, pp. 1-13.
the Accuracy of Child Abuse Reports in a National Survey of 23 Mudaly, N., and C. Goddard, ‘The Ethics of Involving Children
Child Abuse and Neglect, presented at the American Associa- Who Have Been Abused in Child Abuse Research’, Interna-
tion for Public Opinion Research Conference, Anaheim, 2007. tional Journal of Children’s Rights, vol. 17, no. 2, 2009, pp.
4 ChildONEurope, op. cit. 261-281.
24 Pope, C., S. Ziebland and N. Mays, ‘Qualitative Research in
Healthcare. Analysing Qualitative Data’, BMJ, vol. 320, no.
6 Krug et al., op. cit; Kruger, J., et al., ‘A Public Health Approach 7227, 2000, pp. 114-116.
to Violence Prevention in South Africa’, in The Dynamics of
25 Moriarty, J., Qualitative Methods Overview, School for Social
Aggression and Violence in South Africa, edited by R. van
Care Research, London, 2011.
Eeden and M. Wentzel, Human Sciences Research Council,
Pretoria, 1998. 26 United Nations Committee on the Rights of the Child, Gen-
eral Comment No. 13 (2011): The right of the child to free-
7 Krug et al., op. cit.; Houry, D., et al., ‘Emergency Department
dom from all forms of violence, UN document CRC/C/GC/13,
Documentation in Cases of Intentional Assault’, Annals of
Office of the High Commissioner for Human Rights, Geneva,
Emergency Medicine, vol. 34, no. 6, 1999, pp. 715-719.
18 April 2011.
8 Krug et al., op. cit. 27 United Nations Committee on the Rights of the Child, op. cit.
9 Double counting can be an issue with judicial data when a 28 Kadir, A., F. Marais and N. Desmond, ‘Community Perceptions
reported crime includes one or several offenders and/or one of the Social Determinants of Child Health in Western Cape,
or more victims. Additionally, a person can commit more than South Africa: Neglect as a major indicator of child health and
one form of violence at a time, complicating the counting of wellness’, Paediatrics and International Child Health, vol. 33,
offences. Source: ChildONEurope, op. cit. no. 4, 2013, pp. 310-321.
10 Fallon, B., et al., ‘Methodological Challenges in Measuring 29 In the 18- to 24-year-old age group, child neglect was mea-
Child Maltreatment’, Child Abuse & Neglect, vol. 34, no. 1, sured by asking participants how often the following events
2010, pp. 70-79. occurred when they were “a young child (say under the age
11 McClain, P. W., et al., ‘Estimates of Fatal Child Abuse of 12)”: (a) parents or carers expected them to do their own
and Neglect, United States, 1979 through 1988’, laundry under the age of 12; (b) child had regular dental check-
Pediatrics, vol. 91, no. 2, 1993, pp. 338-343. ups; (c) child went to school in clothes that were dirty, torn or
that did not fit because there were no clean ones available; (d)
12 Grimm, B., ‘Child Deaths from Abuse and Neglect: Accurate child went hungry because no one got his/her meals ready or
data, public disclosure needed’, Journal of the National Cen- there was no food in the house; (e) child looked after younger
ter for Youth Law, vol. 28, no. 1, 2007, pp. 1-11. brothers or sisters while parents were out; (f) child was ill,
13 Surveys can be conducted using different settings as primary but no one looked after him/her or took him/her to the doc-
sampling units for data collection, including households, schools tor; or (g) child did not have a safe place to stay. Answers of
or other establishments. Household surveys involve gather- ‘often’ or ‘always’ for all questions except dental check-ups
were coded as neglect; a child rarely or never having a dental
ing information from a sample of households that are usually
check-up was coded as neglect. This measure was adopted
selected at random from a predefined list of all the households
from an earlier study in the United Kingdom by the National
in a specific area. School-based surveys usually involve self-
References
Society for the Prevention of Cruelty to Children conducted in

administered questionnaires given to young people in school
1998. Source: Radford et al., op. cit.
settings. Establishment surveys are carried out in places of
employment, for instance, and can inquire about circum- 30 In this study, a measure of child neglect from the second
stances and behaviours related to the establishment used as revision of the Juvenile Victimization Questionnaire (JVQ-
the sample unit as well as about experiences in other settings. R2) was used. Participating children and youth were asked:
178
“When someone is neglected, it means that the grown-ups 51 Ybarra, M.L., et al., ‘Impact of Asking Sensitive Questions
in their life didn’t take care of them the way they should. They about Violence to Children and Adolescents’, Journal of Ado-
might not get them enough food, take them to the doctor lescent Health, vol. 45, no. 5, 2009, pp. 499-507; Seedat, S.,
when they are sick, or make sure they have a safe place to et al., ‘Ethics of Research on Survivors of Trauma’, Current
stay. At any time in your life, did you get neglected?” Source: Psychiatry Reports, vol. 6, no. 4, 2004, pp. 262-267.
Finkelhor, D., et al., ‘Violence, Crime, and Abuse Exposure in
52 Child Protection Monitoring and Evaluation Reference Group,
a National Sample of Children and Youth: An update’, JAMA
Ethical Principles, Dilemmas and Risks in Collecting Data on
Pediatrics, vol. 167, no. 7, 2013, pp. 614-621.
Violence against Children: A review of available literature, Sta-
31 Public Health Agency of Canada, Canadian Incidence Study tistics and Monitoring Section, Division of Policy and Strategy,
of Reported Child Abuse and Neglect: Major findings, Public United Nations Children’s Fund, New York, 2012.
Health Agency of Canada, Ottawa, 2010.
53 Langhinrichsen-Rohling, J., et al., ‘Sensitive Research with
32 Dubowitz, H., S. C. Pitts and M. M. Black, ‘Measurement of Adolescents: Just how upsetting are self-report surveys any-
Three Major Subtypes of Child Neglect’, Child Maltreatment, way?’, Violence and Victims, vol. 21, no. 4, 2006, pp. 425-444.
vol. 9, no. 4, 2004, pp. 344-356.
33 Straus, M. A., and G. K. Kantor, ‘Definition and Measurement op. cit.
of Neglectful Behavior: Some principles and guidelines’, Child
55 Ybarra et al., op. cit.
Abuse & Neglect, vol. 29, no. 1, 2005, pp. 19-29.
56 Langhinrichsen-Rohling et al., op. cit.
34 Straus and Kantor, op. cit.
57 Dyregrov, K., A. Dyregrov and M. Raundalen, ‘Refugee Fami-
35 DePanfilis, D., Child Neglect: A guide for prevention,
lies’ Experience of Research Participation’, Journal of Trau-
assessment, and intervention, United States Department of
matic Stress, vol. 13, no. 3, 2000, pp. 413-426.
Health and Human Services Administration for Children and
Families, Office on Child Abuse and Neglect, Washington, 58 Ullman, S., ‘Asking Research Participants about Trauma and
DC, 2006. Abuse’, American Psychologist, vol. 62, no. 4, 2007, pp.
329-330.
36 Slack, K. S., et al., ‘Improving the Measurement of Child
Neglect for Survey Research: Issues and recommendations’, 59 Child Protection Monitoring and Evaluation Reference Group,
Child Maltreatment, vol. 8, no. 2, 2003, pp. 98-111. op. cit.
37 Slack et al., op. cit. 60 Child Protection Monitoring and Evaluation Reference Group,
op. cit.; Gallagher, M., ‘Ethics’, in Researching with Children
38 Dubowitz, Pitts and Black, op. cit.
and Young People: Research design, method and analysis,
39 Slack et al., op. cit. edited by E. K. M. Tisdall, J. M. Davis and M. Gallagher, Sage
Publications, London, 2009.
40 Kantor, G. K., et al., ‘Development and Preliminary Psycho-
metric Properties of the Multidimensional Neglectful Behav- 61 Child Protection Monitoring and Evaluation Reference Group,
ior Scale-Child Report’, Child Maltreatment, vol. 9, no. 5, op. cit.; Ondrusek, N., et al., ‘Empirical Examination of the
2004, pp. 409-428. Ability of Children to Consent to Clinical Research’, Journal of
Medical Ethics, vol. 24, no. 3, 1998, pp. 158-165; Abramov-
41 Dubowitz, Pitts and Black, op. cit.
itch, R., et al., ‘Children’s Capacity to Consent to Participation
42 Stowman, S. A., and B. Donohue, ‘Assessing Child Neglect: in Psychological Research: Empirical findings’, Child Develop-
A review of standardized measures’, Aggression and Violent ment, vol. 62, no. 5, 1991, pp. 1100-1109.
Behavior, vol. 10, no. 4, 2005, pp. 491-512; Totsika, V., and
K. Sylva, ‘The Home Observation for Measurement of the
op. cit.
Environment Revisited’, Child and Adolescent Mental Health,
vol. 9, no. 1, 2004, pp. 25-35; Lyons, P., et al., ‘The Child Well- 63 Child Protection Monitoring and Evaluation Reference Group,
Being Scales as a Clinical Tool and a Management Informa- op. cit.; World Health Organization, Draft WHO Ethical
tion System’, Child Welfare, vol. 78, no. 2, 1999, pp. 241- and Safety Guidelines for Sexual and Reproductive Health
258; Lyons, P., et al., ‘Risk Assessment for Child Protective Research and Data Collection with Adolescents, WHO,
Services: A review of the empirical literature on instrument Geneva, 2011; Edmonds, C., Ethical Considerations when
performance’, Social Work Research, vol. 20, no. 3, 1996, pp. Conducting Research on Children in the Worst Forms of Child
143-155. Labour in Nepal, International Programme on the Elimination
of Child Labour, International Labour Organization, Geneva,
43 Fallon et al., op. cit.
2005.
44 ChildONEurope, op. cit.
45 Krug et al., op. cit. op. cit.
46 United Nations Children’s Fund, 2006, op. cit. 65 Clacherty, G., and D. Donald, ‘Child Participation in Research:
Reflections on ethical challenges in the southern African con-
47 Fallon et al., op. cit.; Krug et al., op. cit.
text’, African Journal of AIDS Research, vol. 6, no. 2, 2007,
48 English, D. J., et al., ‘Maltreatment’s Wake: The relationship pp. 147-156.
of maltreatment dimensions to child outcomes’, Child Abuse
66 ChildONEurope, op. cit.; Krug et al., op. cit.
& Neglect, vol. 29, no. 5, 2005, pp. 597-619; MacMillan, H. L.,
et al., ‘Prevalence of Child Physical and Sexual Abuse in the 67 Child Protection Monitoring and Evaluation Reference Group,
References
Community: Results from the Ontario health supplement’, op. cit.

JAMA, vol. 278, no. 2, 1997, pp. 131-135.
49 Fallon et al., op. cit. op. cit.
50 Krug et al., op. cit. 69 Straus, M. A., ‘Measuring Intrafamily Conflict and Violence:
179
The conflict tactics (CT) scales’, Journal of Marriage and the op. cit.
Family, vol. 41, no. 1, 1979, pp. 75-88; Straus, M. A., et al.,
11 Zielinski, D. S., ‘Child Maltreatment and Adult Socioeconomic
‘Identification of Child Maltreatment with the Parent-Child
Well-being’, Child Abuse & Neglect, vol. 33, no. 10, 2009, pp.
Conflict Tactics Scales: Development and psychometric data
666-678.
for a national sample of American parents’, Child Abuse &
Neglect, vol. 22, no. 4, 1998, pp. 249-270. 12 Lansford et al., op. cit.
70 This list includes countries that collected information on child 13 Norman et al., op. cit.; Krug et al., op. cit.
discipline in MICS3, MICS4 or both rounds. Several additional
14 Vagi, K. J., et al., ‘Beyond Correlates: A review of risk and pro-
countries are currently completing the preparation of MICS4
tective factors for adolescent dating violence perpetration’,
reports and therefore this number only represents those with
Journal of Youth and Adolescence, vol. 42, no. 4, 2013, pp.
available results as of May 2014.
633-649; Duke et al., op. cit.; Shields, A., and D. Cicchetti,
71 A few countries conducted surveys using standard DHS mod- ‘Parental Maltreatment and Emotion Dysregulation as Risk
ules and terminology but were not part of the DHS global Factors for Bullying and Victimization in Middle Childhood’,
programme. Journal of Clinical Child & Adolescent Psychology, vol. 30, no.
3, 2001, pp. 349-363.
72 According to the pages on the Global School-based Student
Health Surveys on the WHO website (as of May 2014). 15 Krug et al., op. cit.
73 According to the Health Behaviour in School-aged 16 World Health Organization, Department of Health Statistics
Children Study website (as of May 2014). and Information Systems, Global Health Estimates (GHE),
<www.who.int/healthinfo/global_burden_disease/en/>,
74 Surveys conducted as part of an international survey pro-
gramme were not included in the review.
17 United Nations Office on Drugs and Crime, Global Study on
Homicide 2013: Trends, contexts, data, UNODC, Vienna,
2014.
Chapter 3
18 United Nations Office on Drugs and Crime, op. cit., p. 100.
1 United Nations Committee on the Rights of the Child, Gen-
eral Comment No. 13 (2011): The right of the child to free-
dom from all forms of violence, UN document CRC/C/GC/13, 20 World Health Organization, Global Health Estimates (GHE)
Office of the High Commissioner for Human Rights, Geneva, Summary Tables: Deaths by cause, age, sex and region,
18 April 2011. 2012, WHO, Geneva, 2014, recalculated according to
UNICEF’s regional classification.
2 The National Child Traumatic Stress Network, ‘Child Physical
Abuse Fact Sheet’, 2009, <www.nctsn.org/sites/default/files/ 21 This last cause refers to injuries and killings resulting from the
assets/pdfs/ChildPhysicalAbuse_Factsheet.pdf>, accessed use of force and firearms by the police or other law enforce-
17 June 2014. ment agents in the course of arresting or attempting to arrest
lawbreakers, while maintaining order or during other legal
3 Krug, E. G., et al., editors, World Report on Violence and
actions. See: United Nations Office on Drugs and Crime, op.
Health, World Health Organization, Geneva, 2002.
cit., p. 102.
4 United Nations Committee on the Rights of the Child, op. cit.
22 World Health Organization, op. cit., recalculated according to
5 Lysenko, L. J., E. D. Barker and S. R. Jaffee, ‘Sex Differences UNICEF’s regional classification.
in the Relationship between Harsh Discipline and Conduct
23 Viner, R. M., et al., ‘50-year Mortality Trends in Children
Problems’, Social Development, vol. 22, no. 1, 2013, pp. 197-
and Young People: A study of 50 low-income, middle-
214; McKee, L., et al., ‘Harsh Discipline and Child Problem
income and high-income countries’, The Lancet, vol. 377, no.
Behaviors: The roles of positive parenting and gender’, Jour-
9772, 2011, pp. 1162-1174.
nal of Family Violence, vol. 22, no. 4, 2007, pp. 187-196.
24 Recalculated for UNICEF by the United Nations Office on
6 Finkelhor, D., et al.,’Polyvictimization: Children’s exposure to
Drugs and Crime.
multiple types of violence, crime, and abuse’, OJJDP Juvenile
Justice Bulletin, October 2011. 25 United Nations Office on Drugs and Crime, op. cit., p. 15. It
is important to highlight that the available data do not always
7 Duke, N. N., et al., ‘Adolescent Violence Perpetration: Associ-
distinguish between deaths in which the victim was directly
ations with multiple types of adverse childhood experiences’,
involved in some violent or criminal behaviour and those in
Pediatrics, vol. 125, no. 4, 2010, pp. e778-e786.
which the victim may have been an unintended target or inno-
8 Norman, R. E., et al., ‘The Long-term Health Consequences cent bystander. It is certainly the case that many victims die
of Child Physical Abuse, Emotional Abuse, and Neglect: A by homicide even when they are not willing or active partici-
systematic review and meta-analysis’, PLOS Medicine, vol. pants in crime or gang-related activities.
9, no. 11, 2012, pp. 1-31; Lansford, J. E., et al., ‘A 12-year
26 Given the high proportion of homicides for which the per-
Prospective Study of the Long-term Effects of Early Child
petrator is unknown, this figure may be even higher. See:
Physical Maltreatment on Psychological, Behavioral, and
United Nations Office on Drugs and Crime, op. cit., p. 53.
Academic Problems in Adolescence’, Archives of Pediatrics
& Adolescent Medicine, vol. 156, no. 8, 2002, pp. 824-830; 27 United Nations Office on Drugs and Crime, op. cit., p. 49.
Child Welfare Information Gateway, ‘Factsheet: Long-term
28 United Nations Office on Drugs and Crime, op. cit.
consequences of child abuse and neglect’, US Department of
References
Health and Human Services, Children’s Bureau, Washington, 29 Krug et al., op. cit.
DC, 2013; Krug et al., op. cit.
30 Seelke, C. R., Gangs in Central America, Congressional
9 Child Welfare Information Gateway, op. cit. Research Service, Washington, DC, 2011.
10 Child Welfare Information Gateway, op. cit.; Lansford et al., 31 Bennett, T., and K. Holloway, ‘Gang Membership, Drugs and
180
Crime in the UK’, British Journal of Criminology, vol. 44, no. 43 United Nations Office on Drugs and Crime, op. cit., p. 43.
3, 2004, pp. 305-323; Esbensen, F. A., et al., ‘Similarities and
44 Murray, J., D. R. C. Cerqueira and T. Kahn, ‘Crime and Vio-
Differences in Risk Factors for Violent Offending and Gang
lence in Brazil: Systematic review of time trends, prevalence
Membership’, Australian and New Zealand Journal of Crimi-
rates and risk factors’, Aggression and Violent Behavior, vol.
nology (Australian Academic Press), vol. 42, no. 3, 2009, pp.
18, no. 5, 2013, pp. 471-483.
310-335.
45 Murray, Cerqueira and Kahn, op. cit.
32 Howell, J. C., and A. J. Egley, ‘Moving Risk Factors into
Developmental Theories of Gang Membership’, Youth Vio- 46 de Melo, D. L. B., and I. Cano, Homicídios na Adolescência:
lence and Juvenile Justice, vol. 3, no. 4, 2005, pp. 334-354. IHA 2009-2010, Favela Observatory, Rio de Janeiro, 2012.
33 Seelke, op. cit.; Dupéré, V., et al., ‘Affiliation to Youth Gangs 47 United Nations Children’s Fund, Annual Report 2011 for Bra-
During Adolescence: The interaction between childhood zil, UNICEF Regional Office for Latin America and the Carib-
psychopathic tendencies and neighborhood disadvantage’, bean, Panama, 2012, <www.unicef.org/about/annualreport/
Journal of Abnormal Child Psychology, vol. 35, no. 6, 2007, files/Brazil_COAR_2011.pdf>, accessed 1 June 2014.
pp. 1035-1045; Hill, K. G., et al., ‘Childhood Risk Factors for 48 Krug et al., op. cit.
Adolescent Gang Membership: Results from the Seattle
Social Development Project’, Journal of Research in Crime 49 de Melo and Cano, op. cit.
and Delinquency, vol. 36, no. 3, 1999, pp. 300-322; Lizotte, 50 de Melo and Cano, op. cit.
A. J., et al., ‘Factors Influencing Gun Carrying among Young
Urban Males over the Adolescent-Young Adult Life Course’, 51 For further information, see: PRVL, <http://prvl.org.br/>,
Criminology, vol. 38, no. 3, 2000, pp. 811-834. accessed 1 June 2014; Favela Observatory, <http://observa-
toriodefavelas.org.br/en>, accessed 1 June 2014; Secretariat
34 Portillo, N., ‘Estudios Sobre Pandillas Juveniles en El Salvador for Human Rights, <www.sdh.gov.br/>, accessed 1 June
y Centroamérica: Una revisión de su dimensión participativa’, 2014; and Violence Analysis Laboratory: <www.lav.uerj.br/>,
Apuntes de Psicología, vol. 21, no. 3, 2003, pp. 475-493. accessed 1 June 2014.
35 Klein, M. W., and C. L. Maxson, Street Gang Patterns and 52 Secretariat for Human Rights, United Nations Children’s
Policies, Oxford University Press, New York, 2006; Hawkins, Fund, Favela Observatory and Violence Analysis Labora-
J. D., et al., ‘Predictors of Youth Violence’, Juvenile Justice tory, Municipal Guide for Preventing Lethal Violence against
Bulletin, Office of Juvenile Justice and Delinquency Preven- Adolescents and Young People, Favela Observatory, Rio de
tion, Washington, DC, 2000; Hill et al., op. cit.; Eitle, D., S. Janeiro, 2012. For further information, see: <http://observato-
Gunkel and K. Van Gundy, ‘Cumulative Exposure to Stressful riodefavelas.org.br/categoria/acervo/publicacoes/>, accessed
Life Events and Male Gang Membership’, Journal of Criminal 1 June 2014.
Justice, vol. 32, no. 2, 2004, pp. 95-111.
53 Willadino, R., et al., Prevenção à Violência e Redução de
36 Peterson, D., T. J. Taylor and F. A. Esbensen, ‘Gang Member- Homicídios de Adolescentes e Jovens no Brasil, 2011, Favela
ship and Violent Victimization’, Justice Quarterly, vol. 21, no. Observatory, Rio de Janeiro, 2011.
4, 2004, pp. 793-815; Taylor, T. J., et al., ‘Gang Membership
as a Risk Factor for Adolescent Violent Victimization’, Journal 54 For further information, see: <www.juventude.gov.br/juven-
of Research in Crime and Delinquency, vol. 44, no. 4, 2007, tudeviva/>, accessed 1 June 2014.
pp. 351-380; Melde, C., T. J. Taylor and F. A. Esbensen, ‘“I 55 de Melo and Cano, op. cit.; United Nation’s Children’s Fund,
Got Your Back”: An examination of the protective function of Secretariat for Human Rights, Favela Observatory and Vio-
gang membership in adolescence’, Criminology, vol. 47, no. lence Analysis Laboratory, Homicídios na Adolescência: IHA
2, 2009, pp. 565-594. 2005/2007, Secretariat for Human Rights, Brasilia, 2010.
37 Miller, J., and S. H. Decker, ‘Young Women and Gang Vio- 56 Willadino, op. cit.
lence: Gender, street offending and violent victimization in
57 For further information, see: <http://observatoriodefavelas.
gangs’, Justice Quarterly, vol. 18, no. 1, 2001, pp. 115-140.
org.br/categoria/acervo/publicacoes/>, accessed 1 June
38 Beckett, H., et al., It’s Wrong...But You Get Used to It: A quali- 2014.
tative study of gang-associated sexual violence towards, and
58 See: <http://monitoramento.seppir.gov.br/paineis/pjv/index.
exploitation of, young people in England, University of Bed-
vm?f=vulnerabilidades>, accessed 1 June 2014.
fordshire, Luton, 2013; The Centre for Social Justice, Girls
and Gangs, The Centre for Social Justice, London, 2014. 59 United Nation’s Children’s Fund, UNICEF Seal of Approval,
<www.selounicef.org.br/_selounicef.php?op=1&k=2>,
39 Gatti, U., et al., ‘Youth Gangs, Delinquency and Drug Use: A
accessed 22 May 2014.
test of the selection, facilitation, and enhancement hypoth-
eses’, Journal of Child Psychology and Psychiatry, vol. 46, 60 United Nation’s Children’s Fund, Platform for Urban Centres,
no. 11, 2005, pp. 1178-1190; Katz, C. M., and A. M. Fox, <www.unicef.org/brazil/pt/br_folderPCU_EN_1316.pdf>,
‘Risk and Protective Factors Associated with Gang-involved accessed 22 May 2014.
Youth in Trinidad and Tobago’, Pan American Journal of Public
61 Ministerio de Salud Pública y Asistencia Social, Instituto
Health, vol. 27, no. 3, 2010, pp. 187-202; Swahn, M. H., et al.,
Nacional de Estadística, Centros de Control y Prevención de
‘Alcohol and Drug Use among Gang Members: Experiences
Enfermedades, et al., Encuesta Nacional de Salud Materno
of adolescents who attend school’, Journal of School Health,
Infantil (ENSMI) 2008-2009, Guatemala, 2011.
vol. 80, no. 7, 2010, pp. 353-360.
62 Instituto Nacional de Información de Desarrollo and Ministe-
40 National Gang Center, National Youth Gang Survey Analysis,
References
rio de Salud, Encuesta Nicaragüense de Demografía y Salud

<www.nationalgangcenter.gov/Survey-Analysis>, accessed
(ENDESA) 2006/07: Informe final, Managua, 2008.
12 May 2014.
63 Institut National de la Statistique, Enquête Nationale Sur Le
41 United Nations Office on Drugs and Crime, op. cit., p. 43.
Suivi des Objectifs du Millénaire pour le Développement
42 United Nations Office on Drugs and Crime, op. cit., p. 43. (ENSOMD) à Madagascar, 2012-2013, Antananarivo, 2013.
181
64 Vanuatu Women’s Centre, The Vanuatu National Survey on with at least one child are: Australia (2 per cent), Denmark
Women’s Lives and Family Relationships, Port Vila, 2011. (2 per cent), Mozambique (7 per cent) and the Philippines (2
per cent).
65 ICON-Institut Public Sector, Hacettepe Institute of Popula-
tion Studies and BNB Consulting Ltd., National Research on 77 Kishor, S., and K. Johnson, Profiling Domestic Violence: A
Domestic Violence against Women in Turkey, Republic of multi-country study, ORC Macro, Calverton, 2004.
Turkey Prime Ministry Directorate General on the Status of
78 It is worth noting that those boys aged 13 to 17 years who did
Women, Ankara, 2009.
not report experiencing physical violence at the time of the
66 United Nations Children’s Fund Kenya Country Office, Divi- survey are still at risk of being victimized before they leave
sion of Violence Prevention, National Center for Injury Pre- childhood.
vention and Control, US Centers for Disease Control and
Prevention and Kenya National Bureau of Statistics, Violence
against Children in Kenya: Findings from a 2010 national sur- Chapter 4
vey – Summary report on the prevalence of sexual, physi-
cal and emotional violence, context of sexual violence, and 1 United Nations Committee on the Rights of the Child, Gen-
health and behavioral consequences of violence experienced eral Comment No. 13 (2011): The right of the child to free-
in childhood, Nairobi, 2012. dom from all forms of violence, UN document CRC/C/GC/13,
67 United Nations Children’s Fund Swaziland, A National Study Office of the High Commissioner for Human Rights, Geneva,
on Violence against Children and Young Women in Swaziland, 18 April 2011.
UNICEF Swaziland, Mbabane, 2007. 2 United Nations Children’s Fund, Preventing and
68 United Nations Children’s Fund, US Centers for Disease Con- Responding to Child Sexual Abuse and Exploitation:
trol and Prevention and Muhimbili University of Health and Evidence review, UNICEF, New York, 2014 (forthcoming).
Allied Sciences, Violence against Children in Tanzania: Find- 3 United Nations Children’s Fund, 2014, op. cit.
ings from a national survey, 2009, UNICEF Tanzania, Dar es
Salaam, 2011. 4 United Nations General Assembly, Optional Protocol to the
Convention on the Rights of the Child on the Sale of Children,
69 Zimbabwe National Statistics Agency, United Nations Child Prostitution and Child Pornography, United Nations,
Children’s Fund and Collaborating Centre for Operational New York, 2000, <www.ohchr.org/EN/ProfessionalInterest/
Research and Evaluation, National Baseline Survey on Life Pages/OPSCCRC.aspx>, accessed 14 April 2014.
Experiences of Adolescents, 2011, Harare, 2013.
5 Council of Europe Convention on the Protection of Children
70 It is worth noting that those girls aged 13 to 17 years who did against Sexual Exploitation and Sexual Abuse (CETS no. 201),
not report experiencing physical violence at the time of the <www.coe.int/t/dghl/standardsetting/children/Lanzarote%20
survey are still at risk of being victimized before they leave Convention_EN.pdf>, accessed 5 May 2014.
childhood.
71 In Kenya, the United Republic of Tanzania and Zimbabwe,
the overall prevalence rates for physical violence among girls 7 Gallagher, B., et al., ‘The Sexual Abuse of Children by Strang-
in the VACS exceeded 60 per cent compared to rates of 30 ers: Its extent, nature and victims’ characteristics’, Children
per cent found through the DHS. These discrepancies can and Society, vol. 16, no. 5, 2002, pp. 346-359.
be largely attributed to the different indicator definitions and 8 Grooming refers to a gradual process through which offend-
age ranges covered in the surveys. DHS asked adolescent ers increasingly sexualize a relationship.
girls about their experiences of physical violence since age
15, while the VACS asked adolescents and young adults ret- 9 Mudaly, C., and C. Goddard, The Truth is Longer than a Lie:
Children’s experiences of abuse and professional interven-
rospectively about their experiences of physical violence prior
tions, Jessica Kingsley Publishers, London, 2006.
to age 18, which would include experiences of physical vio-
lence prior to age 15. 10 A wealth of research literature explores the causes, risk fac-
tors, consequences of and responses to sexual violence. The
72 Data were collected in other countries that conducted a DHS
review included in this chapter is meant to provide only a brief
but could not be presented for girls aged 15 to 19 either
overview of what is known about sexual violence in relation
because the datasets are unavailable or restricted or because
to children and should be considered illustrative rather than
results are based on less than 25 unweighted cases.
exhaustive.
73 Devries, K. M., et al., ‘Intimate Partner Violence during Preg-
11 Krahe, B., et al., ‘Childhood Sexual Abuse and Revictimization
nancy: Analysis of prevalence data from 19 countries’, Repro-
in Adolescence’, Child Abuse & Neglect, vol. 23, no. 4, 1999,
ductive Health Matters, vol. 18, no. 36, 2010, pp. 158-170;
pp. 383-394.
Krug et al., op. cit.
12 Chen, L., ‘Sexual Abuse and Lifetime Diagnosis of Psychiatric
74 Ganatra, B. R., K. J. Coyaji and V. N. Rao, ‘Too Far, Too Little,
Disorders: Systematic review and meta-analysis’, Mayo Clinic
Too Late: A community-based case-control study of mater-
Proceedings, vol. 85, no. 7, 2010, pp. 618-629.
nal mortality in rural west Maharashtra, India’, Bulletin of the
World Health Organization, vol. 76, no. 6, 1998, pp. 591-598. 13 Molnar, B., et al., ‘Psychopathology, Childhood Sexual Abuse
and Other Childhood Adversities: Relative links to subsequent
75 Ahmed, S., M. A. Koenig and R. Stephenson, ‘Effects of
suicidal behaviour in the US’, Psychological Medicine, vol. 31,
Domestic Violence on Perinatal and Early-childhood Mortal-
no. 6, 2001, pp. 965-977; Brown, J., et al., ‘Child Abuse and
ity: Evidence from North India’, American Journal of Public
Neglect: Specificity of effects on adolescent and young adult
References
Health, vol. 96, no. 8, 2006, pp. 1423-1428.

depression and suicidality’, Journal of the American Academy
76 The International Violence against Women Surveys also col- of Child and Adolescent Psychiatry, vol. 38, no. 12, 1999, pp.
lected data on violence in pregnancy among adult women. 190-196; Evans, E., et al., ‘Suicidal Phenomena and Abuse
According to these surveys, the rates of physical or sexual in Adolescents: A review of epidemiological studies’, Child
violence during pregnancy among 18- to 69-year-old women Abuse & Neglect, vol. 29, no. 1, 2005, pp. 45-58.
182
14 Dinwiddie, S., et al., ‘Early Sexual Abuse and Lifetime Psy- Child Maltreatment, vol. 3, no. 1, 1998, pp. 63-71; Paine, M.,
chopathology: A co-twin-control study’, Psychological Medi- and D. Hansen, ‘Factors Influencing Children to Self-disclose
cine, vol. 30, no. 1, 2000, pp. 41-52; Brown et al., op. cit. Sexual Abuse’, Clinical Psychology Review, vol. 22, no. 2,
2002, pp. 271-295.
30 Nasjleti, M., ‘Suffering in Silence: The male incest victim’,
16 Pinheiro, P. S., World Report on Violence against Children,
Child Welfare, vol. 59, no. 5, 1980, pp. 269-275.
United Nations Secretary-General’s Study on Violence against
Children, United Nations, Geneva, 2006. 31 Nielsen, T., ‘Sexual Abuse of Boys: Current perspectives’,
Personnel and Guidance Journal, vol. 62, no. 3, 1983, pp.
17 Andrews, G., et al., ‘Child Sexual Abuse’, in Comparative
Quantification of Health Risks: Global and regional burden 139-142.
of disease attributable to selected major risk factors, vol. 2, 32 Nasjleti, op. cit.
edited by M. Ezzati et al., World Health Organization, Geneva,
33 United Nations High Commissioner for Refugees, Rapid
2004; Clemmons, J., et al., ‘Unique and Combined Contribu-
Inter-agency Sexual and Gender-based Violence Assessment,
tions of Multiple Child Abuse Types and Abuse Severity to
Adult Trauma Symptomatology’, Child Maltreatment, vol. 12, Dadaab Refugee Camps and Outskirts: July-August 2011,
no. 2, 2007, pp. 172-181. UNHCR, Geneva, 2011.
18 Howe, D., Child Abuse and Neglect: Attachment, 34 Stoltenborgh, M., et al., ‘A Global Perspective on Child Sexual
development and intervention, Palgrave Macmillan, London, Abuse: Meta-analysis of prevalence around the world’, Child
2005. Maltreatment, vol. 16, no. 2, 2011, pp. 79-101.
19 United Nations Children’s Fund, 2014, op. cit. 35 Ellsberg, M., and L. Heise, Researching Violence against
Women: A practical guide for researchers and activists, World
20 Dinwiddie et al., op. cit.; Molnar et al., op. cit. Health Organization and Program for Appropriate Technology
21 Widom, C., Childhood Victimization: Early adversity, later psy- in Health (PATH), Geneva, 2005.
chopathology, National Institute of Justice, Washington, DC, 36 United Nations Department of Economic and Social Affairs,
2000. Guidelines for Producing Statistics on Violence against
22 Daigneaulta, I., et al., ‘Men’s and Women’s Childhood Sexual Women: Statistical surveys, UNDESA, New York, 2013.
Abuse and Victimization in Adult Partner Relationships: A 37 Ellsberg, M., et al., ‘Researching Domestic Violence against
study of risk factors’, Child Abuse & Neglect, vol. 33, no. 9, Women: Methodological and ethical considerations’, Stud-
2009, pp. 638-647. ies in Family Planning, vol. 32, no. 1, 2001, pp. 1-16; World
23 Wilson, H., and C. Widom, ‘The Role of Youth Problem Behav- Health Organization, Putting Women First: Ethical and safety
iour on the Path from Child Abuse and Neglect to Prostitution: recommendations for research on domestic violence against
A prospective study’, Journal of Research on Adolescence, women, Department of Gender and Women’s Health, Family
vol. 20, no. 1, 2010, pp. 210-236. and Community Health, WHO, Geneva, 2001.
24 Vizard, E., N. Hickey and E. McCrory, ‘Developmental Trajec- 38 Andrews et al., op. cit.; Barth, J., et al., ‘The Current Preva-
tories Associated with Juvenile Sexually Abusive Behaviour lence of Child Sexual Abuse Worldwide: A systematic review
and Emerging Severe Personality Disorder in Childhood: and meta-analysis’, International Journal of Public Health, vol.
3-year study’, British Journal of Psychiatry, vol. 190, no. 49, 58, no. 3, 2013, pp. 469-483; Stoltenborgh et al., op. cit.
2007, pp. 27-32. 39 Barth et al., op. cit.
25 Classen, C. C., et al. ‘Sexual Revictimization: A review of the 40 Peterman A., T. Palermo and C. Bredenkamp, ‘Estimates
empirical literature’, Trauma Violence & Abuse, vol. 6, no. and Determinants of Sexual Violence against Women in the
2, 2005, pp. 103-129; Finkelhor, D., R. K. Ormrod and H. A. Democratic Republic of Congo’, American Journal of Public
Turner, ‘Re-victimization Patterns in a National Longitudinal Health, vol. 101, no. 6, 2011, pp. 1060-1067.
Sample of Children and Youth’, Child Abuse & Neglect, vol.
31, no. 5, 2007, pp. 479-502. 41 Pereda, N., et al., ‘The International Epidemiology of Child
Sexual Abuse: A continuation of Finkelhor (1994)’, Child
26 Finkelhor, D., R. K. Ormrod and H. A. Turner, ‘Poly-victimiza- Abuse & Neglect, vol. 33, no. 6, 2009, pp. 331-342; Collin-
tion and Trauma in a National Longitudinal Cohort’, Develop- Vézina, D., I. Daigneault and M. Hébert, ‘Lessons Learned
ment and Psychopathology, vol. 19, no. 1, 2007, pp. 149-166. from Child Sexual Abuse Research: Prevalence, outcomes,
27 Finklehor, D., and A. Browne. ‘The Traumatic Impact of Child and preventive strategies’, Child and Adolescent Psychiatry
Sexual Abuse: A conceptualization’, American Journal of and Mental Health, vol. 7, no. 22, 2013, pp. 1-9; Andrews et
Orthopsychiatry, vol. 55, no. 4, 1985, pp. 530-541. al., op. cit.; Stoltenborgh et al., op. cit.; Pereda, N., et al., ‘The
Prevalence of Child Sexual Abuse in Community and Student
28 Arata, C. M., et al., ‘Single Versus Multi-type Maltreatment:
Samples: A meta-analysis’, Clinical Psychology Review, vol.
An examination of the long-term effects of child abuse’,
29, no. 4, 2009, pp. 328-338.
Journal of Aggression, Maltreatment & Trauma, vol. 11, no.
4, 2005, pp. 29-52; Finkelhor, Ormrod and Turner, op. cit.; 42 The question on whether respondents were forced with
Elloneni, N., and V. Salmi, ‘Poly-victimization as a Life Condi- threats or in any other way to perform sexual acts when she/
tion: Correlates of poly-victimization among Finnish children’, he did not want to by a spouse or partner was only asked
Journal of Scandinavian Studies in Criminology and Crime in the DHS 2013 in Nigeria, the DHS 2012 in Comoros,
Prevention, vol. 12, no. 1, 2011, pp. 20-44; Ford, J., et al., Gabon, Haiti, Kyrgyzstan and Tajikistan and in the DHS 2011
‘Poly-victimization and Risk of Posttraumatic, Depressive, and in Uganda. In the DHS 2008 in the Philippines, ever-married
References
Substance Use Disorders and Involvement in Delinquency in girls and women were also asked if their current or most
a National Sample of Adolescents’, Journal of Adolescent recent husband or partner ever tried or attempted to force
Health, vol. 46, no. 6, 2010, pp. 545-552. them to have sexual intercourse with him or to perform any
other sexual acts against their will.
29 Arata, C., ‘To Tell or Not to Tell: Current functioning of child
sexual abuse survivors who disclosed their victimization’, 43 World Health Organization, Guidelines for Medico-Legal Care
183
for Victims of Sexual Violence, WHO, Geneva, 2003. respondent’s vagina (in the case of females) or anus with the
penis, hands, fingers, mouth or other objects, or penetrating
44 World Health Organization, 2003, op. cit.
the respondent’s mouth with the penis. For males, sexual
45 Bottoms, B. L., A. G. Rudnicki and M. A. Epstein, ‘A Retro- intercourse could also include someone forcing the respon-
spective Study of Factors Affecting the Disclosure of Child- dent’s penis into his or her mouth, vagina or anus.
hood Sexual and Physical Abuse’, in Disclosing Abuse:
56 Kelly, L., ‘Journeys of Jeopardy: A review of research on
Delays, denials, retractions and incomplete accounts, edited
trafficking in women and children in Europe’, IOM Migration
by M. E. Pipe, et al., Erlbaum, Mahwah, 2007, pp. 175-194.
Research Series, no. 11, International Organization for Migra-
46 Steward, M. S., et al., ‘Implications of Developmental tion, Geneva, 2002.
Research for Interviewing Children’, Child Abuse & Neglect,
57 Chase, E., and J. Statham, The Commercial Sexual Exploita-
vol. 17, no. 1, 1993, pp. 25-37; London, K., et al., ‘Disclosure
tion of Children and Young People: An overview of key litera-
of Child Sexual Abuse: What does the research tell us about
ture and data, Thomas Coram Research Unit, London, 2004.
the ways that children tell?’, Psychology, Public Policy, and
Law, vol. 11, no. 1, 2005, pp. 194-226. 58 Tyldum, G., ‘Limitations in Research on Human Trafficking’,
International Migration, vol. 48, no. 5, 2010, pp. 1-13.
47 Lamb, M. E., et al., ‘Age Differences in Young Children’s
Responses to Open-ended Invitations in the Course of Foren- 59 United Nations Children’s Fund Swaziland, A National Study
sic Interviews’, Journal of Consulting and Clinical Psychology, on Violence against Children and Young Women in Swaziland,
vol. 71, no. 5, 2003, pp. 926-934; London, K., et al., ‘Review UNICEF Swaziland, Mbabane, 2007.
of the Contemporary Literature on How Children Report Sex-
60 United Nations Children’s Fund, US Centers for Disease Con-
ual Abuse to Others: Findings, methodological issues, and
trol and Prevention and Muhimbili University of Health and
implications for forensic interviewers’, Memory, vol. 16, no.
Allied Sciences, Violence against Children in Tanzania: Find-
1, 2008, pp. 29-47.
ings from a national survey 2009, UNICEF Tanzania, Dar es
48 Paine and Hansen, op. cit. Salaam, 2011.
49 Summit, R., ‘Abuse of the Child Sexual Abuse Accommoda- 61 It is possible that the difference in prevalence rates obtained
tion Syndrome’, Journal of Child Sexual Abuse, vol. 1, no. 4, from the two surveys can be partially explained by the fact
1992, pp. 153-163. that different age groups are covered.
50 Paine and Hansen, op. cit. 62 United Nations Children’s Fund Kenya Country Office, Divi-
sion of Violence Prevention, National Center for Injury Pre-
51 Paine and Hansen, op. cit.
vention and Control, US Centers for Disease Control and
52 World Health Organization, 2003, op. cit. Prevention and Kenya National Bureau of Statistics, Violence
against Children in Kenya: Findings from a 2010 National
53 Bruck, M., and S. Ceci, ‘Forensic Developmental Psychology:
Survey – Summary report on the prevalence of sexual, physi-
Unveiling four common misconceptions’, Current Directions
cal and emotional violence, context of sexual violence, and
in Psychological Science, vol. 13, no. 6, 2004, pp. 229-232.
health and behavioral consequences of violence experienced
54 Finkelhor, D., ‘The International Epidemiology of Child Sexual in childhood, Nairobi, 2012.
Abuse’, Child Abuse & Neglect, vol. 18, no. 5, 1994, pp. 409-
63 Zimbabwe National Statistics Agency, United Nations
417; Andrews et al., op. cit.
Children’s Fund and Collaborating Centre for Operational
55 The definitions used in the surveys in Kenya, Swaziland, the Research and Evaluation, National Baseline Survey on Life
United Republic of Tanzania and Zimbabwe are as follows: In Experiences of Adolescents, 2011, Harare, 2013.
the Swaziland survey, sexual violence included the following
64 Wakabi, W., ‘Sexual Violence Increasing in Democratic
acts: physically forced sexual intercourse by a man or boy;
Republic of Congo’, The Lancet, vol. 371, no. 9606, 2008, pp.
coerced intercourse in which a man or boy persuaded or pres-
15-16; Amowitz, M. L., et al., ‘Prevalence of War-related Sex-
sured the respondent to have sexual intercourse against her
ual and Other Human Rights Abuses among Internally Dis-
will; attempted unwanted intercourse in which a man or boy
placed Persons in Sierra Leone’, JAMA, vol. 287, no. 4, 2002,
tried to make the respondent have sexual intercourse when
pp. 513-521; Malemo, K. L., et al., ‘Sexual Violence toward
she did not want to but was unsuccessful; and unwanted
Children and Youth in War-torn Eastern Democratic Repub-
touching in which a man or boy sexually touched the respon-
lic of Congo’, PLOS ONE, vol. 6, no. 1, 2011, e15911. The
dent against her will (such as kissing, grabbing or fondling)
issue of the impact of armed conflict on children is explored
or forced the respondent to touch his private parts against
in great detail in: United Nations, Impact of Armed Conflict on
her will but did not try to force her to have sexual inter-
Children: Report of the expert of the Secretary-General, Ms.
course. Sexual intercourse was considered to be any sexual
Graça Machel, submitted pursuant to General Assembly Res-
act involving vaginal or anal intercourse. In the Kenya, the
olution 48/157, UN document A/51/306, New York, 26 August
United Republic of Tanzania and Zimbabwe surveys, sexual
1996, and its follow-up report: Office of the Special Repre-
violence included the following acts: unwanted touching of
sentative of the Secretary-General for Children and Armed
the respondent in which a perpetrator touched the respon-
Conflict and United Nations Children’s Fund, Machel Study
dent against his or her will in a sexual way, such as unwanted
10-year Strategic Review: Children and conflict in a changing
touching, kissing, grabbing or fondling, but did not try to force
world, UNICEF, New York, 2009.
him or her to have sexual intercourse; attempted unwanted
intercourse in which a perpetrator tried to make the respon- 65 Canning, V., ‘Who’s Human? Developing sociological under-
dent have sexual intercourse when he or she did not want to, standings of the rights of women raped in conflict’, Interna-
but was unsuccessful; physically forced intercourse in which tional Journal of Human Rights, vol. 14, no. 6, 2010, pp. 849-
References
a perpetrator physically forced the respondent to have sexual 864; Giller, J. E., P. J. Bracken and S. Kabaganda, ‘Uganda:
intercourse against his or her will; and coerced intercourse in War, women and rape’, The Lancet, vol. 337, no. 8741, 1991,
which a perpetrator pressured or non-physically forced the p. 604; Swiss, S., and J. E. Giller, ‘Rape as a Crime of War:
respondent to have sexual intercourse against his or her will. A medical perspective’, JAMA, vol. 270, no. 5, 1993, pp.
Sexual intercourse was considered to include penetrating the 612-615.
184
66 Human Rights Watch, Abducted and Abused: Renewed con- project, European Online Grooming Project for the European
flict in northern Uganda, Human Rights Watch, New York, Commission Safer Internet Plus Programme, London, 2010.
2003; Swiss, S., et al., ‘Violence against Women during
the Liberian Civil Conflict’, JAMA, vol. 279, no. 8, 1998, pp.
625-629. 84 Berson and Berson, op. cit.
67 Amowitz et al., op. cit.; Pinheiro, op. cit. 85 Wolak, J., et al., ‘Online “Predators” and their Victims’, Psy-
chology of Violence, vol. 1, no. S, 2010, pp. 13-35.
68 United Nations High Commissioner for Refugees, Sexual
Violence against Refugees: Guidelines on prevention and 86 United Nations General Assembly, op. cit., article 2(c).
response, UNHCR, Geneva, 1995; Ferris, E., ‘Abuse of
Power: Sexual exploitation of refugee women and girls’, Jour-
nal of Women in Culture and Society, vol. 32, no. 3, 2007, 88 United Nations Children’s Fund, 2011, op. cit.
pp. 584-591; World Health Organization, Research Agenda
89 Arnaldo, C. A., editor, Child Abuse on the Internet: Ending
for Sexual Violence in Humanitarian, Conflict and Post-conflict
the silence, United Nations Educational, Scientific and Cul-
Settings, WHO, Geneva, 2012; Child Protection and Gender-
tural Organization, Paris, 2001.
based Violence Sub-Working Group Jordan, Findings from
the Inter-Agency Child Protection and Gender-Based Violence 90 Chase and Statham, op. cit.
Assessment in the Za’atari Refugees Camp, Jordan, 2013.
69 Hossain, M., et al. ‘Men’s and Women’s Experiences of Vio-
92 United Nations Children’s Fund, 2011, op cit.
lence and Traumatic Events in Rural Côte d’Ivoire Before, Dur-
ing and After a Period of Armed Conflict,’ BMJ Open, vol. 4, 93 Davidson, J. C., and E. Martellozzo, ‘Protecting Vulnerable
no. 2, 2014, pp. 1-9. Young People in Cyberspace from Sexual Abuse: Raising
awareness and responding globally’, Police Practice and
70 Malemo et al., op. cit.
Research: An International Journal, vol. 9, no. 4, 2008, pp.
71 Malemo et al., op. cit. 277-289.
72 Longombe, A. O., K. M. Claude and J. Ruminjo, ‘Fistula and 94 Estes, R. J., The Sexual Exploitation of Children: A working
Traumatic Genital Injury from Sexual Violence in a Conflict guide to the empirical literature, School of Social Work, Uni-
Setting in Eastern Congo: Case studies’, Reproductive Health versity of Pennsylvania, Philadelphia, 2001.
Matters, vol. 16, no. 31, 2008, pp. 132-141; Loncar, M., et
95 Chase and Statham, op. cit.; Webster et al., op. cit.
al., ‘Psychological Consequences of Rape on Women in the
1991-1995 War in Croatia and Bosnia and Herzegovina’, Croa- 96 Davidson and Martellozzo, op. cit.
tian Medical Journal, vol. 47, no. 1, 2006, pp. 67-75.
97 American Professional Society on the Abuse of Children,
73 Casey, S. E., et al., ‘Care-seeking Behavior by Survivors of ‘APSAC Statement on the Harm to Child Pornography Vic-
Sexual Assault in the Democratic Republic of the Congo’, tims’, October 2013, <www.apsac.org/assets/documents/
American Journal of Public Health, vol. 101, no. 6, 2011, pp. apsac%20statement%20on%20harm%20to%20child%20
1054-1055; Duroch, F., M. McRae and R. F. Grais, ‘Descrip- pornography%20victims%2010.29.13.pdf>, accessed 20
tion and Consequences of Sexual Violence in Ituri Province, June 2014.
Democratic Republic of Congo’, BMC International Health
98 The National Society for the Prevention of Cruelty against
and Human Rights, vol. 11, no. 5, 2011, pp. 1-8; Johnson,
Children working definition of sexual bullying. See: <www.
K., et al., ‘Association of Sexual Violence and Human Rights
nspcc.org.uk/Inform/resourcesforteachers/classroom-
Violations with Physical and Mental Health in Territories of the
resources/sexual_bullying_definition_wdf68769.pdf>,
Eastern Democratic Republic of the Congo’, JAMA, vol. 304,
no. 5, 2010, pp. 553-562.
99 Alvarez, A. R. G., ‘“IH8U”: Confronting cyberbullying and
74 Hossain, et al., op. cit.
exploring the use of cybertools in teen dating relationships’,
75 Berson, I. R., and M. J. Berson, ‘Challenging Online Behav- Journal of Clinical Psychology: In Session, vol. 68, no. 11,
iors of Youth: Findings from a comparative analysis of young 2012, pp. 1205-1214.
people in the United States and New Zealand’, Social Science
100 Questions on experiences of sexual violence were asked
Computer Review, vol. 23, no. 1, 2005, pp. 29-38.
among boys and men in the DHS 2007 in Ukraine, but data
76 Berson and Berson, op. cit. were not presented in the final DHS report and could not be
calculated from the dataset.
77 United Nations Children’s Fund, Child Safety Online: Global
challenges and strategies, UNICEF Innocenti Research Cen- 101 It is worth noting that those boys aged 13 to 17 who had not
tre, Florence, 2011. experienced sexual violence at the time of the survey are
still at risk of being victimized before they leave childhood.
78 Livingstone, S., and L. Haddon, EU Kids Online: Final report,
EU Kids Online, London School of Economics and Political 102 Questions on age at first experience of sexual violence
Science, London, 2009. among girls and women were also asked in other countries
that conducted a DHS, but results are based on less than
79 Council of Europe Convention on the Protection of Children
25 unweighted cases and therefore are not included here.
against Sexual Exploitation and Sexual Abuse, op cit.
These questions were also asked in the DHS 2011 in Equa-
80 Finkelhor, D., Child Sexual Abuse: New theory and research, torial Guinea, but the data could not be recalculated because
Free Press, New York, 1984. access to the dataset was restricted.
References
81 Canadian Centre for Child Protection, Child Sexual Abuse: It 103 Centro Paraguayo de Estudios de Población (CEPEP),
is your business, <www.cybertip.ca/pdfs/C3P_ChildSexualA- Encuesta Nacional de Demografía y Salud Sexual y Repro-
buse_ItIsYourBusiness_en.pdf>, accessed 20 June 2014. ductiva 2008 (ENDSSR 2008): Informe final, Asunción, 2009.
82 Webster, S., et al., Scoping Report: European online grooming 104 Serbanescu F., A. Ruiz and D. B. Suchdev, Reproductive
Health Survey Jamaica 2008: Final report, US Centers for
185
Disease Control and Prevention, Atlanta, 2010. Suivi des Objectifs du Millénaire pour le Développement
(ENSOMD) à Madagascar, 2012-2013, Antananarivo, 2013.
105 Turkish Statistical Institute, The National Research on
Domestic Violence against Women in Turkey, Ankara, 2009. 116 Petersen, I., A. Bhana and M. McKay, ‘Sexual Violence and
Youth in South Africa: The need for community-based pre-
106 Secretariat of the Pacific Community, Kiribati Family Health
and Support Study: A study on violence against women and vention interventions’, Child Abuse & Neglect, vol. 29, no.
children, Secretariat of the Pacific Community, Noumea, 11, 2005, pp. 1233-1248.
New Caledonia, 2010. 117 Petersen, Bhana and McKay, op. cit.
107 Secretariat of the Pacific Community, Solomon Islands Fam- 118 South African Police Service, Crime Report 2010/2011,
ily Health and Safety Study: A study on violence against <www.pmg.org.za/files/docs/110908crimestats2011.pdf>,
women and children, Secretariat of the Pacific Community, accessed 11 June 2014. The Sexual Offences Act of 2007
Honiara, 2009. broadened the definition of sexual offences beyond rape to
108 Vanuatu Women’s Centre, Vanuatu National Survey on include other forms of both penetrative and non-penetra-
Women’s Lives and Family Relationships, Vanuatu Wom- tive sexual contact (including, for instance, sexual assault,
en’s Centre, Port Vila, 2011. indecent exposure, incest, child sexual exploitation, child
pornography, sexual harassment and trafficking for sexual
109 Questions on forced first sex among girls and women were purposes).
also asked in the MICS 2010-2011 in Kazakhstan and the
DHS 2007 in Tuvalu, but results are based on less than 25 119 South African Police Service, Annual Report 2008/2009,
unweighted cases and therefore are not included here. In <http://db3sqepoi5n3s.cloudfront.net/files/docs/090922
the DHS 2007 in Ukraine, estimates on forced first sex crimestats.pdf>, accessed 11 June 2014.
were only calculated for the number of girls and women 120 Dawes, A., and C. L. Ward, ‘Levels, Trends and Determi-
who reported having sex for the first time before age 20. nants of Child Maltreatment in the Western Cape Province’,
Questions on forced first sex among girls and women were in The State of Population in the Western Cape Province,
asked in the DHS 2011 in Equatorial Guinea, but data are not edited by R. Marindo et al., HSRC Press, Cape Town, 2008.
presented in the final DHS report and access to the dataset
was restricted. 121 Jewkes, R. et al., ‘What We Know and Don’t Know: Single
and multiple perpetrator rape in South Africa’, South African
110 Questions on forced first sex among boys and men were
Crime Quarterly, no. 41, 2012, pp. 11-19.
asked in the DHS 2011 in Cameroon, but data could not be
calculated from the dataset. 122 Burton, P., and L. Leoschut, School Violence in South Africa:
Results of the 2012 National School Violence Study, Centre
111 Awusabo-Asare, K., et al., Adolescent Sexual and Reproduc-
of Justice and Crime Prevention, Cape Town, 2013.
tive Health in Ghana: Results from the 2004 National Sur-
vey of Adolescents, Occasional Report, no. 22, Guttmacher 123 This study involved a representative sample of men aged
Institute, New York, 2006; Munthali, A., et al., Adolescent 18 to 49 in the Eastern Cape and KwaZulu-Natal provinces.
Sexual and Reproductive Health in Malawi: Results from the See: Jewkes, R., et al. ‘Gender Inequitable Masculinity and
2004 National Survey of Adolescents, Occasional Report, Sexual Entitlement in Rape Perpetration South Africa: Find-
no. 24, Guttmacher Institute, New York, 2006; Neema, ings of a cross-sectional study’, PLOS ONE, vol. 6, no. 12,
S., et al., Adolescent Sexual and Reproductive Health in 2011, e29590.
Uganda: Results from the 2004 Uganda National Survey
124 The sexual entitlement variable included responses to state-
of Adolescents, Occasional Report, no. 25, Guttmacher
Institute, New York, 2006; Guiella, G., and V. Woog, Santé ments about wanting to ‘have sex’, such as ‘I wanted her
Sexuelle et de la Reproduction des Adolescents au Burkina sexually’, ‘Wanted to prove I could do it’ and ‘Experimenting
Faso: Résultats de l’enquête nationale sur les adolescents with sex’.
du Burkina Faso 2004, Occasional Report, no. 21, 125 Kriminologisches Forschungsinstitut Niedersachsen, Erster
Guttmacher Institute, New York, 2006. Forschungsbericht zur Repräsentativbefragung Sexueller
112 Questions on perpetrators of sexual violence among girls and Missbrauch 2011, Hanover, 2011.
women were also asked in other countries that conducted 126 Data were collected from a nationally representative sample
a DHS, but results are based on less than 25 unweighted of girls and women aged 16 to 70. See: Instituto Nazionale
cases and therefore are not included here. These questions di Statistica, Violence and Abuses against Women Inside
were also asked in the DHS 2011 in Equatorial Guinea, but and Outside the Family, Rome, 2006.
the data could not be recalculated because access to the
dataset was restricted. 127 Data were collected from a representative sample of about
10,000 ninth-graders (aged 15 to 17) attending public
113 Instituto Nacional de Información de Desarrollo and Ministe- schools. See: Averdijk, M., K. Müller-Johnson and M. Eisner,
rio de Salud, Encuesta Nicaragüense de Demografía y Salud Sexual Victimization of Children and Adolescents in Switzer-
(ENDESA) 2006/07: Informe final, Managua, 2008. land, 2nd ed., UBS Optimus Foundation, Zurich, 2012.
114 Questions on perpetrators of sexual violence among boys 128 The Optimus Study is a large-scale research project
and men were also asked in the DHS 2011 in Mozambique launched by the Optimus Foundation of UBS (a global finan-
and the DHS 2008 in the Plurinational State of Bolivia, but cial services company). As part of this initiative, population
results are based on less than 25 unweighted cases and and agency-based surveys were conducted in China and
therefore are not included here. These questions were not
Switzerland between 2008 and 2011 to improve the record-
asked in the DHS 2011 in Cameroon. Questions on perpe-
ing and prevention of sexual abuse and sexual victimization
References
trators of sexual violence were asked among boys and men

of adolescents and children. Further studies and projects are
in the DHS 2007 in Ukraine, but data are not presented in
planned for the coming years. At the time this report was
the final DHS report and could not be calculated from the
prepared, only the findings from the Switzerland study had
dataset.
been released. Information about the initiative can be found
115 Institut National de la Statistique, Enquête Nationale sur le at <www.optimusstudy.org>, accessed 5 May 2014.
186
129 National Society for the Prevention of Cruelty to Children, Developmental Outcomes Beyond That of Other Forms of
Child Abuse and Neglect in the UK Today, NSPCC, London, Maltreatment?’, Child Abuse & Neglect, vol. 29, no. 5, 2005,
2011. pp. 513-532.
130 Finkelhor, D., et al., ‘Violence, Crime, and Abuse Exposure in 6 United Nations General Assembly, Convention on the Rights
a National Sample of Children and Youth: An update’, JAMA- of the Child, UN document A/RES/44/25, United Nations,
Pediatrics, vol. 167, no. 7, 2013, pp. 614-621. New York, 20 November 1989.
131 Awusabo-Asare et al., op. cit.; Munthali et al., op. cit.; 7 United Nations Committee on the Rights of the Child, op. cit.
Neema et al., op. cit.; Guiella and Woog, op. cit.
8 Data coverage was insufficient to calculate averages for most
132 Girls were asked whether they received anything, such regions using this indicator.
as money, gifts or something else, in exchange for sexual
9 Erickson and Egeland, op. cit; Schneider et al., op. cit.
intercourse.
10 Dietz, T. L., ‘Disciplining Children: Characteristics associ-
133 Goran-Svedin, C., ‘Experiences of Sexual Abuse,
ated with the use of corporal punishment and non-violent
Selling Sex and Use of Pornography’, in The Baltic
discipline’, Child Abuse & Neglect, vol. 24, no. 12, 2002, pp.
Sea Regional Study on Adolescents’ Sexuality, edited
1529-1542; Smith, D. J., Love, Fear, and Discipline: Every-
by S. Mossige, M. Ainsaar and C. Goran-Svedin,
day violence toward children in Afghan families, Afghanistan
NOVA, Oslo, 2007.
Research and Evaluation Unit, Kabul, 2008; Tang, C. S., ‘The
134 The sample in Estonia included 16- to 19-year-old students. Rate of Physical Child Abuse in Chinese Families: A commu-
The study in Lithuania included 16- to 20-year-old second- nity survey in Hong Kong’, Child Abuse & Neglect, vol. 22, no.
ary and technical school students. The sample in Norway 5, 1998, pp. 381-391.
consisted of students aged 17 to 19 attending senior high
school in the nine largest ‘town regions’ in the country. The 11 Dietz, op. cit.; Hunter, W. M., et al., ‘Risk Factors for Severe
sample in Poland involved 17- to 18-year-old students. The Child Discipline Practices in Rural India’, Journal of Pediatric
sample in the Russian Federation included 16- to 18-year-old Psychology, vol. 25, no. 6, 2000, pp. 435-447.
secondary and technical school students from four cities in 12 Durrant, J. E., Positive Discipline: What it is and how to do
the Northwest region. The sample in Sweden consisted of it, 2nd ed., Save the Children Sweden, Southeast Asia and
students aged 17 to 19 attending general or practical/voca- Pacific Region, Bangkok, 2011.
tional schools in five cities.
13 Stith, S. M., et al., ‘Risk Factors in Child Maltreatment: A
135 Offering sexual services for pay included involvement in the meta-analytic review of the literature’, Aggression and Vio-
following paid actions: exposing sexual organs to someone, lent Behavior, vol. 14, no. 1, 2009, pp. 13-29; Hashima, P.
involvement in nude photographs/film, showing off mastur- Y., and P. R. Amato, ‘Poverty, Social Support, and Parental
bation, performing oral or anal sex, sexual intercourse and/ Behavior’, Child Development, vol. 65, no. 2, 1994, pp. 394-
or being photographed or filmed while performing sexual 403; Belsky, J., et al., ’Socioeconomic Risk, Parenting During
activities. the Preschool Years and Child Health Age 6 Years’, European
Journal of Public Health, vol. 17, no. 5, 2007, pp. 508-513.
14 The study was also conducted in Jordan, Sweden and the
Chapter 5 United States. Families were recruited from public and pri-
vate schools serving socially and economically diverse pop-
1 United Nations Committee on the Rights of the Child, General ulations in selected cities within each of the participating
Comment No. 8 (2006): The right of the child to protection countries. Although the samples were not nationally repre-
from corporal punishment and other cruel or degrading forms sentative, families from high-, middle- and low-income house-
of punishment, UN document CRC/C/GC/8, Office of the High holds were included in the approximate proportions in which
Commissioner for Human Rights, Geneva, 2 March 2007. these income groups represented the local population. See:
2 Straus, M. A., and M. J. Paschall, ‘Corporal Punishment by Lansford, J., et al., ‘Corporal Punishment of Children in Nine
Mothers and Development of Children’s Cognitive Ability: Countries as a Function of Child Gender and Parent Gender’,
A longitudinal study of two nationally representative age International Journal of Pediatrics, vol. 2010, 2010, Article ID
cohorts’, Journal of Aggression, Maltreatment & Trauma, vol. 672780, pp. 1-12.
18, no. 5, 2009, pp. 459-483. 15 The Canadian Incidence Study of Reported Child Abuse and
3 Erickson, M. F., and B. Egeland, ‘A Developmental View of Neglect employed a multi-stage sampling design in selecting
the Psychological Consequences of Maltreatment’, School a representative sample of 112 child welfare agencies nation-
Psychology Review, vol. 16, no. 2, 1987, pp. 156-168. ally and a sample of child maltreatment cases within these
agencies. Data were collected directly from child protection
4 Gershoff, E. T., ‘Corporal Punishment by Parents and Associ-
workers on a representative sample of 15,980 investigations
ated Child Behaviors and Experiences: A meta-analytic and
conducted during the fall of 2008. This sample was weighted
theoretical review’, Psychological Bulletin, vol. 128, no. 4,
to mirror provincial annual estimates.
2002, pp. 539-579; McCord, J., ‘On Discipline’, Psychologi-
cal Inquiry, vol. 8, no. 3, 1997, pp. 215-217; Straus, M. A., 16 Jud, A., and N. Trocmé, Physical Abuse and Physical Pun-
D. B. Sugarman and J. Giles-Sims, ‘Spanking by Parents and ishment in Canada, Child Canadian Welfare Research Portal
Subsequent Antisocial Behavior of Children’, Archives of Information Sheet No. 122E, 2012, <http://cwrp.ca/infos-
Pediatrics and Adolescent Medicine, vol. 151, no. 8, 1997, heets/physical-abuse-and-physical-punishment-canada>,
pp. 761-767. accessed 13 June 2014.
References
5 Lau, J. T. F., et al., ‘The Relationship between Physical Mal- 17 Data were collected in Sweden as well, and are presented
treatment and Substance Use among Adolescents: A survey together with other national findings in the country sum-
of 95,788 adolescents in Hong Kong’, Journal of Adolescent mary later in this section. See: Bussmann, K. D., C. Erthal
Health, vol. 37, no. 2, 2005, pp. 110-119; Schneider, M. W., and A. Schroth, The Effect of Banning Corporal Punishment
et al., ‘Do Allegations of Emotional Maltreatment Predict in Europe: A five-nation comparison, Martin-Luther-University,
187
Halle-Wittenberg, 2009. S. S., M. J. Dykas and J. Cassidy, ‘Loneliness and Peer Rela-
tions in Adolescence’, Social Development, vol. 21, no. 2,
18 Bussmann, Erthal and Schroth, op. cit.
2012, pp. 273-293.
19 Radford, L., et al., Child Abuse and Neglect in the UK Today,
5 Hartup and Stevens, op. cit.; Brendgen, M., F. Vitaro and W.
National Society for the Prevention of Cruelty to Children,
M. Bukowski, ‘Deviant Friends and Early Adolescents’ Emo-
London, 2011.
tional and Behavioral Adjustment’, Journal of Research on
20 These non-violent forms of discipline included being told why Adolescence, vol. 10, no. 2, 2000, pp. 173-189.
something is wrong, being distracted when doing something
6 See, for example: Erath, S. A., K. S. Flanagan and K.
wrong, withdrawal of privileges and being sent to one’s room.
L. Bierman, ‘Early Adolescent School Adjustment:
21 Verbal aggression included being embarrassed or humiliated, Associations with friendship and peer victimization’, Social
shouted or screamed at, threatened with being smacked, Development, vol. 17, no. 4, 2008, pp. 853-870; Hartup and
sworn at, called names such as ‘stupid’ or ‘lazy’ and threat- Stevens, op. cit.; Waldrip, A. M., K. T. Malcolm and L. A.
ened with being sent away. Jensen-Campbell, ‘With a Little Help from Your Friends: The
22 Lansford et al., op. cit. importance of high-quality friendships on early adolescent
adjustment’, Social Development, vol. 17, no. 4, 2008, pp.
23 Zolotor, A. J., et al., ‘The Emergence of Spanking among a 835-852.
Representative Sample of Children under 2 Years of Age in
North Carolina’, Frontiers in Child and Neurodevelopmental 7 Van Zalk, M. H. W., et al., ‘Peer Contagion and Adolescent
Psychiatry, vol. 2, no. 36, 2011, pp. 1-8. Depression: The role of failure anticipation’, Journal of Clinical
Child and Adolescent Psychology, vol. 39, no. 6, 2010, pp.
24 Ministry of Health, New Zealand, A Portrait of Health: Key 837-848; Rose, A. J., W. Carlson and E. M. Waller, ‘Prospec-
results of the 2006/07 New Zealand Health Survey, Ministry tive Associations of Co-rumination with Friendship and Emo-
of Health, Wellington, 2008. tional Adjustment: Considering the socioemotional trade-offs
25 Global Initiative to End All Corporal Punishment of Children, of co-rumination’, Developmental Psychology, vol. 43, no. 4,
Corporal Punishment of Children in Sweden, <www.end- 2007, pp. 1019-1031.
corporalpunishment.org/pages/pdfs/states-reports/Sweden. 8 Brendgen, Vitaro and Bukowski, op. cit.
pdf>, accessed 13 June 2014.
9 Murray, J., and D. P. Farrington, ‘Risk Factors for Conduct Dis-
26 Ziegert, K. A., ‘The Swedish Prohibition of Corporal Punish- order and Delinquency: Key findings from longitudinal stud-
ment: A preliminary report’, Journal of Marriage and Family, ies’, Canadian Journal of Psychiatry, vol. 55, no. 10, 2010, pp.
vol. 45, no. 4, 1983, pp. 917–926; Durrant, J. E., ‘Evaluating 633-642; Heinze, H. J., P. A. Toro and K. A. Urberg, ‘Antisocial
the Success of Sweden’s Corporal Punishment Ban’, Child Behavior and Affiliation with Deviant Peers’, Journal of Clini-
Abuse & Neglect, vol. 23, no. 5, 1999, pp. 435-448; Durrant, cal Child and Adolescent Psychology, vol. 33, no. 2, 2004, pp.
J. E., ‘Legal Reform and Attitudes toward Physical Punish- 336-346; Brendgen, Vitaro and Bukowski, op. cit.
ment in Sweden’, International Journal of Children’s Rights,
vol. 11, no. 2, 2003, pp. 147-174; Lazerele, R. I., and B. John- 10 Card, N. A., and E. V. E. Hodges, ‘Shared Targets for Aggres-
son, ‘Evaluation of the Effects of Sweden’s Spanking Ban on sion by Early Adolescent Friends’, Developmental Psychol-
Physical Child Abuse Rates: A literature review’, Psychological ogy, vol. 42, no. 6, 2006, pp. 1327-1338.
Reports, vol. 85, no. 2, 1999, pp. 381-392. 11 Goodman, M. R., E. A. Stormshak and T. J. Dishion, ‘The
27 Lansford et al., op. cit. Significance of Peer Victimization at Two Points in Develop-
ment’, Journal of Applied Developmental Psychology, vol.
28 Bussmann, Erthal and Schroth, op. cit. 22, no. 5, 2001, pp. 507-526; Kochenderfer-Ladd, B., and J.
29 Ellonen et al., ‘Adolescents’ Experiences of Parental L. Wardrop, ‘Chronicity and Instability of Children’s Peer Vic-
Violence in Danish and Finnish Families: A compara- timization Experiences as Predictors of Loneliness and Social
tive perspective’, Journal of Scandinavian Studies in Satisfaction Trajectories’, Child Development, vol. 72, no. 1,
Criminology and Crime Prevention, vol. 12, no. 2, 2011, pp. 2001, pp. 134-151; Craig, W. M., ‘The Relationship among
173-197. Bullying, Victimization, Depression, Anxiety, and Aggression
in Elementary School Children’, Personality and Individual Dif-
30 Ellonen et al., op. cit. ferences, vol. 24, no. 1, 1998, pp. 123-130.
12 Rubin, K. H., W. M. Bukowski and J. G. Parker, ‘Peer Inter-
actions, Relationships, and Groups’, in Handbook of Child
Chapter 6 Psychology, Vol. 3: Social, emotional and personality develop-
ment, 6th ed., edited by N. Eisenberg, Wiley, New York, 2006.
1 See, for example: Hartup, W. W., and N. Stevens, ‘Friend-
ships and Adaptation in the Life Course’, Psychological Bul- 13 Students were given the following definition when asked
letin, vol. 121, no. 3, 1997, pp. 355-370; Rubin, K. H., et al., if they had been physically attacked in the past 12 months:
‘Attachment, Friendship, and Psychosocial Functioning in “A physical attack occurs when one or more people hit or
Early Adolescence’, The Journal of Early Adolescence, vol. strike someone, or when one or more people hurt another
24, no. 4, 2004, pp. 326-356. person with a weapon (such as a stick, knife or gun). It is
not a physical attack when two students of about the same
2 See, for example: Perlman, D., ‘Loneliness: A life span, devel-
strength or power choose to fight each other.” An important
opmental perspective’, in Families and Social Networks,
point to keep in mind when interpreting these data is that the
edited by R. M. Milardo, Sage, Newbury Park, 1988, pp. 190-
perpetrators of physical attacks may also be persons other
220; Steinberg, L., and S. B. Silverberg, ‘The Vicissitudes of
References
than the adolescents’ peers, since students were not asked

Autonomy in Early Adolescence’, Child Development, vol. 57,
to report on only those incidents in which another student
no. 4, 1986, pp. 841-851.
attacked them.
3 Hartup and Stevens, op. cit.
14 The 2012 National School Violence Study consisted of a
4 Hartup and Stevens, op. cit.; Rubin et al., op. cit.; Woodhouse, nationally representative sample of 5,939 secondary school
188
students from a total of 121 schools from each of South Afri- Survey Series, which addresses the health and well-being
ca’s nine provinces. of New Zealand secondary school students. The survey col-
lected data from a nationally representative sample of around
15 Burton, P., and L. Leoschut, School Violence in South Africa:
8,500 students attending composite and secondary schools
Results of the 2012 National School Violence Study, Centre of
in the country. For more information about the survey, see:
Justice and Crime Prevention, Cape Town, 2013.
Clark, T. C., et al., Youth’12 Overview: The health and wellbe-
16 The first wave of the NatSCEV (NatSCEV I) was conducted in ing of New Zealand secondary school students in 2012, The
2008. The second wave (NatSCEV II) was carried out in 2011 University of Auckland, Auckland, 2013.
and consisted of a nationally representative sample of 4,503
26 Adolescent Health Research Group, op. cit.
children aged 1 month to 17 years. Children were randomly
selected for telephone interviews. Participants aged 10 to 17 27 See, for example: Harris, M. B., ‘Sex, Race, and Experiences
years were interviewed about their own experiences while of Aggression’, Aggressive Behavior, vol. 18, no. 3, 1992, pp.
interviews with primary caregivers were conducted for chil- 201-217; Eagly, A. H., and V. J. Steffen, ‘Gender and Aggres-
dren under age 10. See: Finkelhor, D., et al., ‘Violence, Crime, sive Behavior: A meta-analytic review of the social psycho-
and Abuse Exposure in a National Sample of Children and logical literature’, Psychological Bulletin, vol. 100, no. 3, 1986,
Youth: An update’, JAMA-Pediatrics, vol. 167, no. 7, 2013, pp. 309-330.
pp. 614-621.
28 See, for example: Plomin, R., K. Nitz and D. C. Rowe, ‘Behav-
17 Excludes threats, bullying, teasing or emotional bullying and ioral Genetics and Aggressive Behavior in Childhood’, in
Internet harassment. Handbook of Developmental Psychopathology, edited by
18 Finkelhor et al., op. cit. M. Lewis and S. M. Miller, Plenum Press, New York, 1990;
Smuts, B., ‘Male Aggression against Women: An evolution-
19 Pinheiro, P. S., World Report on Violence against Children, ary perspective’, Human Nature, vol. 3, no. 1, 1992, pp. 1-44.
UN Secretary-General’s Study on Violence against Children,
United Nations, Geneva, 2006. 29 See, for example: Harris, M. B., ‘How Provoking! What
makes men and women angry?’, Aggressive Behavior, vol.
20 Laufer, A., and Y. Harel, ‘The Role of Family, Peers and School 19, no. 3, 1993, pp. 199-211; Burbank, V. K., ‘Cross-cultural
Perceptions in Predicting Involvement in Youth Violence’, Perspectives on Aggression in Women and Girls: An introduc-
International Journal of Adolescent Medicine and Health, vol. tion’, Sex Roles, vol. 30, no. 3/4, 1994, pp. 169-176.
15, no. 3, 2003, pp. 235-244; Walsh S. D., et al., ‘Physical and
Emotional Health Problems Experienced by Youth Engaged 30 Ricardo, C., et al., ‘Gender, Sexual Behaviour and Vulnerability
in Physical Fighting and Weapon Carrying’, PLOS ONE, vol. among Young People’, in Promoting Young People’s Sexual
8, no. 2, 2013, p. e56403; Rudatsikira, E., A. S. Muula and S. Health: International perspectives, edited by R. Ingham and
Siziya, ‘Prevalencia e Correlaciones de la Lucha Física entre P. Aggleton, Routledge, London, 2005; Baker, G., ‘Boys in
Adolescentes Escolares en Santiago de Chile, Chile’, Revista the Picture’, 2000, <www.promundo.org.br/wp-content/
Brasileira de Psiquiatria, vol. 30, no. 3, 2008, pp. 197-202. uploads/2009/01/Boys-in-the-picture_2000.pdf>, accessed
22 May 2014.
21 In GSHS and HBSC surveys, adolescents are asked how
many times during the last 12 months they have been 31 Adolescent Health Research Group, op. cit.
involved in a physical fight. Response options range from 32 Currie, C., et al., editors, ‘Social Determinants of Health and
“I have not been in a physical fight in the last 12 months” Well-being Among Young People: Health Behaviour in School-
to “four times or more”. The findings presented here show aged Children (HBSC) Study – International report from the
the proportion that reported fighting at least once in the past 2009/10 survey’, Health Policy for Children and Adolescents,
12 months. In the GSHS, students were given the following no. 6, WHO Regional Office for Europe, Copenhagen, 2012.
definition when asked if they had been involved in a physical
fight in the past 12 months: “A physical fight occurs when 33 US Centers for Disease Control and Prevention, op. cit.
two students of about the same strength or power choose 34 Adolescent Health Research Group, op. cit.
to fight each other.”
35 Pepler, D. J., and W. M. Craig, Making a Difference in Bully-
22 Data on children’s involvement in physical fights are available ing, LaMarsh Centre for Research on Violence and Conflict
for the United States from other nationally representative Resolution, York University, Toronto, 2000.
sources and are generally consistent with the rate that was
recorded by the 2009/2010 HBSC (that is, 35 per cent among 36 Olweus, D., ‘Bully/Victim Problems among School Children:
children aged 13 to 15). For example, the 2013 Youth Risk Some basic facts and effects of a school-based intervention
Behavior Survey (YRBS) found that 25 per cent of students in program’, in The Development and Treatment of Childhood
grades 9 to 12 reported having been in a physical fight at least Aggression, edited by D. Pepler and K. H. Rubin, Erlbaum,
once during the previous 12 months, and 8 per cent said they Hillsdale, 1991; Pepler, D. J., and W. M. Craig, ‘A Peek Behind
had been in such a fight on school property (US Centers for the Fence: Naturalistic observations of aggressive children
Disease Control and Prevention, ‘Youth Risk Behavior Surveil- with remote audiovisual recording’, Developmental Psychol-
lance: United States 2013’, Morbidity and Mortality Weekly ogy, vol. 31, no. 4, 1995, pp. 548-553.
Report, Surveillance Summaries, vol. 63, no. 4, Atlanta, 13
37 David-Ferdon, C., and M. F. Hertz, ‘Electronic Media, Vio-
June 2013). For more information about the YRBS, see:
lence, and Adolescents: An emerging public health problem’,
<www.cdc.gov/healthyyouth/yrbs/index.htm>, accessed 19
Journal of Adolescent Health, vol. 41, no. 6 (Suppl. 1), 2007,
June 2014.
pp. S1-S5.
23 Adolescent Health Research Group, The Health and Well-
38 Only a brief overview of the available literature on risk factors
being of New Zealand Secondary School Students in 2012:
References
and impacts of bullying is provided here.

Youth’12 prevalence tables, The University of Auckland,
Auckland, 2013. 39 Shields, A., and D. Cicchetti, ‘Parental Maltreatment and
Emotion Dysregulation as Risk Factors for Bullying and Vic-
24 US Centers for Disease Control and Prevention, op. cit.
timization in Middle Childhood’, Journal of Clinical Child &
25 The 2012 New Zealand Youth Survey is part of the Youth 2000 Adolescent Psychology, vol. 30, no. 3, 2001, pp. 349-363.
189
40 Laeheem, K., M. Kuning and N. McNeil, ‘Bullying: Risk fac- 54 Ttofi and Farrington, op. cit.; Van der Wal, M. F., C. A. M. De
tors becoming “bullies”’, Asian Social Science, vol. 5, no. 5, Wit and R. A. Hirasing, ‘Psychosocial Health among Young
2009, pp. 50-57. Victims and Offenders of Direct and Indirect Bullying’, Pediat-
rics, vol. 111, no. 6, 2003, pp. 1312-1317; Kaltiala-Heino, R.,
41 Farrington, D., and A. Baldry, ‘Individual Risk Factors for
et al., ‘Bullying, Depression, and Suicidal Ideation in Finnish
School Bullying’, Journal of Aggression, Conflict and Peace
Adolescents: School survey’, BMJ, vol. 319, no. 7206, 1999,
Research, vol. 2, no. 1, 2010, pp. 4-16.
pp. 348-351; Glew et al., op. cit.
42 Nansel, T. R., et al., ‘Bullying Behaviors among US Youth:
55 Farrington, D. P., et al., ‘Bullying Perpetration and Victim-
Prevalence and association with psychosocial adjustment’,
ization as Predictors of Delinquency and Depression in the
JAMA, vol. 285, no. 16, 2001, pp. 2094-2100.
Pittsburgh Youth Study’, Journal of Aggression, Conflict and
43 Currie et al., op. cit.; Farrington and Baldry, op. cit. Peace Research, vol. 3, no. 2, 2011, pp. 74-81.
44 Farrington and Baldry, op. cit.; Wang, J., R. J. Iannotti and T. 56 Nansel et al., 2001, op. cit.
R. Nansel, ‘School Bullying among Adolescents in the United
57 Nansel, T. R., et al., ‘Relationships Between Bullying and Vio-
States: Physical, verbal, relational, and cyber’, Journal of Ado-
lence among US Youth’, Archives of Pediatrics & Adolescent
lescent Health, vol. 45, no. 4, 2009, pp. 368-375; Rigby, K.,
Medicine, vol. 157, no. 4, 2003, pp. 348-353.
‘Why Do Some Children Bully at School? The contributions of
negative attitudes towards victims and the perceived expec- 58 Cross, D., et al., Australian Covert Bullying Prevalence Study
tations of friends, parents and teachers’, School Psychology (ACBPS), Child Health Promotion Research Centre, Edith
International, vol. 26, no. 2, 2005, pp. 147-161. Cowan University, Perth, 2009.
45 Farrington and Baldry, op. cit. 59 Cyberbullying Research Center, <http://cyberbullying.us/
about-us/>, accessed 6 April 2014.
46 Nansel et al., 2001, op. cit.; Monks, C. P., et al., ‘Bullying in
Different Contexts: Commonalities, differences and the role 60 United Nations Children’s Fund, Child Safety Online: Global
of theory’, Aggression and Violent Behavior, vol. 14, no. 2, challenges and strategies, UNICEF Innocenti Research Cen-
2009, pp. 146-156. tre, Florence, 2011.
47 Monks et al., op. cit.; Kouwenberg, M., et al., ‘Peer Victim- 61 Bennett, D. C., et al., ‘College Students’ Electronic Victim-
ization Experienced by Children and Adolescents Who are ization in Friendships and Dating Relationships: Anticipated
Deaf or Hard of Hearing’, PLOS ONE, vol. 7, no. 12, 2012, distress and associations with risky behaviors’, Violence and
p. e52174; Monks, C. P., R. Ortega-Ruiz and A. J. Rodríguez- Victims, vol. 26, no. 4, 2011, pp. 410-429.
Hidalgo, ‘Peer-victimization in Multicultural Schools in Spain
62 Pinheiro, op. cit.
and England’, European Journal of Developmental Psychol-
ogy, vol. 5, no. 4, 2008, pp. 507-535. 63 United Nations Children’s Fund, op. cit.
48 Warwick, I., et al., Homophobia, Sexual Orientation and 64 Cross et al., op. cit.
Schools: A review and implications for action, Research
65 Cappadocia, M. C., W. M. Craig and D. Pepler, ‘Cyberbullying
report RR594, Department for Education and Skills, Univer-
Prevalence, Stability, and Risk Factors During Adolescence’,
sity of London, London, 2004.
Canadian Journal of School Psychology, vol. 28, no. 2, 2013,
49 Patterson, G., ‘The Bully as Victim?’, Paediatric Nursing, vol. pp. 171-192.
17, no. 10, 2005, pp. 27-30.
66 Livingstone, S., et al., EU Kids Online II: Final report 2011, EU
50 Gobina, I., et al. ‘Bullying and Subjective Health among Ado- Kids Online, London School of Economics & Political Science,
lescents at Schools in Latvia and Lithuania’, International London, 2011.
Journal of Public Health, vol. 53, no. 5, 2008, pp. 272-276;
67 Austria, Belgium, Bulgaria, Cyprus, the Czech Republic, Den-
Ttofi, M. M., and D. P. Farrington, ‘Bullying: Short-term and
mark, Estonia, Finland, France, Germany, Greece, Hungary,
long-term effects, and the importance of defiance theory
Ireland, Italy, Lithuania, the Netherlands, Norway, Poland,
in explanation and prevention’, Victims and Offenders, vol.
Portugal, Romania, Slovenia, Spain, Sweden, Turkey and the
3, no. 2-3, 2008, pp. 289-312; Craig, W. M., ‘The Relation-
United Kingdom.
ship among Bullying, Victimization, Depression, Anxiety, and
Aggression in Elementary School Children’, Personality and 68 Burton and Leoschut, op. cit.
Individual Differences, vol. 24, no. 1, 1998, pp. 123-130.
69 US Centers for Disease Control and Prevention, op. cit.
51 Striegel-Moore, R. H., et al., ‘Abuse, Bullying, and Discrimi-
70 Finkelhor et al., op. cit.
nation as Risk Factors for Binge Eating Disorder’, American
Journal of Psychiatry, vol. 159, no. 11, 2002, pp. 1902-1907. 71 US Department of Justice, Bureau of Justice Statistics,
School Crime Supplement (SCS) to the National Crime Vic-
52 Olweus, D., ‘Bullying at School: Basic facts and effects of a
timization Survey, 2011, as cited in Robers, S., J. Kemp and
school based intervention program’, Journal of Child Psychol-
J. Truman, Indicators of School Crime and Safety: 2012,
ogy and Psychiatry and Allied Disciplines, vol. 35, no. 7, 1994,
National Center for Education Statistics, US Department of
pp. 1171–1190; Cook, C. R., et al., ‘Predictors of Bullying and
Education and Bureau of Justice Statistics, Office of Justice
Victimization in Childhood and Adolescence: A meta-analytic
Programs, US Department of Justice, Washington, DC, 2013.
investigation’, School Psychology Quarterly, vol. 25, no. 2,
2010, pp. 65-83; Lawson, M. A., et al., ‘Analyzing Sub-popu- 72 Brazilian Institute of Geography and Statistics (IBGE), ‘Pes-
lation Profiles and Risk Factors for School Bullying’, Children quisa Nacional de Saúde do Escolar 2012’. For more informa-
and Youth Services Review, vol. 35, no. 6, 2013, pp. 973-983; tion on the survey, see: <www.ibge.gov.br/home/estatistica/
References
Glew, G. M., et al., ‘Bullying and School Safety’, The Journal populacao/pense/2012/default.shtm>, accessed 6 April 2014.
of Pediatrics, vol. 152, no. 1, 2008, pp. 123-128.
73 For more information on the Second Regional Comparative and
53 Bond, L., et al., ‘Does Bullying Cause Emotional Problems? Explanatory Study, see: <www.unesco.org/new/en/santiago/
A prospective study of young teenagers’, BMJ, vol. 323, no. education/education-assessment/second-regional-compara-
7311, 2001, pp. 480-484; Ttofi and Farrington, op. cit. tive-and-explanatory-study-serce/>, accessed 6 April 2014.
190
74 United Nations Educational, Scientific and Cultural Orga- they are discriminated against because they are female.”
nization (UNESCO) Regional Bureau for Education in Latin See: <www.unicef.org/protection/57929_58001.html>,
America and the Caribbean, Executive Summary: Student accessed 31 July 2014.
achievement in Latin America and the Caribbean – Results
2 Hong Le, M. T., et al., ‘Early Marriage and Intimate Partner
of the Second Regional Comparative and Explanatory Study
Violence among Adolescents and Young Adults in Viet Nam’,
(SERCE), OREALC/UNESCO, Santiago, 2008.
Journal of Interpersonal Violence, vol. 29, no. 5, 2014, pp.
75 Cross et al., op. cit. 889-910.
76 Defined as being picked on by being chased or grabbed or 3 United Nations Children’s Fund, A Study on Violence against
by being made to do something the respondent did not want Girls, UNICEF Innocenti Research Centre, Florence, 2009.
to do.
4 United Nations Children’s Fund, Child Marriage in India: An
77 Defined as being called names, having mean things said analysis of available data, UNICEF, New Delhi, 2012; Field, E.,
to them or being told by others that they were not wanted and A. Attila, ‘Early Marriage, Age of Menarche, and Female
around. Schooling Attainment in Bangladesh’, Journal of Political
78 US Department of Justice, Bureau of Justice Statistics, op. cit. Economy, vol. 116, no. 5, 2008, pp. 881-930.
79 Finkelhor et al., op. cit. 5 United Nations Children’s Fund, The State of the World’s Chil-
dren 2014: In numbers – Every child counts, UNICEF, New
80 United Nations Children’s Fund, Eliminating Discrimination York, 2014.
against Children and Parents Based on Sexual Orientation
and/or Gender Identity, UNICEF Position Paper, no. 9, May 6 United Nations Children’s Fund, The State of the World’s Chil-
2014 (unpublished). dren 2011: Adolescence – An age of opportunity, UNICEF,
New York, 2011.
7 United Nations Children’s Fund, Early Marriage: Child
82 The acronym LGBT is used here as shorthand for children spouses, UNICEF Innocenti Research Centre, Florence, 2001.
with primary or significant same-sex attraction and/or non-
traditional gender identities. 8 Santhya, K. G., et al., ‘Consent and Coercion: Examining
unwanted sex among married young women in India’, Inter-
83 United Nations Children’s Fund, 2014, op. cit. national Family Planning Perspectives, vol. 33, no. 3, 2007,
84 The National School Climate Survey was based on a nationally pp. 124-132.
representative sample of 8,584 LGBT students between the 9 Oshiro, A., et al., ‘Intimate Partner Violence among General
ages of 13 and 20 years. See: Kosciw, J. G., et al., The 2011 and Urban Poor Populations in Kathmandu, Nepal’, Journal
National School Climate Survey: The experiences of lesbian, of Interpersonal Violence, vol. 26, no. 10, 2011, pp. 2073-
gay, bisexual and transgender youth in our nation’s schools, 2092; Speizer, I. S., and E. Pearson, ‘Association between
the Gay, Lesbian and Straight Education Network (GLSEN), Early Marriage and Intimate Partner Violence in India: A focus
New York, 2012. on youth from Bihar and Rajasthan’, Journal of Interpersonal
85 However, the odds of planning a suicide attempt among les- Violence, vol. 26, no. 10, 2011, pp. 1963-1981; Raj, A., et
bians and of “medically serious attempts” among gay males al., ‘Association between Adolescent Marriage and Marital
were no longer significantly different than heterosexual peers Violence among Young Adult Women in India’, International
after controlling for other risk and demographic factors. See: Journal of Gynecology and Obstetrics, vol. 110, no. 1, 2010,
Stone, D. M., ‘Sexual Orientation and Suicide Ideation, Plans, pp. 35-39.
Attempts, and Medically Serious Attempts: Evidence from 10 Rahman, M., et al., ‘Association between Adolescent Mar-
local Youth Risk Behavior Surveys, 2001-2009’, American riage and Intimate Partner Violence: A study of young adult
Journal of Public Health, vol. 104, no. 2, 2014, pp. 262-271. women in Bangladesh’, Asia-Pacific Journal of Public Health,
86 See, for example: Wichstrøm, L., and K. Hegna, ‘Sexual Ori- vol. 26, no. 2, 2014, pp. 160-168.
entation and Suicide Attempt: A longitudinal study of the Nor- 11 Hong Le et al., op. cit.
wegian adolescent population’, Journal of Abnormal Psychol-
ogy, vol. 112, no. 1, 2003, pp. 144-151; Van Bergen, D. D., et 12 Erulkar, A., ‘Early Marriage, Marital Relations and Intimate
al., ‘Victimization and Suicidality among Dutch Lesbian, Gay, Partner Violence in Ethiopia’, International Perspectives on
and Bisexual Youth’, American Journal of Public Health, vol. Sexual and Reproductive Health, vol. 39, no. 1, 2013, pp. 6-13.
103, no. 1, 2013, pp. 70-72. 13 US Centers for Disease Control and Prevention, ‘Understand-
87 Coker T. R., S. B. Austin and M. A. Schuster, ‘The Health ing Teen Dating Violence: Fact sheet’, 2014, <www.cdc.gov/
and Health Care of Lesbian, Gay, and Bisexual Adolescents’, violenceprevention/pdf/teen-dating-violence-2014-a.pdf>,
Annual Review of Public Health, vol. 31, 2010, pp. 457-477. accessed 13 May 2014.
14 Children’s Safety Network, ‘Teen Dating Violence as a Pub-
lic Health Issue’, 2012, <www.childrenssafetynetwork.org/
Chapter 7 sites/childrenssafetynetwork.org/files/TeenDatingViolen-
ceasaPublicHealthIssue.pdf>, accessed 22 May 2014.
1 Kishor, S., and K. Johnson, Profiling Domestic Violence: A
15 US Centers for Disease Control and Prevention, 2014, op.
multi-country study, ORC Macro, Calverton, 2004. According
cit.; Smith, P. H., J. W. White and L. J. Holland, ‘A Longitu-
to UNICEF, “Gender-based violence (GBV) is a term used for
dinal Perspective on Dating Violence among Adolescent and
describing harmful acts perpetrated against a person based
College-age Women’, American Journal of Public Health, vol.
on socially ascribed differences between females and males.
References
93, no. 7, 2003, pp. 1104-1109.

While the broadest interpretation of GBV is sometimes
understood to include specific types of violence against boys 16 Hickman, L. J., L. H. Jaycox and J. Aronoff, ‘Dating Violence
and men, the term has historically been and continues to be among Adolescents: Prevalence, gender distribution, and pre-
used primarily as a way to highlight the vulnerabilities of girls vention program effectiveness’, Trauma Violence & Abuse,
and women to various forms of violence in settings where vol. 5, no. 2, 2004, pp. 123-142; Davis, A., Interpersonal and
191
Physical Dating Violence among Teens, National Council on Negative Emotion in Women Experiencing Intimate Partner
Crime and Delinquency, Oakland, 2008. Violence: Shame, guilt, and PTSD’, Behavior Therapy, vol. 42,
no. 4, 2011, pp. 740-750; Becker, K. D., J. Stuewig and L. A.
17 Halpern, C. T., et al., ‘Partner Violence among Adolescents
McCloskey, ‘Traumatic Stress Symptoms of Women Exposed
in Opposite-sex Romantic Relationships: Findings from the
to Different Forms of Childhood Victimization and Intimate
National Longitudinal Study of Adolescent Health’, American
Partner Violence’, Journal of Interpersonal Violence, vol. 25,
Journal of Public Health, vol. 91, no. 10, 2001, pp. 1679-1685.
no. 9, 2010, pp. 1699-1715.
18 Vagi, K. J., et al., ‘Beyond Correlates: A review of risk and pro-
21 Foshee, V. A., et al., ‘A Longitudinal Examination of Psy-
tective factors for adolescent dating violence perpetration’,
chological, Behavioral, Academic, and Relationship
Journal of Youth and Adolescence, vol. 42, no. 4, 2013, pp.
Consequences of Dating Abuse Victimization among
633-649.
a Primarily Rural Sample of Adolescents’, Journal of
19 Martin, S. L., et al., ‘Pregnancy-associated Violent Deaths: Adolescent Health, vol. 53, no. 6, 2013, pp. 723-729;
The role of intimate partner violence’, Trauma Violence & Exner-Cortens, D., J. Eckenrode and E. Rothman,
Abuse, vol. 8, no. 2, 2007, pp. 135-148; Campbell, J. C., ‘Longitudinal Associations between Teen Dating Violence
‘Health Consequences of Intimate Partner Violence’, The Victimization and Adverse Health Outcomes’, Pediatrics, vol.
Lancet, vol. 359, no. 9314, 2002, pp. 1331-1336; Abrahams, 131, no. 1, 2013, pp. 71-78.
N., et al., ‘Mortality of Women from Intimate Partner Violence
22 Maharaj, P., and C. Munthree, ‘Coerced First Sexual Inter-
in South Africa: A national epidemiological study’, Violence
course and Selected Reproductive Health Outcomes among
and Victims, vol. 24, no. 4, 2009, pp. 546-556; Devries, K., et Young Women in KwaZulu-Natal, South Africa’, Journal of
al., ‘Violence against Women is Strongly Associated with Sui- Biosocial Science, vol. 39, no. 2, 2007, p. 231-244; Coker, A.
cide Attempts: Evidence from the WHO multi-country study L., ‘Does Physical Intimate Partner Violence Affect Sexual
on women’s health and domestic violence against women’, Health? A systematic review’, Trauma Violence & Abuse, vol.
Social Science and Medicine, vol. 73, no. 1, 2011, pp. 79-86; 8, no. 2, 2007, pp. 149-177.
Vizcarra, B., et al., ‘Partner Violence as a Risk Factor for Men-
tal Health among Women from Communities in the Philip- 23 Wolfe, D. A., et al., ‘The Effects of Children’s Exposure to
pines, Egypt, Chile, and India’, Injury Control and Safety Pro- Domestic Violence: A meta-analysis and critique’, Clinical
motion, vol. 11, no. 2, 2004, pp. 125-129; Ellsberg, M., et al., Child and Family Psychology Review, vol. 6, no. 3, 2003, pp.
‘Intimate Partner Violence and Women’s Physical and Mental 171-187.
Health in the WHO Multi-country Study on Women’s Health 24 Levendosky, A. A., and S. A. Graham-Bermann,
and Domestic Violence: An observational study’, The Lancet, ‘Parenting in Battered Women: The effects of domestic vio-
vol. 371, no. 9619, 2008, pp. 1165-1172; Garcia-Moreno, C., lence on women and their children’, Journal of Family Vio-
et al., ‘Prevalence of Intimate Partner Violence: Findings from lence, vol. 16, no. 2, 2001, pp. 171-192.
the WHO multi-country study on women’s health and domes-
tic violence’, The Lancet, vol. 368, no. 9543, 2006, pp. 1260- 25 Grych, J. H., et al., ‘Patterns of Adjustment among Children of
1269; Coker, A. L., et al., ‘Physical and Mental Health Effects Battered Women’, Journal of Consulting and Clinical Psychol-
of Intimate Partner Violence for Men and Women’, Ameri- ogy, vol. 68, no. 1, 2000, pp. 84-94.
can Journal of Preventive Medicine, vol. 23, no. 4, 2002, pp. 26 Holt, S., H. Buckley and S. Whelan, ‘The Impact of Expo-
260-268; Coker, A. L., et al., ‘Physical Health Consequences sure to Domestic Violence on Children and Young People: A
of Physical and Psychological Intimate Partner Violence’, review of the literature’, Child Abuse & Neglect, vol. 32, no. 8,
Archives of Family Medicine, vol. 9, no. 5, 2000, pp. 451-457. 2008, pp. 797-810; Stewart, L. A., and J. Power, ‘Profile and
20 Wong, J. Y. H., et al., ‘Depression among Women Experi- Programming Needs of Federal Offenders with Histories of
encing Intimate Partner Violence in a Chinese Community’, Intimate Partner Violence’, Journal of Interpersonal Violence,
Nursing Research, vol. 60, no. 1, 2011, pp. 58-65; Chandra, 2014, pp. 168-188.
P. S., V. A. Satyanarayana and M. P. Carey, ‘Women Report- 27 Hong Le et al., op. cit.; Whitfield, C. L., et al., ‘Violent Child-
ing Intimate Partner Violence in India: Associations with PTSD hood Experiences and the Risk of Intimate Partner Violence in
and depressive symptoms’, Archives of Women’s Mental Adults: Assessment in a large health maintenance organiza-
Health, vol. 12, no. 4, 2009, pp. 203-209; Nerøien, A. I., and tion’, Journal of Interpersonal Violence, vol. 18, no. 2, 2003,
B. Schei, ‘Partner Violence and Health: Results from the first pp. 166-185; Speizer and Pearson, op. cit.
national study on violence against women in Norway’, Scan-
dinavian Journal of Public Health, vol. 36, no. 2, 2008, pp. 28 The domestic violence module was also used in the Tuvalu
161-168; Pico-Alfonso, M. A., et al., ‘The Impact of Physical, DHS 2007, but results for all forms of partner violence are
Psychological, and Sexual Intimate Male Partner Violence on based on less than 25 unweighted cases and therefore are
Women’s Mental Health: Depressive symptoms, posttrau- not included here.
matic stress disorder, state anxiety, and suicide’, Journal of 29 The expression ‘ever-married’ is used in this chapter to refer
Women’s Health, vol. 15, no. 5, 2006, pp. 599-611; Savas, N., to boys and girls who have ever been married or lived with a
and G. Agridag, ‘The Relationship between Women’s Mental partner as if married.
Health and Domestic Violence in Semirural Areas: A study in
30 Questions on emotional violence perpetrated by a husband or
Turkey’, Asia-Pacific Journal of Public Health, vol. 23, no. 3,
partner were not asked in the Bangladesh DHS 2007.
2011, pp. 399-407; Vos, T., et al., ‘Measuring the Impact of
Intimate Partner Violence on the Health of Women in Victoria, 31 Questions on sexual violence perpetrated by a husband or
Australia’, Bulletin of the World Health Organization, vol. 84, partner were not asked in the Pakistan DHS 2012-2013.
no. 9, 2006, pp. 739-744; Dillon, G., et al., ‘Mental and Physi-
32 Centro de Estudios de Población y Desarrollo Social (CEPAR),
cal Health and Intimate Partner Violence against Women: A
References
Encuesta Demográfica y de Salud Materna e Infantil 2004,

review of the literature’, International Journal of Family Medi-
Quito, 2005.
cine, 2013, Article ID 313909; Avdibegovi, E., and O. Sinanovi,
‘Consequences of Domestic Violence on Women’s Mental 33 La Asociación Demográfica Salvadoreña (ADS), Encuesta
Health in Bosnia and Herzegovina’, Croatian Medical Journal, Nacional de Salud Familiar (FESAL-2008): Informe final, San
vol. 47, no. 5, 2006, pp. 730-741; Beck, J. G., et al., ‘Exploring Salvador, 2009.
192
34 Ministerio de Salud Pública y Asistencia Social (MSPAS)/Insti- 40 Institute of Public Health (IPH), Albania Ministry of Health
tuto Nacional de Estadística (INE)/Centros de Control y Pre- and the Institute of Statistics (INSTAT), Albania Reproductive
vención de Enfermedades (CDC), Encuesta Nacional de Salud Health Survey 2002, Tirana, 2005.
Materno Infantil 2008-2009, Guatemala, 2011.
41 National Center for Disease Control and Public Health (NCDC),
35 Instituto Nacional de Información de Desarrollo (INIDE) and Ministry of Labour, Health, and Social Affairs (MoLHSA) and
Ministerio de Salud (MINSA), Encuesta Nicaragüense de the National Statistics Office of Georgia, Reproductive Health
Demografía y Salud 2011/12: Informe preliminar, Managua, Survey Georgia 2010, Tbilisi, 2012.
2013.
42 Ministry of Health (Romania), World Bank, United Nations
36 Centro Paraguayo de Estudios de Población (CEPEP), Population Fund, United States Agency for International
Encuesta Nacional de Demografía y Salud Sexual y Reproduc- Development and United Nations Children’s Fund, Reproduc-
tiva 2008: Informe final, Asunción, 2009. tive Health Survey Romania 2004: Summary report, Buzau,
2005.
37 Serbanescu, F., A. Ruiz and D. B. Suchdev, Reproductive
Health Survey Jamaica 2008: Final report, US Centers for Dis- 43 In these surveys, partner physical violence included the same
ease Control and Prevention, Atlanta, 2010. acts as those measured in the DHS, except that in Albania it
did not include pulling her hair, dragging her, burning her or
38 Instituto Nacional de Estadística y Geografía (INEGI),
attacking her with a weapon, and in Romania it did not include
Encuesta Nacional sobre la Dinámica de las Relaciones en los
shaking her or twisting her arm. Partner sexual violence was
Hogares (2011): Panorama de violencia contra las mujeres en
defined as physically forced sexual intercourse and partner
Estados Unidos Mexicanos, Aguascalientes, 2013, as cited
emotional violence included insulting her or swearing at her
in Tovar, J., ‘Measuring Violence against Women: The case
and threatening to hurt her or someone she cares about.
of Mexico’, presentation given at the International Seminar
on Gender Statistics, Incheon, Republic of Korea, 13 Novem- 44 Physical violence was defined as slapping or throw-
ber 2013, <http://unstats.un.org/unsd/demographic/meetings/ ing something at her that could hurt her; pushing her,
wshops/Korea/2013/list_of_docs.htm>, accessed 10 June shoving her or pulling her hair; hitting her with a fist or some-
2014. thing else that could hurt her; kicking her, dragging her or
beating her up; choking or burning her; and threatening to
39 In all of these countries, except Ecuador and Mexico, physical
use or actually using a gun, knife or other weapon against her.
violence included the same acts as those included in the DHS
Girls and women were considered to have been subjected to
module on domestic violence. The survey in Ecuador did not
sexual violence if they were physically forced into sexual inter-
include questions on whether girls and women were choked
course, had sexual intercourse when they did not want to for
or burned, while in Mexico, physical violence included any of
fear of what their partner would do or were forced to do some-
the following: pushed her or pulled her hair; tied her; kicked
thing sexual that they found humiliating or degrading. Emo-
her; threw objects at her; hit her with the hands or objects;
tional violence included insulting or cursing at her, belittling
tried to strangle or hang her; attacked her with a weapon; and
her or humiliating her in front of others, scaring or threatening
shot her. In El Salvador, Guatemala, Nicaragua and Paraguay,
her and threatening to hurt her or someone she cares about.
the definition of sexual violence used was: being physically
ICON-Institut Public Sector, Hacettepe Institute of Population
forced to have sexual intercourse, forced to perform other
Studies and BNB Consulting, National Research on Domestic
sexual acts and agreeing to have sexual intercourse for fear
Violence against Women in Turkey, Ankara, 2009.
of what the partner might do. In Ecuador and Jamaica, sexual
violence was defined only as forced sexual intercourse. In 45 Ministry of Gender and Family (Maldives), The Maldives
Mexico, sexual violence included demanding or forcing a girl Study on Women’s Health and Life Experiences: Initial results
or woman to have sexual intercourse or to perform sexual on prevalence, health outcomes and women’s responses to
acts against her will. In addition to the forms of emotional violence, Male, 2006.
violence covered in the DHS module, the surveys conducted
46 Secretariat of the Pacific Community, The Samoa Family
in El Salvador, Nicaragua and Paraguay included doing things
Health and Safety Study, Noumea, 2006.
on purpose to frighten or intimidate her; the survey in Gua-
temala also included this form of emotional violence as well 47 Vanuatu Women’s Centre, The Vanuatu National Survey on
as threats to take away the child(ren). The Jamaica survey Women’s Lives and Family Relationships, Port Vila, 2011.
did not include a question on whether the partner humili-
48 Ministry of Health, General Statistics Office (Viet Nam), World
ated her in front of others. The Ecuador survey included the
Health Organization and United Nations Children’s Fund, Sur-
same acts of emotional violence as those in the DHS module.
vey Assessment of Vietnamese Youth Round 2 (SAVY 2),
The Mexico survey defined emotional violence as including:
Hanoi, 2010.
shaming, belittling or humiliating her; ignoring her or not pro-
viding affection; making her fearful; threatening to leave or 49 In the Maldives, Samoa and Vanuatu surveys, physical vio-
damage or hide her property; threatening to take away the lence was defined as being slapped or having something
child(ren) or financial resources; threatening to kill her or the thrown at her that could hurt her; pushed, shoved or had her
child(ren); threatening her with a weapon; locking her inside hair pulled or cut; hit with a fist or something else that could
the house or forbidding her to see anyone; turning child(ren) hurt her; kicked, dragged or beaten up; choked or burned;
or relatives against her; spying on or monitoring her; bar- and being threatened with or having a weapon used against
ring her from working or studying; comparing her with other her. Sexual violence was defined as physically forced sexual
women; and getting angry at her because food is not made intercourse, having sexual intercourse when she did not want
or housework is not done. In Mexico, information was also to for fear of what her partner would do, and being forced to
collected on the following forms of economic violence com- do something sexual that she found humiliating or degrad-
References
mitted by a partner: asks how she spends money; is restric- ing. Emotional violence included any of the following acts:
tive with giving her money, including for household expenses, insulted her or made her feel bad about herself; belittled or
even if he has money; fails to make money or threatens not humiliated her in front of others; scared or intimidated her on
to give her money; spends money needed for the home; and purpose; and threatened to hurt her or someone she cares
has taken over or removed money or property. about.
193
50 Data on partner emotional violence are not available by age (UCR2) Survey.
groups for Vanuatu, but the lifetime prevalence rate among all
61 Sexual interference is defined as the direct or indirect touch-
girls and women aged 15 to 49 who have ever been married
ing of a person under the age of 14 using a body part or an
is 68 per cent.
object for sexual purposes.
51 Physical abuse included: having something thrown at her that
62 Nocentini, A., et al., ‘Physical Dating Aggression in Adoles-
could hurt; twisting her arm or pulling her hair; attacks that
cence: Cultural and gender invariance’, European Psycholo-
resulted in physical injuries; pushing her strongly; attacking
gist, vol. 16, no. 4, 2011, pp. 278-287.
her with a weapon; a beating on the head that results in a
coma; hitting her with an object such as a belt or stick; stran-
gling or attempting to strangle her; grabbing her strongly;
slapping her face; breaking her bones; and burning her on pur- Chapter 8
pose. Sexual abuse included a husband refusing to allow her
to use contraceptives; physically forcing her to have sexual 1 Heise, L., and C. García-Moreno, ‘Violence by Intimate Part-
intercourse; forcing her in different ways to engage in sexual ners’, in World Report on Violence, edited by E. G. Krug et
acts she is not happy with; and threatening to force sexual al., World Health Organization, Geneva, 2002; García-Moreno,
intercourse or other sexual acts she is not happy with. Psy- C., et al., WHO Multi-country Study on Women’s Health and
chological abuse included being cursed at or insulted; having Domestic Violence against Women, World Health Organiza-
something thrown at her that could hurt; having her property tion, Geneva, 2005.
destroyed or ‘spoiled’; being yelled or shouted at; and say- 2 United Nations Children’s Fund Kenya Country Office, Divi-
ing things in order to provoke or upset her. The survey also sion of Violence Prevention, National Center for Injury Preven-
collected information on two additional forms of spousal vio- tion and Control, US Centers for Disease Control and Preven-
lence: social abuse, which included being forced to quit work tion and Kenya National Bureau of Statistics, Violence against
and being prohibited from meeting friends or going out with Children in Kenya: Findings from a 2010 national survey –
female neighbours; and economic abuse, which was defined Summary report on the prevalence of sexual, physical and
as a husband refusing to give her enough money for house- emotional violence, context of sexual violence, and health
hold expenses (even though he has enough money to spend and behavioural consequences of violence experienced in
on other things), asking in detail about how money is spent, childhood, Nairobi, 2012.
withdrawing money from accounts or credit cards without
her permission, controlling her belongings, prohibiting her 3 Zimbabwe National Statistics Agency, United Nations
from working and trying to exploit her legacy without her Children’s Fund and Collaborating Centre for Operational
permission. Palestinian Central Bureau of Statistics (PCBS), Research and Evaluation, National Baseline Survey on Life
Violence Survey in the Palestinian Society 2011: Main find- Experiences of Adolescents, 2011, Harare, 2013.
ings, Ramallah, 2012. 4 In reading these data, it is important to keep in mind that posi-
52 Data may have been collected in other countries that con- tive attitudes towards wife-beating should not necessarily be
ducted a DHS but are not presented in the final DHS reports interpreted as approval of wife-beating and do not imply that
and therefore are not included here. a woman or girl will inevitably become a victim of domestic
violence.
53 Although these data are based on all girls and women
between the ages of 15 and 49, the general patterns are 5 Data coverage was insufficient to calculate regional aver-
expected to hold for adolescent girls. ages for the Middle East and North Africa, East Asia and the
Pacific, Latin America and the Caribbean and CEE/CIS.
54 The domestic violence module was also used in the Ghana
DHS 2008 and the Uganda DHS 2011, but results for all forms 6 Head, S. K., et al., Women’s Lives and Challenges: Equality
of partner violence are based on less than 25 unweighted and empowerment since 2000, ICF International, Rockville,
cases and therefore are not included here. 2014.
55 This survey was representative of only selected micro-urban 7 Kishor, S., and K. Johnson, 2004, Profiling Domestic Violence
areas with a population of at least 100,000. Coverage was – A Multi-Country Study, ORC Macro, Calverton, 2004.
estimated to reach roughly 88 per cent of the national pop-
8 Head et al., op. cit.
ulation aged 16 to 65. Blima SchraiberI, L., et al., ‘Intimate
Partner Sexual Violence among Men and Women in Urban 9 The standard child discipline module used in MICS and in a
Brazil, 2005’, Revista de Saude Publica, vol. 42, no. 1, 2008, number of DHS and other nationally representative house-
pp. 127-137. hold surveys includes a question about whether respondents
believe that physical punishment is necessary in order to
56 Instituto Nacional de Información de Desarrollo (INIDE) and
properly raise/educate children. Prior to 2010, the child dis-
Ministerio de Salud (MINSA), op. cit.
cipline module was administered only to mothers or primary
57 This information was also collected in the Ghana DHS 2008 caregivers of children aged 2 to 14. Beginning with MICS4
and Uganda DHS 2011, but results are based on less than 25 (2009-2011), the methodology was changed so that any adult
unweighted cases and therefore are not included here. household member, not just the mother or primary caregiver,
can respond to questions about child discipline.
58 These data are based on 25-49 unweighted cases.
10 Nobes, G., et al., ‘Physical Punishment by Mothers and
59 US Centers for Disease Control and Prevention, ‘Youth Risk
Fathers in British Homes’, Journal of Interpersonal Violence,
Behavior Surveillance: United States 2013’, Morbidity and
vol. 14, no. 8, 1999, pp. 887-902; Power, T. G., et al., ‘Compli-
Mortality Weekly Report, Surveillance Summaries, vol. 63,
ance and Self-assertion: Young children’s responses to moth-
References
no. 4, Atlanta, 13 June 2013.

ers versus fathers’, Developmental Psychology, vol. 30, no.
60 Mahony, T. H., ‘Police-reported Dating Violence in Canada 6, 1994, pp. 980-989; Regalado, M., et al., ‘Parents’ Disci-
2008’, Juristat, vol. 30, no. 2, Summer 2010, pp. 1-26. Data pline of Young Children: Results from the National Survey of
in this article are from Statistics Canada, Canadian Centre for Early Childhood Health, Pediatrics, vol. 113, no. 6, 2004, pp.
Justice Statistics, Incident-based Uniform Crime Reporting 1952-1958.
194
11 Rydstrøm, H., ‘“Like a White Piece of Paper”: Embodiment children had left home. Department for Children, Schools
and the moral upbringing of Vietnamese children’, Ethnos: and Families, Review of Section 58 of the Children Act 2004,
Journal of Anthropology, vol. 66, no. 3, 2001, pp. 394-413; DCSF, London, 2007.
Rydstrøm, H., ‘Masculinity and Punishment: Men’s upbring-
22 National Society for the Prevention of Cruelty to Children,
ing of boys in rural Vietnam’, Childhood, vol. 13, no. 3, 2006,
Public Attitudes to the NSPCC Full Stop Campaign and a
pp. 329-348.
Range of Children’s Issues in Northern Ireland, NSPCC, Bel-
12 Mahoney, A., et al., ‘Mother and Father Self-reports of Cor- fast, 2000.
poral Punishment and Severe Physical Aggression toward
23 Data were also collected in the Republic of Moldova, the
Clinic-referred Youth’, Journal of Clinical Child Psychology,
former Yugoslav Republic of Macedonia and Ukraine but are
vol. 29, no. 2, 2000, pp. 266-281; Lansford, J. E., et al., ‘Cor-
not presented here since comparable data are available from
poral Punishment of Children in Nine Countries as a Function
MICS and reported elsewhere in this chapter.
of Child Gender and Parent Gender’, International Journal of
Pediatrics, vol. 2010, 2010, Article ID 672780, pp. 1-12. 24 Data were collected from nationally representative samples
in each country. The sample consisted of 994 young adults
13 Lansford et al., op. cit.
and adults aged 15 and older in Bulgaria; 1,015 adults aged 18
14 Data were collected via phone interviews from a nationally to 74 in Latvia; 500 children and adults aged 15 to 74 in Lithu-
representative sample of 720 parents. Tucci, J., J. Mitchell ania; and 955 adults aged 18 and older in Poland. Sajkowska,
and C. R. Goddard, Crossing the Line: Making the case for M., The Problem of Child Abuse: Attitudes and experiences
changing Australian laws about the physical punishment of in seven countries of Central and Eastern Europe, Nobody’s
children, Australian Childhood Foundation, Ringwood, 2006. Children Foundation, Institute of Applied Social Sciences,
15 Data were collected from a nationally representative sample Warsaw University, Warsaw, 2006.
of 1,650 children from 6 of the 15 regions in Chile. United 25 National Center for Injury Prevention and Control, US Centers
Nations Children’s Fund, Final Report: Fourth Study on Child for Disease Control and Prevention, Understanding Sexual
Maltreatment and Family Relations, Department of Sociology, Violence: Fact sheet, 2007, <www.cdc.gov/ncipc/pub-res/
Pontificia Universidad Católica de Chile, Santiago, 2012. images/sv%20factsheet.pdf >, accessed 17 April 2014.
16 Data were also collected in Jordan but are not presented here 26 Data were collected from a total representative sample of
since comparable data are available from DHS and reported 841 women and men from Anguilla, Barbados, Dominica,
elsewhere in this chapter. Families were recruited from public
Grenada, Montserrat, and Saint Kitts and Nevis. The result-
and private schools serving socially and economically diverse
ing sampling distribution was found to be biased towards
populations in selected cities within each of the participating
women, which was explained by the higher rates of refusal
countries. Although the samples were not nationally repre-
among men.
sentative, families from high-, middle- and low-income house-
holds were included in the approximate proportion in which 27 Jones, A. D., and E. Trotman Jemmott, Child Sexual Abuse
these income groups represented the local population. Lans- in the Eastern Caribbean, United Nations Children’s Fund,
ford et al., op. cit. Action for Children and University of Huddersfield, Hudder-
sfield, 2009.
17 Sariola, H., Violence Against Children and Child Sexual Abuse
in Finland, presentation at the Central Union for Child Wel- 28 Data were collected from a sample of 357 parents and care-
fare, Helsinki, 30 August 2012; Central Union for Child Wel- givers. Ige, O. K., and O. I. Fawole, ‘Preventing Child Sexual
fare, Attitudes to Disciplinary Violence, Finland, 2012, Cen- Abuse: Parents’ perceptions and practices in urban Nige-
tral Union for Child Welfare and Taloustutkimus Oy, Helsinki, ria’, Journal of Child Sexual Abuse, vol. 20, no. 6, 2011, pp.
2012. 695-707.
18 Data were collected via a telephone survey from a nation- 29 Data were collected from a sample of 652 parents of third-
ally representative sample of 750 adults aged 18 and over. grade students from seven schools in Jingzhou city. Chen, J.,
Wood, B., Physical Punishment of Children in New Zealand: M. P. Dunne and P. Han, ‘Prevention of Child Sexual Abuse
Six years after law reform, EPOCH New Zealand, Wellington, in China: Knowledge, attitudes and communication practices
2013. of parents of elementary school children’, Child Abuse &
19 Bunting, L., M. A. Webb and J. Healy, ‘In Two Minds? Paren- Neglect, vol. 31, no. 7, 2007, pp. 747-755.
tal attitudes toward physical punishment in the UK’, Children 30 Chen, Dunne and Han, op. cit.
& Society, vol. 24, no. 5, 2010, pp. 359-370; Anderson S., L.
31 Ige and Fawole, op. cit.
Murray and J. Brownlie, Disciplining Children: Research with
parents in Scotland, Scottish Executive Central Research 32 Data were collected from a representative sample of 296
Unit, Edinburgh, 2002. adults aged 18 to 64. Tennfjord, O. S., ‘Prediction of Attitudes
Towards Child Sexual Abuse among Three Different Norwe-
20 This study involved a nationally representative sample of
gian Samples’, Journal of Sexual Aggression, vol. 12, no. 3,
1,250 mothers and fathers of children aged 0 to 12. Ghate,
D., et al., The National Study of Parents, Children and Disci- 2006, pp. 245-263.
pline in Britain: Key findings, Economic and Social Research 33 Chen, Dunne and Han, op. cit.
References
Council, Swindon, 2003.

34 Tennfjord, op. cit.
21 Nationally representative samples of 1,822 parents across
35 Ige and Fawole, op. cit.
England and Wales participated in the study, including both
parents with children who still lived at home and those whose 36 Chen, Dunne and Han, op. cit.
195
statistical table
Experience of Lifetime
Homicide rate physical violence experience of Intimate partner Attitudes towards
(per 100,000) Violent discipline at home Violence among peers (%) since age 15 (%) sexual violence violence (%) wife-beating (%)
2012 (%) 2005-2013* 2003-2013* 2005-2013* (%) 2004-2013* 2005-2013* 2002-2013*
Attitudes
Psychological
towards
punishment
Any violent
aggression
physical
discipline
Physical
Physical
Physical
Bullying
punishment
attacks
Female
Female
Female
Female
Female
fights
Male
Male
Male
Male
Male
(%)
Total
Countries and areas 2005-2013*
Afghanistan 11 4 8 74 69 62 – – – – – – – – – – 84 41
Albania 3 2 2 77 61 71 – – – – – – – – – 37 24 13
Algeria 2 1 2 88 75 84 52 48 – – – – – – – – 66 15
Andorra – – – – – – – – – – – – – – – – – –
Angola 7 4 5 – – – – – – – – – – – – – – –
Antigua and Barbuda – – – – – – 25 48 – – – – – – – – – –
Argentina 5 1 3 72 46 65 25 34 25 – – – – – – – 2 4
Armenia 1 0 1 70 43 66 10 51 – – – – – – – 21 8 3
Australia 1 1 1 – – – – – – – – – – – – – – –
Austria 0 0 0 – – – 41 37 – – – – – – – – – –
Azerbaijan 4 1 3 77 51 74 – – – – 9 – 0.4 – 14 63 39 22
Bahamas 7 2 5 – – – – – – – – – – – – – – –
Bahrain 1 0 1 – – – – – – – – – – – – – – –
Bangladesh 1 2 1 – – – – – – – – – – – 47 y – 33 y –
Barbados 3 0 3 75 56 62 13 38 – – – – – – – – 5 36
Belarus 0 0 1 65 y 34 y 59 y – – – – – – – – – 3 3 8
Belgium 2 0 1 – – – 37 37 – – – – – – – – – –
Belize 10 1 6 71 57 54 31 36 – – – – – – – – 11 26
Benin 5 4 5 – – – 42 32 – – – – – – – 16 15 –
Bhutan 1 1 1 – – – – – – – – – – – – – 70 –
Bolivia (Plurinational
9 5 7 – – – 30 33 – 11 y 37 y 1y 5y (62) y 48 y – 17 –
State of)
Bosnia and Herzegovina 0 0 0 55 40 42 – – – – – – – – – 5 1 14
Botswana 9 5 7 – – – 52 48 56 – – – – – – – – –
Brazil 31 3 17 – – – – – – – – – – – – – – –
Brunei Darussalam 0 0 0 – – – – – – – – – – – – – – –
Bulgaria 1 0 1 – – – – – – – – – – – – – – –
Burkina Faso 5 4 5 83 58 79 – – – – 21 – – – 8 40 39 37
Burundi 5 4 5 – – – – – – – – – – – – 56 74 –
Cabo Verde 1 0 1 – – – – – – – 16 – – – (19) 24 23 –
Cambodia 4 3 4 – – – 22 14 – – 21 y – 0.3 – 19 24 42 –
Cameroon 6 3 5 93 78 87 – – – 41 45 2 22 – 54 43 50 43
Canada 2 1 2 – – – 32 34 – – – – – – – – – –
Central African Republic 8 2 5 92 81 84 – – – – – – – – – 83 79 31
Chad 5 4 4 84 77 71 – – – – – – – – – – 59 38
Chile 3 1 2 – – – 15 29 – – – – – – – – – –
China 1 1 1 – – – 29 y 19 y – – – – – – – – – –
Colombia 22 3 13 – – – 32 y 33 y 26 y – 13 y – 3y – 34 y – – –
Comoros 5 3 5 – – – – – – – 11 – 5 – 9 17 43 –
Congo 8 3 5 87 69 80 – – – – – – – – – 76 y 73 y 22
Cook Islands – – – – – – 33 41 – – – – – – – – – –
Costa Rica 8 2 5 46 30 31 19 22 – – – – – – – – 3 12
Côte d'Ivoire 10 7 9 91 73 88 – – – – 33 – 5y – 24 y 51 51 39
Croatia 0 1 0 – – – 16 39 – – – – – – – – – –
Cuba 3 0 2 – – – – – – – – – – – – – – –
Cyprus 1 0 0 – – – – – – – – – – – – – – –
Czech Republic 0 0 0 – – – 16 43 – – – – – – – – – –
statistical table
Democratic People's
4 2 3 – – – – – – – – – – – – – – –
Republic of Korea
Democratic Republic of
20 8 14 92 80 82 – – – – 56 – 21 y – 70 – 72 32
the Congo
Denmark 1 0 0 – – – 17 27 – – – – – – – – – –
Djibouti 6 4 5 72 67 57 41 60 56 – – – – – – – – 32
Dominica – – – – – – 27 39 – – – – – – – – – –
196
Experience of Lifetime Attitudes

Homicide rate physical violence experience of Intimate partner Attitudes towards towards
(per 100,000) Violent discipline at home Violence among peers (%) since age 15 (%) sexual violence violence (%) wife-beating (%) physical
2012 (%) 2005-2013* 2003-2013* 2005-2013* (%) 2004-2013* 2005-2013* 2002-2013* punishment
(%)
Psychological
2005-2013*
punishment
Any violent
aggression
discipline
Physical
Physical
Physical
Bullying
attacks
Female
Female
Female
Female
Female
fights
Male
Male
Male
Male
Male
Total
Countries and areas
Dominican Republic 5 2 4 67 45 y 50 – – – – 13 – 8 – 33 14 7 8y
Ecuador 10 2 6 – – – 29 y 34 y 36 y – – – 6y – – – – –
Egypt 2 1 1 91 82 83 70 45 59 – 40 y – – – 21 – 50 y 42
El Salvador 42 11 27 – – – – – – – – – 7y – – – – –
Equatorial Guinea 12 4 9 – – – – – – – 42 – 17 – (73) 56 57 –
Eritrea 3 2 3 – – – – – – – – – – – – 60 51 –
Estonia 1 2 1 – – – 37 28 – – – – – – – – – –
Ethiopia 9 4 6 – – – – – – – – – – – – 51 64 –
Fiji 2 2 2 72 y – – 42 47 – – – – – – – – – –
Finland 0 1 1 – – – 28 28 – – – – – – – – – –
France 0 0 0 – – – 32 33 – – – – – – – – – –
Gabon 6 2 4 – – – – – – – 39 – 14 – 58 47 58 –
Gambia 7 5 6 90 74 81 – – – – – – – – – – 74 39
Georgia 1 0 1 67 50 59 – – – – – – – – – – 5 13
Germany 0 1 0 – – – 30 20 – – – – – – – – – –
Ghana 5 5 5 94 73 89 62 53 60 34 32 – 17 – 35 37 53 50
Greece 1 0 0 – – – 29 51 – – – – – – – – – –
Grenada – – – – – – 27 38 41 – – – – – – – – –
Guatemala 32 12 22 – – – – – – – – – 3y – – – – –
Guinea 6 5 5 – – – – – – – – – – – – 63 89 –
Guinea-Bissau 6 5 6 82 74 68 – – – – – – – – – – 39 y 25
Guyana 8 3 6 76 63 66 38 38 – – – – – – – 25 18 23
Haiti 27 11 19 85 79 64 – – – – 27 – 10 – 43 22 24 30
Holy See – – – – – – – – – – – – – – – – – –
Honduras 15 11 13 – – – 32 28 – – 15 – 6 – 28 y 18 15 –
Hungary 1 0 1 – – – 24 41 – – – – – – – – – –
Iceland 0 0 0 – – – 16 28 – – – – – – – – – –
India 3 1 2 – – – – – – – 21 – 5 – 34 47 45 –
Indonesia 1 0 1 – – – 50 34 40 – – – – – – 48 y 45 –
Iran (Islamic Republic of) 2 1 2 – – – – – – – – – – – – – – –
Iraq 6 2 4 79 63 75 28 37 – – – – – – – – 50 22
Ireland 0 0 0 – – – 26 35 – – – – – – – – – –
Israel 1 0 1 – – – – – – – – – – – – – – –
Italy 0 0 0 – – – 9 35 – – – – – – – – – –
Jamaica 18 7 13 85 68 72 40 50 – – – – 11 y – – 28 y 8 27
Japan 0 0 0 – – – – – – – – – – – – – – –
Jordan 4 2 3 90 67 88 41 47 38 – 31 y – – – 26 – 84 y 23
Kazakhstan 2 1 1 49 29 43 – – – – 4 – 1 – 23 14 9 7
Kenya 3 2 3 – – – 57 48 – – 27 – 11 – 34 54 57 –
Kiribati – – – 81 y – – 37 35 – – – – – – – 65 77 –
Kuwait 1 0 0 – – – 28 45 33 – – – – – – – – –
Kyrgyzstan 1 0 0 54 y 37 y 43 y – – – – 6 – 0.1 – 3 40 25 8
Lao People's Democratic
4 2 3 76 44 71 – – – – – – – – – 50 56 42
Republic
Latvia 0 1 1 – – – 43 39 – – – – – – – – – –
Lebanon 1 0 1 82 56 y 80 25 49 – – – – – – – – 22 y 24 y
Lesotho 23 14 18 – – – – – – – – – – – – 54 48 –
statistical table
Liberia 4 3 4 90 76 84 – – – – 39 – 13 – 46 37 48 61
Libya 1 0 1 – – – 35 40 30 – – – – – – – – –
Liechtenstein – – – – – – – – – – – – – – – – – –
Lithuania 1 1 1 – – – 52 38 – – – – – – – – – –
Luxembourg 0 0 0 – – – 28 32 – – – – – – – – – –
Madagascar 9 7 8 – – – – – – – – – – – – 44 47 –
197
2012 (%) 2005-2013* 2003-2013* 2005-2013* (%) 2004-2013* 2005-2013* 2002-2013* punishment
(%)
Psychological
2005-2013*
punishment
Any violent
aggression
discipline
Physical
Physical
Physical
Bullying
attacks
Female
Female
Female
Female
Female
fights
Male
Male
Male
Male
Male
Total
Countries and areas
Malawi 2 1 1 – – – 45 23 – – 21 – 18 – 40 21 16 –
Malaysia 1 0 1 – – – 21 30 – – – – – – – – – –
Maldives 1 0 1 – – – 37 30 – – – – – – – – 41 y –
Mali 5 5 5 – – – – – – – – – – – 14 – 83 –
Malta 0 0 0 – – – – – – – – – – – – – – –
Marshall Islands – – – – – – – – – – (35) – – – (42) 71 47 –
Mauritania 4 3 3 87 78 82 47 58 – – – – – – – – 36 36
Mauritius 1 1 1 – – – 36 36 27 – – – – – – – – –
Mexico 12 2 7 – – – – – – – – – – – – – – –
Micronesia (Federated
– – – – – – – – – – – – – – – – – –
States of)
Monaco – – – – – – – – – – – – – – – – – –
Mongolia 3 1 2 46 25 38 31 44 – – – – – – – 9 14 16
Montenegro 1 0 1 63 45 56 – – – – – – – – – – 6 5
Morocco 2 1 1 91 67 y 89 19 43 – – – – – – – – 64 41 y
Mozambique 8 5 7 – – – – – – 28 22 3 9 30 37 20 24 –
Myanmar 5 8 7 – – – 19 15 21 – – – – – – – – –
Namibia 9 4 6 – – – 52 50 – – 32 – – – – 44 38 –
Nauru – – – – – – 39 45 – – – – – – – – – –
Nepal 2 1 1 – – – – – – – 10 – 11 – 23 27 24 –
Netherlands 1 0 0 – – – 21 31 – – – – – – – – – –
New Zealand 1 1 1 – – – – – – – – – – – – – – –
Nicaragua 6 2 4 – – – – – – – – – 7y – – – 19 y –
Niger 4 3 3 82 66 77 – – – – – – – – – 41 54 45
Nigeria 15 12 14 91 79 81 – – – – 27 – 6 – 15 25 33 62
Niue – – – – – – – 33 – – – – – – – – – –
Norway 0 0 0 – – – 23 – – – – – – – – – – –
Oman 1 0 0 – – – 47 50 – – – – – – – – – –
Pakistan 4 3 4 – – – 41 37 – – 30 y – – – 28 y (33) y 53 y –
Palau – – – – – – – – – – – – – – – – – –
Panama 26 4 15 45 y 29 y 33 y – – – – – – – – – – – 8
Papua New Guinea 4 2 3 – – – – – – – – – – – – – – –
Paraguay 12 2 7 – – – – – – – – – 2y – – – – –
Peru 2 1 2 – – – 47 37 – – 14 y – – – 27 y – – –
Philippines 6 1 3 – – – 48 38 – – 15 – 5y – 31 y – 15 –
Poland 1 0 1 – – – 24 32 – – – – – – – – – –
Portugal 0 0 0 – – – 37 27 – – – – – – – – – –
Qatar 1 0 0 – – – 42 51 – – – – – – – – – –
Republic of Korea 1 1 1 – – – – – – – – – – – – – – –
Republic of Moldova 1 1 1 76 48 69 – – – – 14 – 7 – 28 14 13 15
Romania 1 1 1 – – – 41 41 – – – – – – – – – –
Russian Federation 2 1 2 – – – 33 37 – – – – – – – – – –
Rwanda 14 11 13 – – – – – – – 16 – 12 – 43 y 35 56 –
Saint Kitts and Nevis – – – – – – 23 38 – – – – – – – – – –
Saint Lucia – – – 68 44 60 25 41 34 – – – – – – – 15 21
Saint Vincent and the
– – – – – – 30 46 39 – – – – – – – – –
Grenadines
Samoa – – – – – – 74 68 – – – – – – – 50 58 –
statistical table
San Marino – – – – – – – – – – – – – – – – – –
Sao Tome and Principe – – – – – – – – – – 26 – 9 – 37 25 23 –
Saudi Arabia 1 1 1 – – – – – – – – – – – – – – –
Senegal 2 1 1 – – – – – – – – – – – – 31 61 –
Serbia 1 0 1 67 37 60 – – – – – – – – – 6 2 7
Seychelles – – – – – – 51 – – – – – – – – – – –
198

2012 (%) 2005-2013* 2003-2013* 2005-2013* (%) 2004-2013* 2005-2013* 2002-2013* punishment
(%)
Psychological
2005-2013*
punishment
Any violent
aggression
discipline
Physical
Physical
Physical
Bullying
attacks
Female
Female
Female
Female
Female
fights
Male
Male
Male
Male
Male
Total
Countries and areas
Sierra Leone 11 8 9 82 65 74 – – – – – – – – – 57 63 43
Singapore 0 1 1 – – – – – – – – – – – – – – –
Slovakia 0 0 0 – – – 25 40 – – – – – – – – – –
Slovenia 0 0 0 – – – 20 40 – – – – – – – – – –
Solomon Islands 3 1 3 72 y – – 67 53 – – – – – – – 73 72 –
Somalia 6 4 5 – – – – – – – – – – – – – 75 y –
South Africa 12 4 8 – – – – – – – – – – – – – – –
South Sudan 9 6 8 – – – – – – – – – – – – – 72 –
Spain 0 0 0 – – – 14 35 – – – – – – – – – –
Sri Lanka 2 1 1 – – – 38 47 48 – – – – – – – 54 y –
State of Palestine – – – 93 76 90 56 y 44 y – – – – – – – – – 21
Sudan 7 5 6 – – – – – – – – – – – – – 52 –
Suriname 0 0 0 86 60 82 26 21 – – – – – – – – 19 13
Swaziland 22 10 16 89 66 82 32 19 – – – – – – – 34 42 82
Sweden 0 0 0 – – – 11 30 – – – – – – – – – –
Switzerland 0 1 0 – – – 34 – – – – – – – – – – –
Syrian Arab Republic 2 1 1 89 78 84 – – – – – – – – – – – 13
Tajikistan 2 1 1 78 60 73 7 22 25 – 7 – 1 – 16 – 47 15
Thailand 3 1 2 – – – 27 34 33 – – – – – – – 10 –
The former Yugoslav
0 0 0 69 52 56 20 31 20 – – – – – – – 14 3
Timor-Leste 2 2 2 – – – – – – – 30 – 2 – 31 72 81 –
Togo 8 5 7 93 77 86 – – – – – – – – – – 41 35
Tonga – – – – – – 50 49 – – – – – – – – – –
Trinidad and Tobago 20 5 12 77 54 68 15 36 41 – – – – – – – 10 25
Tunisia 1 0 1 93 74 90 31 47 46 – – – – – – – 27 44
Turkey 3 1 2 – – – – – – – – – – – – – 30 y –
Turkmenistan 5 2 4 – – – – – – – – – – – – – – –
Tuvalu – – – – – – – – – – – – – – – 83 69 –
Uganda 11 8 10 – – – 46 36 – 54 54 7 19 – 46 52 62 –
Ukraine 1 1 1 61 30 57 41 39 – 41 y 6 – 0.3 2y (2) 2 2 11
United Arab Emirates 2 1 2 – – – 23 47 – – – – – – – – – –
United Kingdom 1 0 0 – – – 25 y 33 y – – – – – – – – – –
United Republic of
7 4 6 – – – 28 y 40 y 56 y – 24 – 13 – 46 39 52 –
Tanzania
United States 6 2 4 – – – 26 35 – – – – – – – – – –
Uruguay 3 1 2 – – – 19 26 19 – – – – – – – – –
Uzbekistan 2 1 1 – – – – – – – – – – – – 63 63 –
Vanuatu – – – 84 72 77 67 51 – – – – – – – 63 56 51
Venezuela (Bolivarian
37 2 20 – – – 35 y 30 y – – – – – – – – – –
Republic of)
Viet Nam 1 1 1 74 55 55 26 22 – – – – – – – – 35 17
Yemen 2 1 2 95 86 92 41 56 51 – – – – – – – – 44
Zambia 7 5 6 – – – 65 53 – – 33 – 16 – 43 55 61 –
Zimbabwe 6 3 5 – – – 61 y 41 y – – 23 – 18 – 58 48 48 –
# For a complete list of countries and areas in the regions and sub-regions and for country compared for the Dominican Republic, Jamaica and Sierra Leone since data sources for girls
categories, see page 200. are more recent than those for boys. WHO Global Health Estimates are available for WHO
Member States whose population were over 250,000 in 2012. Zeroes appearing in the table
statistical table
– Data not available.

do not necessarily mean that there were no homicide victims in these countries, but rather
( ) Figures are based on 25-49 unweighted cases. that the recalculated homicide rates came to 0 after rounding.
y Data differ from the standard definition or refer to only part of a country. Source: UNICEF global databases, 2014, based on the Health Behaviour in School-aged Chil-
* Data refer to the most recent year available during the period specified in the column dren Study (HBSC), Demographic and Health Surveys (DHS), Global School-based Student
heading. Health Surveys (GSHS), Multiple Indicator Cluster Surveys (MICS), and other national sur-
veys. Data on homicide rates are from: World Health Organization, Global Health Estimates
Notes: Data on attitudes towards wife-beating among boys and girls cannot be directly (GHE) Summary Tables: Deaths by cause, age, sex and region, 2012, WHO, Geneva, 2014.
199
Indicator definitions Regional classification
Homicide rate: Number of homicide victims among children ‘Countries outside of these regions’ includes places in which
and adolescents aged 0 to 19 years per 100,000 population. UNICEF does not have programmes. They are primarily high-
and high-middle-income countries located in Australasia, Eu-
Any violent discipline: Percentage of children aged 2 to 14
rope and North America.
years who experienced any violent discipline (psychological
aggression and/or physical punishment) in the past month.
Sub-Saharan Africa: Eastern and Southern Africa; West and
Physical punishment: Percentage of children aged 2 to Central Africa; Djibouti; Sudan
14 years who experienced physical punishment in the past Eastern and Southern Africa: Angola; Botswana; Burundi;
month. Comoros; Eritrea; Ethiopia; Kenya; Lesotho; Madagascar;
Malawi; Mauritius; Mozambique; Namibia; Rwanda;
Psychological aggression: Percentage of children aged 2 Seychelles; Somalia; South Africa; South Sudan; Swaziland;
to 14 years who experienced psychological aggression in the Uganda; United Republic of Tanzania; Zambia; Zimbabwe
past month.
West and Central Africa: Benin; Burkina Faso; Cabo Verde;
Bullying: Percentage of adolescents aged 13 to 15 years Cameroon; Central African Republic; Chad; Congo; Côte
who reported being bullied at least once in the past couple d’Ivoire; Democratic Republic of the Congo; Equatorial Guinea;
of months. Gabon; Gambia; Ghana; Guinea; Guinea-Bissau; Liberia; Mali;
Mauritania; Niger; Nigeria; Sao Tome and Principe; Senegal;
Physical fights: Percentage of adolescents aged 13 to 15 Sierra Leone; Togo
years who reported being in a physical fight one or more
times during the past 12 months. Middle East and North Africa: Algeria; Bahrain; Djibouti;
Egypt; Iran (Islamic Republic of); Iraq; Jordan; Kuwait;
Physical attacks: Percentage of students aged 13 to 15 years Lebanon; Libya; Morocco; Oman; Qatar; Saudi Arabia; State
who reported being physically attacked one or more times in of Palestine; Sudan; Syrian Arab Republic; Tunisia; United
the past 12 months. Arab Emirates; Yemen
Experience of physical violence since age 15: Percentage South Asia: Afghanistan; Bangladesh; Bhutan; India;
of girls and boys aged 15 to 19 years who experienced any Maldives; Nepal; Pakistan; Sri Lanka
physical violence since age 15. East Asia and the Pacific: Brunei Darussalam; Cambodia;
China; Cook Islands; Democratic People’s Republic of Korea;
Lifetime experience of sexual violence: Percentage of girls Fiji; Indonesia; Kiribati; Lao People’s Democratic Republic;
and boys aged 15 to 19 years who ever experienced forced Malaysia; Marshall Islands; Micronesia (Federated States of);
sexual intercourse or any other forced sexual acts (including Mongolia; Myanmar; Nauru; Niue; Palau; Papua New Guinea;
in childhood). Philippines; Republic of Korea; Samoa; Singapore; Solomon
Islands; Thailand; Timor-Leste; Tonga; Tuvalu; Vanuatu; Viet
Intimate partner violence: Percentage of ever-married girls
Nam
and boys aged 15 to 19 years who ever experienced any
physical, sexual or emotional violence committed by their Latin America and the Caribbean: Antigua and Barbuda;
spouses or partners. Argentina; Bahamas; Barbados; Belize; Bolivia (Plurinational
State of); Brazil; Chile; Colombia; Costa Rica; Cuba; Dominica;
Attitudes towards wife-beating: Percentage of girls and Dominican Republic; Ecuador; El Salvador; Grenada;
boys aged 15 to 19 years who think that a husband/partner Guatemala; Guyana; Haiti; Honduras; Jamaica; Mexico;
is justified in hitting or beating his wife or partner for at least Nicaragua; Panama; Paraguay; Peru; Saint Kitts and Nevis;
one of the specified reasons, i.e., if his wife burns the food, Saint Lucia; Saint Vincent and the Grenadines; Suriname;
argues with him, goes out without telling him, neglects the Trinidad and Tobago; Uruguay; Venezuela (Bolivarian Republic
children or refuses sexual relations. of)
Attitudes towards physical punishment: Percentage of CEE/CIS: Albania; Armenia; Azerbaijan; Belarus; Bosnia
adults who think that physical punishment is necessary to and Herzegovina; Bulgaria; Croatia; Georgia; Kazakhstan;
raise/educate children. Kyrgyzstan; Montenegro; Republic of Moldova; Romania;
Russian Federation; Serbia; Tajikistan; The former Yugoslav
Republic of Macedonia; Turkey; Turkmenistan; Ukraine;
Uzbekistan
Countries outside of these regions: Andorra; Australia;
Austria; Belgium; Canada; Cyprus; Czech Republic; Denmark;
Estonia; Finland; France; Germany; Greece; Hungary; Iceland;
Ireland; Israel; Italy; Japan; Latvia; Liechtenstein; Lithuania;
statistical table
Luxembourg; Malta; Monaco; Netherlands; New Zealand;

Norway; Poland; Portugal; San Marino; Slovakia; Slovenia;
Spain; Sweden; Switzerland; United Kingdom; United States
200
201
Ten facts about
UNICEF
Data and Analytics Section
Division of Data, Research
and Policy
violence against 3 United Nations Plaza

New York, NY 10017, USA
Tel: +1 212 326 7000
Email: [email protected]
children ISBN: 978-92-806-4767-9
1 In 2012 alone, homicide took the lives of about 95,000 children and
adolescents under the age of 20 – almost 1 in 5 of all homicide
victims that year.
2 Around 6 in 10 children between the ages of 2 and 14 worldwide

(almost a billion) are subjected to physical punishment by their
caregivers on a regular basis.
3 Close to 1 in 3 students between the ages of 13 and 15 worldwide

report involvement in one or more physical fights in the past year.
4 Slightly more than 1 in 3 students between the ages of 13 and 15

worldwide experience bullying on a regular basis.
5 About 1 in 3 adolescents aged 11 to 15 in Europe and North

America admit to having bullied others at school at least once
in the past couple of months.
6 Almost one quarter of girls aged 15 to 19 worldwide

(almost 70 million) report being victims of some form of physical
violence since age 15.
7 Around 120 million girls under the age of 20 (about 1 in 10)

have been subjected to forced sexual intercourse or other forced
sexual acts at some point in their lives. Boys are also at risk, although
a global estimate is unavailable due to the lack of comparable data in
most countries.
8 1 in 3 adolescent girls aged 15 to 19 worldwide (84 million)

have been the victims of any emotional, physical or sexual violence
committed by their husbands or partners at some point in their lives.
9 About 3 in 10 adults worldwide believe that physical punishment

is necessary to properly raise or educate children.
10 Close to half of all girls aged 15 to 19 worldwide (around

126 million) think a husband is sometimes justified in hitting or
beating his wife.
Note: Estimates are based on a subset of countries with available data covering 50 per cent or more of the global population of children
or adults within the respective age ranges.
Sources: UNICEF global databases, 2014, based on Demographic and Health Surveys (DHS), Global School-based Student Health Surveys
(GSHS), Health Behaviour in School-aged Children (HBSC) study, Multiple Indicator Cluster Surveys (MICS), other national surveys,
and relevant studies. Population data are from: United Nations, Department of Economic and Social Affairs, Population Division, World
Population Prospects: The 2012 revision, CD-ROM edition, United Nations, New York, 2013.

Hidden in Plain Sight: A Statistical Analysis of Violence Against Children (2014)

Uploaded by

Copyright:

Available Formats

Hidden in Plain Sight: A Statistical Analysis of Violence Against Children (2014)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hidden in Plain Sight: A Statistical Analysis of Violence Against Children (2014)

Uploaded by

Copyright:

Available Formats

HIDDEN IN

© United Nations Children’s Fund

Permission is required to reproduce

To request permission and for any

All reasonable precautions have

Suggested citation: United Nations

02 Using data to make violence visible . . . . . . 14

03 Physical violence: A leading cause of

04 Sexual violence: Not limited to girls . . . . 60

05 The most common form of violence

06 When children hurt one another. . . . . . . . 112

07 Violent unions among adolescents . . . . . . 130

08 Exploring attitudes and social norms . . . 146

09 Key findings and CONCLUSIONS. . . . . . . . . . . . 164

Statistical table . . . . . . . . . . . . . . . . . . . . . . . . 196

mental Violence is often described as psychological maltreatment, mental abuse, verbal

Neglect or negligent treatment means the failure to meet children’s

“All forms of violence against children, however light,

United Nations Committee on the Rights of the Child, General

Over the last decade, recognition of the pervasive

victims do denounce an abuse, the legal system

fails to respond and child protection services are

Each chapter follows a similar structure, beginning

In addition to the possible overlap of various types of

witnesses to violence – all at the same time. Children

who suffer violence are sometimes driven to hurt

“1. States Parties shall take

treatment and follow-up of instances

of child maltreatment described

The social and economic costs of

Prenatal period and birth early childhood

middle childhood Late childhood and adolescence

self- and political awareness and with independent

A full understanding of violence against children

circumstances related to incidents of abuse.

Despite the wide-ranging types of data needed,

CHAPTER 2: USING DATA TO MAKE VIOLENCE VISIBLE

CHAPTER 2: USING DATA TO MAKE VIOLENCE VISIBLE

Measuring child neglect: Many

CHAPTER 2: USING DATA TO MAKE VIOLENCE VISIBLE

The potential impact of children’s

participation in research on violence

CHAPTER 2: USING DATA TO MAKE VIOLENCE VISIBLE

questions about violence against children

CHAPTER 2: USING DATA TO MAKE VIOLENCE VISIBLE

on violence against children

CHAPTER 2: USING DATA TO MAKE VIOLENCE VISIBLE

the confidence intervals are likely to be relatively wide,

confidentiality and safety of all concerned. Moreover,

often cited as perpetrators. Acts of physical force

against children are also common among peers.

Although all children are vulnerable to physical

To measure physical violence committed by a

Preventable deaths Figure 3.1B

sex. As children enter the second decade of their

deaths due to intentional injuries (out of all causes)

Most of these young victims (around 85,000 or 90

largest global share of homicides among individuals

Central and Eastern Europe

Homicide rates among boys are higher than