The economic
dimensions of
interpersonal
violence
WORLD HEALTH
ORGANIZATION
Geneva
The Economic Dimensions
of Interpersonal Violence
DEPARTMENT OF INJURIES AND VIOLENCE PREVENTION
WORLD HEALTH ORGANIZATION
20 AVENUE APPIA
1211 GENEVA 27
SWITZERLAND
WHO Library Cataloguing-in-Publication Data
The economic dimensions of interpersonal violence / Hugh Waters … [et al.].
1.Violence – economics 2.Violence – prevention and control 3.Costs and cost analysis
4.Review literature I.Waters, Hugh.
ISBN 92 4 159160 9 (LC/NLM classification: HV 6625)
© World Health Organization 2004
All rights reserved. Publications of the World Health Organization can be obtained from Marketing and
Dissemination,World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland (tel: +41 22 791 2476 ;
fax: +41 22 791 4857 ; email: [email protected]). Requests for permission to reproduce or translate WHO publications
– whether for sale or for noncommercial distribution – should be addressed to Publications, at the above
address (fax: +41 22 791 4806; email: [email protected]).
The designations employed and the presentation of the material in this publication do not imply the expression of
any opinion whatsoever on the part of the World Health Organization concerning the legal status of any country, territory,
city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on
maps represent approximate border lines for which there may not yet be full agreement.
The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or
recommended by the World Health Organization in preference to others of a similar nature that are not mentioned.
Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters.
The World Health Organization does not warrant that the information contained in this publication is complete and
correct and shall not be liable for any damages incurred as a result of its use.
Printed in France
Editorial Committee and Project Team
Department of International Health, Johns Hopkins Bloomberg School of Public Health
Hugh Waters, MS, PhD
Adnan Hyder, MD, PhD
Yogesh Rajkotia, MSc
Suprotik Basu, MHS
Julian Ann Rehwinkel, MPH
Injuries and Violence Prevention Department,World Health Organization
Alexander Butchart, MA, PhD
Acknowledgements
The input of the following individuals who peer reviewed earlier drafts of this report is
gratefully acknowledged.
David J. Ball, Middlesex University, United Kingdom.
Nancy Cardia, Centre for the Study of Violence, University of São Paulo, Brazil.
Phillip J Cook, Duke University, USA.
Phaedro Corso, Centers for Disease Control and Prevention, USA.
Elizabeth Eckermann, Deakin University, Australia.
Patricia Hernandez,World Health Organization, Geneva.
Pat Mayhew, Australian Institute of Criminology, Australia.
James Mercy, Centers for Disease Control and Prevention, USA.
Staff of the WHO Department of Injuries and Violence Prevention who also provided
comments on earlier drafts of the report includes Dr Andrés Villaveces, Ms Alison Phinney,
Dr David Meddings and Ms Laura Sminkey.
The development and publication of this handbook has been made possible by the generous
financial support of the Government of Sweden
Suggested citation
Waters H, Hyder A, Rajkotia Y, Basu S, Rehwinkel JA, Butchart A. The economic dimensions of
interpersonal violence. Department of Injuries and Violence Prevention, World Health
Organization, Geneva, 2004.
v
Contents
Abbreviatiovii
Foreword ….viii
Summary ……x
1. Introduction …………………………………………………………………………………………………………..1
1.1 Definitions …………………………………………………………………………………………………………2
1.2 An ecological framework for assessing the economic dimensions of interpersonal violence ..4
1.3 Types of costs ……………………………………………………………………………………………………6
1.4 Economic evaluation of interventions ……………………………………………………………………7
1.5 Key methodological issues ………………………………………………………………………………….8
1.6 Search strategy ………………………………………………………………………………………………..10
1.7 Characteristics of included and excluded literature………………………………………………11
2. Costs of interpersonal violence …………………………………………………………………………….13
2.1 Overview of the costs ……………………………………………………………………………………….13
2.2 Child abuse and neglect……………………………………………………………………………………15
2.3 Intimate partner violence……………………………………………………………………………………17
2.4 Sexual violence ………………………………………………………………………………………………..21
2.5 Workplace violence …………………………………………………………………………………………..22
2.6 Youth violence ………………………………………………………………………………………………….24
2.7 Guns, drugs and gangs ……………………………………………………………………………………24
2.8 Effects on public finances………………………………………………………………………………….26
3. Economic effects of interventions to reduce interpersonal violence……………………….28
4. Effects of economic factors and policies on interpersonal violence ……………………….34
4.1 Absolute income levels and interpersonal violence ………………………………………………35
4.2 Employment and social networks ……………………………………………………………………….36
4.3 Income inequalities …………………………………………………………………………………………..37
4.4 Inter-generational effects……………………………………………………………………………………39
4.5 Policy responses ………………………………………………………………………………………………40
4.6 Future directions in exploring the role of economic factors in interpersonal violence ..40
5. Conclusion …………………………………………………………………………………………………………..42
5.1 Methodological issues……………………………………………………………………………………….42
5.2 Gaps in the literature and directions for future research ……………………………………….43
Appendix 1 Types of cost evaluations ……………………………………………………………………45
Appendix 2 Electronic databases and web sites searched ……………………………………..46
Appendix 3 Categories used to abstract literature ………………………………………………….48
References ……………………………………………………………………………………………………….49
Tables and Figures
Box 1 The value of a human life …………………………………………………………………………..9
Figure 1 Ecological model for understanding interpersonal violence …………………………..6
Figure 2 Costs and benefits of interpersonal violence………………………………………………10
Table 1 Costs of social violence in Latin America ………………………………………………….14
Table 2 Costs of child abuse – selected studies……………………………………………………..16
vi
Table 3 Costs of intimate partner violence – selected studies ………………………………….18
Table 4 Costs of sexual violence – selected studies………………………………………………..22
Table 5 Costs of workplace violence – selected studies …………………………………………23
Table 6 Costs of youth violence – selected studies ………………………………………………..24
Table 7 Economic evaluations of interventions to reduce violence – selected studies ..28
Table 8 Selected studies exploring the relationship between economic factors and
violence ………………………………………………………………………………………………..34
vii
Abbreviations
CBA cost-benefit analysis
CDC Centers for Disease Control and Prevention (USA)
CEA cost-effectiveness analysis
CUA cost-utility analysis
DALY disability-adjusted life year
HMO healthcare maintenance organization
IADB Inter-American Development Bank
ILO International Labour Organisation
NCVS National Crime Victimization Survey (USA)
NVAWS National Violence against Women Survey (USA)
QALY quality-adjusted life year
WHO World Health Organization
WRVH World Report on Violence and Health
viii
Foreword
WHO’s World report on violence and health (published in 2002) makes a strong
case for violence prevention. It reviewed available scientific evidence. It showed
the need to work at all levels of the ecological model – with individuals, families,
communities and societies – and to draw upon the contributions of multiple
sectors, such as justice, education, welfare, employment and health. It concluded
that violence prevention is complex, but is possible. The present report, on The
economic dimensions of interpersonal violence, strengthens the case for investing in
prevention even further by highlighting the enormous economic costs of the
consequences of interpersonal violence, and reviewing the limited but
nonetheless striking evidence for the cost-effectiveness of prevention
programmes.
The first section of this report presents an ecological model for assessing the
economic dimensions of interpersonal violence, and addresses some of the
methodological issues around the costing of violence, its consequences and
efforts to prevent it. The second section reviews the available evidence for the
direct and indirect economic costs of child abuse and neglect; intimate partner
violence; sexual assault; workplace violence and youth violence; and the effect on
public finances of selected risk factors for interpersonal violence, including
firearms, alcohol, drugs and gangs. Section three reviews cost-effectiveness
studies of programmes to prevent child abuse and neglect, intimate partner
violence, youth violence and gun violence. The fourth section examines selected
recent studies that explore the relationship between interpersonal violence and
economic factors that can potentially be modified through policy interventions,
including poverty, economic inequality, employment and social networks. The
report concludes by identifying the many gaps in the literature (for instance,
there were no studies of the costs of elder abuse, few from developing countries
and few cost-effectiveness studies), and defining a research agenda for future
studies of the costs of interpersonal violence, which it recommends should be
based upon a standardized methodology that, unlike the currently dominant
human capital approach, allows for the comparison of the costs of interpersonal
violence across countries and different economies.
After reading this report one is left with three key messages. First, that the
consequences of interpersonal violence are extremely costly. Second, that
prevention studies show evidence of cost effectiveness. Third, that for most of
the developing world and many developed countries there is not even descriptive
information about the direct costs of treating the consequences of interpersonal
violence. Together, these messages outline one of the major challenges in the
years ahead, which is to systematically establish a solid base of evidence about
ix
the costs of interpersonal violence in all societies, and then to feed this evidence
into policy-making and advocacy where it can complement and strengthen
moral arguments for the prevention of interpersonal violence. To do any less
would mean that public health has failed to meet its obligations to promote
health and safety for all.
Dr Etienne Krug
Director
Department of Injuries and Violence Prevention
World Health Organization
Geneva, Switzerland
x
Summary
Interpersonal violence is expensive. For instance, estimates of the cost of
violence in the United States of America reach 3.3% of the gross domestic
product. In England and Wales, the total costs from violence – including
homicide, wounding and sexual assault – amount to an estimated $40.2 billion
annually1.
Interpersonal violence is defined to include violence between family members
and intimate partners and violence between acquaintances and strangers that is
not intended to further the aims of any formally defined group or cause. Selfdirected
violence, war, state-sponsored violence and other collective violence are
specifically excluded from these definitions.
This report, based on an extensive review of peer reviewed articles and
published and unpublished reports, treats the following themes:
The economic effects of interpersonal violence in a variety of
socioeconomic and cultural settings.
The economic effects of interventions intended to reduce interpersonal
violence.
The effects of economic conditions and policies on interpersonal
violence – with particular reference to poverty, structural adjustment,
income equality and social investment.
Interpersonal violence disproportionately affects low- and middle-income
countries. The economic effects are also likely to be more severe in poorer
countries. However, as this report shows, there is a scarcity of studies of the
economic effects of this violence in low- and middle-income countries.
Comparisons with high-income countries are complicated by the fact that
economic losses related to productivity tend to be undervalued in low-income
countries since these losses are typically based on foregone wages and income.
For example, a single homicide is calculated to cost, on average, $15 319 in
South Africa, $602 000 in Australia, $829 000 in New Zealand, and more than
$2 million in the USA.
Many of the studies detailing the costs of violence are from the USA where child
abuse results in $94 billion in annual costs to the economy – 1.0% of the gross
domestic product. Direct medical treatment costs per abused child have been
calculated by different studies to range from $13 781 to $42 518. Intimate
partner violence costs the USA economy $12.6 billion on an annual basis – 0.1%
of the gross domestic product – compared to 1.6% of the gross domestic product
in Nicaragua and 2.0% of the gross domestic product in Chile. Gun violence –
which includes suicides – has alone been calculated at $155 billion annually in
1 Throughout this document, to enable comparisons and to adjust for inflation and varying exchange rates,
monetary values have been converted to 2001 US dollars. This was done by converting other currency
amounts to US dollars using the exchange rate at the mid-point of the year of the estimate, then converting
the resulting US dollar estimate to 2001 US dollars using the official US consumer price index. The
exchange rates used are those from international markets, and are not adjusted for purchasing power parity
xi
the USA, with lifetime medical treatment costs per victim ranging from $37 000
to $42 000.
Evidence abounds that the public sector – and thus society in general – bears
much of the economic burden of interpersonal violence. Several studies in the
USA showed that from 56% to 80% of the costs of care for gun and stabbing
injuries are either directly paid by public financing or are not paid at all – in
which case they are absorbed by the government and society in the form of
uncompensated care financing and overall higher payment rates. In low- and
middle-income countries, it is also probable that society absorbs much of the
costs of violence through direct public expenditures and negative effects on
investment and economic growth.
There are relatively few published economic evaluations of interventions
targeting interpersonal violence. Available studies showed that preventive
interventions to stop interpersonal violence occurring cost less than the money
that they save, in some cases by several orders of magnitude. The 1994 Violence
Against Women Act in the USA has resulted in an estimated net benefit of $16.4
billion, including $14.8 billion in averted victim’s costs. A separate analysis
showed that providing shelters for victims of domestic violence resulted in a
benefit to cost ratio between 6.8 and 18.4. Similarly, the costs of a programme
to prevent child abuse through counselling equalled 5.0% of the costs of child
abuse itself. Implementation of a gun registration law in Canada cost $70
million, in comparison with a total annual cost of $5.6 billion for firearm-related
injuries in that country. Interventions that targeted juvenile offenders – including
aggression replacement training and foster care treatment – resulted in economic
benefits that were more than 30 times greater than the corresponding costs.
The approaches taken to several key methodological issues differed substantially
across the studies reviewed. Studies documenting the economic effects of
interpersonal violence have used a broad range of categories of costs. Those
estimating indirect costs – including the opportunity cost of time, lost
productivity and reduced quality of life – provided higher cost estimates than
studies that limited the costs of violence to direct costs alone. Other key
methodological issues included the economic values assigned to human life, lost
productive time and psychological distress. The rate at which future costs and
benefits are discounted, in accounting terms, also varied across studies.
Given the wide range of methodological differences and extensive gaps in the
existing literature on the economics of interpersonal violence, there is a clear
need for systematic future research into the costs of violence. Such research
should follow rigorous methodological guidelines, be inclusive of both direct and
indirect cost categories, and – perhaps most importantly – be comparable across
countries and settings.
The World report on violence and health (Krug et al., 2002) also showed that
effective interventions are available – particularly in the areas of child abuse and
neglect by caregivers, youth violence and gun-related violence. Given the
overwhelming evidence of the high costs of interpersonal violence, detailed
analysis of the economic feasibility of interventions is a clear research priority.
1
The 2002 World Report on Violence and Health (WRVH), published by the World
Health Organization (WHO), documented in detail the extent and
consequences of violence around the world. Global statistics showed that
520 000 individuals were victims of homicide in 2000; this number was almost
certainly underestimated given incomplete reporting systems and the illicit
nature of violence.
Violence disproportionately affects low- and middle-income countries. The
WRVH estimated that more than 90% of all violence-related deaths occurred in
these countries. The estimated rate of violent death in low- and middle-income
countries was 32.1 per 100 000 people in 2000, compared to 14.4 per 100 000
in high-income countries. As a result, the economic effects of violence were also
likely to be proportionally more severe in poorer countries. Nonetheless, as this
document shows, there was a scarcity of studies of the economic effects of
violence in low- and middle-income countries2.
The WRVH divided violence into interpersonal violence, self-directed violence
and collective violence.This report focuses on the relationships between the first
type – interpersonal violence – and economic factors. Based on an extensive
review of published articles and unpublished reports, the report treats the
following themes:
The economic effects of interpersonal violence in a wide variety of
socioeconomic and cultural settings (Chapter 2). Economic effects
were measured at the individual level as direct economic costs and
benefits, lost earnings, psychological costs and lost investments in
human capital – and at the aggregate level in terms of the effects of
interpersonal violence on investment, social cohesion and economic
growth.
The economic effects of interventions intended to reduce interpersonal
violence (Chapter 3).
The effects of economic conditions and policies on interpersonal
violence – with particular reference to poverty, structural adjustment,
income equality and social investment (Chapter 4).
Throughout this report, a public health economics perspective is taken to the
problem of interpersonal violence. Cost-benefit analysis, cost-effectiveness
analysis and the documentation of the economic costs of violence in terms of
1 Introduction
2 This report classifies countries by income level using the following categories from the 2003 World
Development Report (World Bank, 2003): low-income – $745 per capita or less; lower middle-income –
$746 to $2975; upper middle-income – $2976 to $9205; high-income – $9206 or more.
THE ECONOMIC DIMENSIONS OF INTERPERSONAL VIOLENCE
2
lost productivity and human capital dominated the methodologies used in the
literature reviewed here. It is important to note that there are several other
approaches to assessing the human toll of violence, including distinct literatures
from the sociological, anthropological, epidemiological and biomedical
perspectives. While related to each of these, the economic approach is unique in
its attempt to quantify, in monetary terms, the multifaceted causes and effects
of interpersonal violence. Since this report focuses on the economic approach, a
review of the literature reflecting these other perspectives lies beyond its scope.
The primary intended audience for this report consists of researchers working in
the field of violence prevention for whom a global overview of what is known
about the economic causes and effects of violence can provide pointers for a
future research and programmatic agenda. A goal of the report is to help frame
a research agenda for the field of economic evaluation related to interpersonal
violence. The report demonstrates the gaps in the present literature, in terms of
both content and methodology. The conclusion discusses the shortcomings of
existing studies and proposes initial guidelines for future research that will help
to promote comparability of findings across countries, settings and types of
violence.
Policy-makers and technical experts working on violence and injury prevention
are an important secondary audience for the report, which seeks to make this
group aware of the considerable economic costs of interpersonal violence – and
of the potential for savings offered by prevention. An important goal is to provide
policy-makers with the context necessary to promote and prioritize pertinent
research initiatives related to the economics of interpersonal violence – and to
position the costs of violence within national-level policy priorities.
1.1 Definitions
WHO defines violence as:
«The intentional use of physical force or power, threatened or actual,
against oneself, another person, or against a group or community, that
either results in or has a high likelihood of resulting in injury, death,
psychological harm, maldevelopment or deprivation» (Krug et al., 2002).
This definition explicitly includes psychological harm and deprivation among
the effects of violence, with corresponding implications for calculation of the
economic effects of violence (Chapter 2). While there is general agreement that
psychological distress is an important component of the economic burden of
violence, most studies have not quantified it in calculating the economic effects
of violence. Among those that have, there is little agreement in the
methodologies used.
This document defines interpersonal violence to include violence between
family members and intimates, and violence between acquaintances and
strangers that is not intended to further the aims of any formally defined group
or cause.Within the broad category of interpersonal violence, family and partner
1. INTRODUCTION
3
violence includes child abuse, intimate partner violence and elder abuse.
Acquaintance and stranger violence includes stranger rape or sexual assault,
youth violence, violence occurring during property crimes and violence in
institutional settings such as schools, workplaces and nursing homes. Selfdirected
violence, war, state-sponsored violence and other collective violence are
specifically excluded from these definitions.
Borrowing from the 2002 WRVH and a broad survey of the literature, this
document classifies subcategories of interpersonal violence, with corresponding
definitions, as follows:
Child abuse and neglect: «All forms of physical and/or emotional illtreatment,
sexual abuse, neglect or negligent treatment or commercial
or other exploitation, resulting in actual or potential harm to the child’s
health, survival, development or dignity in the context of a relationship
of responsibility, trust or power» (Krug et al., 2002).
Intimate partner violence: Behaviour within an intimate relationship
that causes physical, sexual or psychological harm, including acts of
physical aggression, sexual coercion, psychological abuse and
controlling behaviours. The term covers violence by both current and
former spouses and partners. Though women can be violent toward
men in relationships, and violence exists in same-sex partnerships, the
largest burden of intimate partner violence is inflicted by men against
their female partners (Krug et al., 2002).
Abuse of the elderly: «A single or repeated act, or lack of appropriate
action, occurring within any relationship where there is an expectation
of trust which causes harm or distress to an older person, including
physical, psychological or sexual abuse, and neglect.»
Sexual violence: «Any sexual act, attempt to obtain a sexual act,
unwanted sexual comments or advances, or acts to traffic, or otherwise
directed, against a person’s sexuality using coercion, by any person
regardless of their relationship to the victim.» This definition includes
rape, defined as physically forced or otherwise coerced penetration of
the vulva or anus, using a penis, other body parts or an object (Krug et
al., 2002).
Workplace violence: Violence committed in a place of employment.
Youth violence: Violence committed by or against individuals between
the ages of 10 and 29.
Other violent crime.
Chapter 2 describes in detail what is known about the economic effects of each
of these types of interpersonal violence, with the exception of elder abuse, where
the effects remain largely undocumented.
THE ECONOMIC DIMENSIONS OF INTERPERSONAL VIOLENCE
4
1.2 An ecological framework for assessing the economic
dimensions of interpersonal violence
To assess the economic dimensions of interpersonal violence, it is necessary to
understand the causes and identify the factors that increase the likelihood of
people becoming victims and perpetrators of such violence. According to the
WRVH no single factor can explain why one individual, community or society is
more or less likely to experience interpersonal violence. Instead, the Report
showed that interpersonal violence is a complex phenomenon rooted in the
interaction of many factors ranging from the biological to the political. To
capture this complexity, the WRVH adopted an ecological model that organizes
the risk factors for interpersonal violence into four interacting levels: the
individual level, relationships, community contexts and societal factors
(Figure 1).
Ecological model for understanding interpersonal violence
Individual-level risks include demographic factors such as age, income
and education; psychological and personality disorders; alcohol and
substance abuse; and a history of engaging in violent behaviour or
experiencing abuse.
At the relationship level, factors such as poor parenting practices and
family dysfunction, marital conflict around gender roles and resources,
and associating with friends who engage in violent or delinquent
behaviour increase the risk for most types of interpersonal violence.
The community level refers to the contexts in which social relationships
occur such as neighbourhoods, schools, workplaces and other
institutions. Poverty, high residential mobility and unemployment,
social isolation, the existence of a local drug trade, and weak policies
and programmes within institutions increase the risk of interpersonal
violence.
Societal-level risks are broad factors that create a climate in which
interpersonal violence is encouraged, including economic, social,
health and education polices that maintain or increase economic and
social inequalities; social and cultural norms that support the use of
violence; the availability of means (such as firearms) and weak criminal
justice systems that leave perpetrators immune to prosecution.
Figure 1
Societal Community Relationship Individual
1. INTRODUCTION
5
Interventions to prevent interpersonal violence are likewise usefully categorized
by the ecological model. Based on findings from the WRVH, interventions
shown through scientific evaluation to be of proven or promising effectiveness in
preventing interpersonal violence include the following:
Approaches for changing individual behaviour include pre-school
enrichment and social development programmes, as well as vocational
training and incentives to complete secondary schooling. These are
designed to ensure academic success, manage anger and build skills,
and are effective in preventing youth violence. Similar life-skills and
educational approaches around issues of gender, relationships and
power have been used to address physical and sexual violence against
women. Effective treatment and counselling can reduce the potential
for further physical and psychosocial harm after interpersonal violence
has been experienced.
Relationship-level interventions include those delivered in early
childhood, such as parenting programmes, the provision of support and
advice through home visitation in the first 3 years of a child’s life, and
family therapy for dysfunctional families.These types of approaches, for
instance, have been associated with reductions in child abuse and with
long-term reductions in violent and delinquent behaviour among
young people. Strong mentoring is another approach.
Community-level interventions include reducing the availability of
alcohol; changing institutional settings – e.g. schools, workplaces,
hospitals and long-term care institutions for the elderly – by means of
appropriate policies, guidelines and protocols; providing training to
better identify and refer people at-risk for interpersonal violence; and
improving emergency care and access to health services.
At the societal level, promising interventions include providing accurate
public information about the causes of interpersonal violence, its risks
and its preventability; strengthening law enforcement and judicial
systems3; implementing policies and programmes to reduce poverty
and inequalities of all kinds; improving support for families; and
reducing access to firearms and other means of violence.
Experiences demonstrating the economic effects of interventions directly
intended to reduce interpersonal violence are described in Chapter 3. The
effects on interpersonal violence of economic factors at the community and
societal levels, and of government policies to address them, are described in
Chapter 4.
3 Economic theory predicts that criminal behaviour will respond to incentives, including the threat of
punishment. Becker (1993) initiated a line of research using a general cost-benefit framework to model
criminals responses to economic incentives.
THE ECONOMIC DIMENSIONS OF INTERPERSONAL VIOLENCE
6
1.3 Types of costs
Studies documenting the economic effects of interpersonal violence have used a
broad range of categories of costs. Much of the difference in terms of the overall
estimates made by the studies reviewed in this report was due to the inclusion
or exclusion of different categories of costs, rather than to different
methodologies in tracking costs.
Costs and benefits of interpersonal violence
As shown in Figure 2, cost categories can be broadly grouped into direct costs
and benefits – those resulting directly from acts of violence or attempts to
prevent them – and indirect costs and benefits. The most commonly cited direct
costs were medical care and the costs of the judicial and penal systems – policing
and incarceration. Indirect costs included the long-term effects of acts of
violence on perpetrators and victims, such as lost wages and psychological costs,
also referred to as pain and suffering (Hornick, Paetsch & Bertrand, 2002).The
calculation of psychological costs was a common practice in legal cases seeking
to assess the monetary value of reimbursement to victims of violence.
Psychological costs were generally significantly greater than the direct economic
losses incurred by victims (Miller, Cohen & Rossman, 1993). Some studies
attempted to place a value on the negative affect of violence on housing values –
a cost to society. For example, in the USA, a doubling in homicide rates was
associated with a 12.5% decline in property values (IADB, 2002). Other
indirect cost categories quantified the effects of violence beyond the immediate
perpetrators and victims – for example, a negative impact on investment in
countries with high rates of violence and higher insurance rates for all of society.
Figure 2
Interpersonal violence
– Child abuse and neglect
– Intimate partner violence
– Elder abuse
– Sexual violence
– Workplace violence
– Youth violence
– Other violent crime
Direct costs and benefits
– Costs of legal services
– Direct medical costs
– Direct perpetrator control costs
– Costs of policing
– Costs of incarceration
– Costs of foster care
– Private security contracts
– Economic benefits to perpetrators
Indirect costs and benefits
– Lost earnings and lost time
– Lost investments in human capital
– Indirect protection costs
– Life insurance costs
– Benefits to law enforcement
– Productivity
– Domestic investment
– External investment and tourism
– Psychological costs
– Other non-monetary costs
1. INTRODUCTION
7
1.4 Economic evaluation of interventions
The economic evaluation of interventions is undertaken to guide decisionmaking
so that scarce resources can be allocated in the most effective way.
Accordingly, one of the main principles of economic evaluation is that it should
involve a comparison of the costs and benefits of multiple options (Gold, Siegel
& Weinstein, 2001). An economic evaluation can be conducted from a variety of
perspectives, such as societal, sectoral or organizational. Each perspective differs
in the costs that are selected for evaluation. The selection of a perspective will
largely depend on the primary stakeholder; but when multiple major
stakeholders are present, as often is the case, it is not uncommon to conduct an
economic evaluation from multiple perspectives.
A range of economic analyses have commonly been used for comparing violence
interventions, including cost-utility analysis, cost-benefit analysis and costeffectiveness
analysis (see Appendix 1 for further explanation of the typologies).
The type of evaluation conducted will depend on the outcome indicator used –
for example, quality-adjusted life years (QALYs), monetary units or cases
averted. While intervention-specific indicators allow for more accurate
assessments of particular interventions, they also limit the ability of making
cross-intervention comparisons. For example, if intervention A has a ratio of
$0.01 per rehabilitation session attended, and intervention B has a ratio of $10
per crime averted, it is difficult to determine which intervention is a better use
of resources. Therefore, when choosing an outcome indicator, it is essential to
consider all of the plausible comparisons so that the evaluation can be effectively
used as a decision-making tool (Drummond & McGuire, 2001).
Intervention Costs
Programme costs arise from the development and implementation of
interventions aimed at reducing the burden of interpersonal violence. This will
include the costs of all inputs – both fixed capital investments and recurrent
programme costs – necessary to provide the intervention. Common examples of
such costs include operating costs, labour costs and capital costs (Drummond &
McGuire, 2001). Programme costs are especially important when conducting
economic evaluations to compare interventions, since these costs will likely vary
between interventions and can greatly influence their relative cost-effectiveness.
Therefore, when reading the later sections in this report regarding the benefits
of individual interventions, particular scrutiny should be given to their
associated programme costs.
From an economic perspective, the reduction of direct and indirect costs
resulting from an intervention can be referred to as the benefits of that
intervention. Programme costs, however, can be thought of as the investment
necessary to achieve those benefits. Therefore, programme costs will most often
be found in the numerator of a cost-effectiveness analysis (as costs), whereas the
reduction of direct or indirect costs will be found in the denominator (as
benefits).
THE ECONOMIC DIMENSIONS OF INTERPERSONAL VIOLENCE
8
Intervention Benefits
A wide variety of indicators can be used to measure the benefits of an
intervention, and their selection will largely depend on the goals of that
intervention. For example, a violence prevention intervention could be
measured in terms of saved lives or violent acts averted, while an intervention
targeting prior offenders could be measured by recidivism rates. Less
straightforward, however, is the measurement of the benefits gained by
interventions aimed at the victims of violent acts. An indicator for this type of
intervention not only would have to take into account its impact on the quality
of life of the victim, but should also be a metric that allows for comparison
between interventions.A potentially powerful candidate for this type of indicator
is the QALY.
The basic concept of QALY is straightforward. Utility scores for particular
health states are first elicited from members of the targeted population. Health
utility scores can range between 0 and 1, where 0 is the equivalent of being dead
and 1 represents perfect health, although some health states are regarded as
being worse than death and have negative valuations. These scores can be
elicited in a number of ways, but the most commonly used are the time-tradeoff,
standard gamble, and visual analogue methods (Drummond & McGuire, 2001).
The amount of time spent in a particular health state is then weighted by the
utility score attributed to that health state. A perfect health (utility score 1) of 1
year would equal 1 QALY, but 1 year in a health state with half of that utility
(utility score .5) would equal .5 QALYs. Thus, an intervention that generates 4
additional years in a health state valued at 0.75 will generate 1 more QALY than
an intervention that generates 4 additional years in a health state valued at 0.5.
The use of QALYs as an outcome indicator for interventions aimed at victims of
violence has so far been limited. Therefore, further research needs to be
conducted to determine the feasibility as well as the appropriate methodology
for collecting health utility scores from victims of violence. In addition,
standardized utility scores – similar to the «EuroQol» survey for health states
(Dolan, 1997) – for the different types and degrees of violence should also be
developed. Evaluations expressed in other units, such as cost per case averted,
could then be modelled to derive a cost per QALY ratio.The overall benefit of
this research would be the establishment of a common metric to compare the
effectiveness of a wide range of interventions, including violence prevention,
offender rehabilitation and victim counselling.
1.5 Key methodological issues
In addition to differences in terms of the categories of costs and benefits
included, there were several other methodological issues where there were
significant disparities among the studies reviewed. There were important
differences in how rates of interpersonal violence were estimated. Sources for
estimates included crime reports, hospital records and household surveys.
Substantial numbers of violent acts – particularly intimate partner violence – go
unreported and untreated. As a result, all of these sources were likely to result in
1. INTRODUCTION
9
underestimates of the true incidence of violence.
As with any attempt to quantify the costs of morbidity and mortality, a principal
methodological difference was in the dollar values assigned to a human life (see
Box 1), lost productive time and psychological distress. Another important
difference among the studies was the varied perspective from which costs were
calculated. The majority of the studies of the costs of violence used a societal
perspective – in other words, in principle all costs were included whether they
accrued to the victim, the perpetrator, a third party payer or society at large.
Several studies, however, included only costs to the victims, without counting
the social costs of prevention, law enforcement, incarceration and lost
productivity.
A further key methodological difference among studies was the time frame used
to calculate costs. Most of the cost estimates of the aggregate economic losses
caused by violence were for a 1-year time period. But the time frame varied,
making direct comparisons difficult. Studies undertaken from the individual
perspective often calculated direct and indirect costs for the lifetime of the
individual.
Nearly all studies that calculated costs and benefits beyond a 1-year time frame
used some kind of discount rate to estimate future costs and benefits – based on
the principle that humans value consumption and quality of life in the present
more than they do an equivalent amount of consumption in the future. This
concept is rooted in uncertainty about the future – making it more desirable to
consume or benefit from life in the short-run than to wait for the equivalent
amount of consumption in the future. For economists, the concept of
consumption is most often considered equivalent to and measurable by the level
of expenditures for an individual or a household. However, the concept of
quality of life itself was not consistently defined in the economic literature;
generally it was equated with individuals’ willingness to pay for improvements in
their lives, whether such improvements were material or intangible.
BOX 1
The value of a human life
Among studies that quantify the value
of lost human life, there is
considerable variation in the monetary
value assigned to one life. The value of
life is most commonly calculated using
estimates of the quality of life, wage
premiums for risky jobs, willingness to
pay for safety measures and individual
behaviour related to safety measures
such as using seatbelts (Boardman et
al., 1996).
The values used among studies
reviewed in this document ranged
from $3.1 million to $6.8 million. These
estimates are in line with those
generally used in the literature. Miller
(1989) reviewed 29 cost-benefit
studies and found that the mean value
given to a human life in these studies
was $4.2 million. Fisher, Chestnut &
Violette (1989) reviewed 21 studies
and found a range of $2.6 million to
$13.7 million. Walker (1997) used a
figure of $602 000, but this did not
include the costs of the judicial system
or psychological costs.
Finally, Viscusi (1993) examined 24
studies using wage-risk trade-offs to
estimate the value of life. Most of these
studies placed the value of life
between $4.0 million and $9.4 million.
Viscusi also pointed out that risk was a
less robust predictor of wage levels
than other factors, particularly
education.
THE ECONOMIC DIMENSIONS OF INTERPERSONAL VIOLENCE
10
The discount rates used in the studies reviewed here ranged from 2.0% to 10%.
It should be noted, however, that only a small proportion of the studies reviewed
actually gave the discount rates they used, further complicating comparison of
the findings between them. The United States Panel on Cost-Effectiveness in
Health and Medicine has recommended using a real rate of 3.0% for cost
evaluations in health care (Gold, Siegel & Weinstein, 2001). This rate reflects a
wide range of studies documenting individuals’ preferences for present
consumption compared to future consumption and interest rates for private
investment. In theory, both of these factors influence the discount rate for future
costs and benefits in the context of financial and health-related gains and losses.
As stated above, monetary values in this document have been converted to 2001
US dollars to enable comparisons and to adjust for inflation and varying
exchange rates.Values expressed in other currencies in original documents, and
US dollar values from previous years, have been converted to 2001 US dollars
using the US consumer price index and applicable international exchange rates
from the year of the original estimates. Costs expressed as a percentage of the
gross domestic product were calculated using the gross domestic product from
the year the costs were reported.
1.6 Search strategy
We conducted a wide-ranging literature search using electronic databases and
the Internet as detailed in Appendix 2. The search was conducted without
restrictions on the language of publication and included publication dates from
January 1980 to May 2003.We used the following keywords, representing types
of violent behaviour and factors associated with violent behaviour combined
with economic variables:
After the initial database and Internet literature searches, additional sources
were identified through the reference lists of collected articles and through
consultation with resource experts. Studies were included in this review if they
contained: a costing component of the violent behaviour’s effects or factors
associated with this behaviour (the cost of treating firearm-related injuries for
instance); costing information on interventions (such as the cost of the
Violent behaviour and related factors Economic variables
Violence: interpersonal violence, family Costs: cost-effectiveness, cost-benefit
violence, partner violence, domestic violence
Abuse: child abuse, domestic abuse, Economics: economic policy
partner abuse, girl abuse
Assault Benefits
Homicide Investments
Injury and intentional injury Human capital
Firearms Expenses
1. INTRODUCTION
11
intervention itself or the economic effects of the intervention), or a framework
for analyzing the relationship between violence and its various economic
dimensions. After the review, a total of 119 studies were retained, of which 54
are from the peer reviewed literature and 65 are not peer reviewed. Of these
studies, 79 pertain to the first theme of the review – i.e. the economic effects of
interpersonal violence. There are 27 studies relevant to the second theme – the
economic effects of interventions intended to reduce interpersonal violence –
and 13 pertain to the third theme – the effects of economic conditions and
policies on interpersonal violence. The contents of these studies were
systematically abstracted using the information categories listed in Appendix 3.
1.7 Characteristics of included and excluded literature
While 119 studies were retained for analysis in this review, a total of 248 were
considered based on the keywords described in the search strategy. As a
relatively large proportion of all studies examined was excluded, it is important
to describe in further detail the exclusion criteria and the characteristics of
excluded studies so that, ideally, future research into the economic effects of
interpersonal violence might follow more consistently the characteristics of the
included studies.
A clearly measurable costing component was a key prerequisite for inclusion in
the review. Whereas searches of the social science and policy literature yielded a
bounty of research examining various aspects of violence – including strategies
for prevention, social environments that foster violence, roles of various
stakeholders in violence prevention, and the relationship between violence and
social capital – these studies did not generally determine direct or indirect costs
related to interpersonal violence. The strength of much of this social science
literature is a testament to the importance of considering sociopolitical variables
and their relationships with violence and violence prevention. However, the
relative lack of economic data on actual monetary costs – direct or indirect –
highlights an essential area for increased attention, given the importance of
costing data in any accurate reflection of the burden of violence. A number of
studies based on theoretical models predicting violence were likewise excluded
if they did not have an empirical component.
It is clear from the review that data on economic dimensions of interpersonal
violence from low- and middle-income countries are scarce. Much of the raw
data from high-income countries have been extracted from central government
sources, such as the United States Department of Justice and the Australian
Institute of Criminology. A partial explanation for the lack of costing data from
low- and middle-income countries is the absence of reliable data collection
mechanisms from government sources, leaving little from which researchers can
examine trends and draw conclusions. Furthermore, a significant portion of the
costing data has been extracted from hospital-based accounting and recordkeeping
systems – areas in which lower income countries are at a significant
disadvantage.
THE ECONOMIC DIMENSIONS OF INTERPERSONAL VIOLENCE
12
We have presented here a range of costing data to accurately reflect the available
literature and have pointed out where there are possible variations in the quality
and rigour of the included studies. The discussion of the economic correlates of
violence (Chapter 4) briefly reviews key sociological literature relevant to this
field and only provides a glimpse into the extensive literature on the relationship
between interpersonal violence and factors such as economic inequality,
employment rates and welfare expenditure.
2. COSTS OF INTERPERSONAL VIOLENCE
13
A total of 119 studies and documents discussing the costs of violence were
retained for this review: 54 are from the peer reviewed literature and 65 are from
other sources, including governments and international organizations (see
Chapter 1 for a discussion of the criteria used to include documents for review
in this report). Because no systematically documented studies of the economic
effects of abuse of the elderly were found, this category has been dropped from
the review. As the 2002 WRVH shows, elder abuse is common in countries of all
income levels, indicating the urgent need for further research of this topic.
2.1 Overview of the costs
There are widely varying estimates of the cost of violence internationally,
depending on the definitions used, the types of costs included and the
methodologies used. In the USA, the Department of Justice (1994) reported the
direct costs of violent crime to victims, based on reported crimes and on
responses to the National Crime Victimization Survey (NCVS)4. In 1994, these
costs amounted to $1.8 billion – 0.02% of the USA gross domestic product for
that year.
When indirect costs are included, estimates of the costs of violence in the USA
soar. The National Crime Prevention Council (1999) included effects on
employment and workers’ productivity as well as psychological costs to calculate
an estimated $46.8 billion for the costs of violent crime in the USA – equivalent
to 0.5% of the gross domestic product. Miller, Cohen & Rossman (1993)
estimated an annual cost of intentional injuries of $84.1 billion in the USA for
the time period 1987-1990. Fromm (2001) estimated an annual cost of $94
billion to the USA economy resulting from child abuse alone – 1.0% of the gross
domestic product. These estimates included psychological costs to the victim –
the equivalent of pain and suffering – extrapolated into the future using a
discount rate ranging from 3.0% to 5.0%.Throughout the literature on the costs
of violence, psychological costs greatly outweighed the direct costs of violence –
partially explaining the wide variance in the estimates that are available.
Miller, Fisher & Cohen (2001) used archival data on all 377 000 violent crimes
reported in Pennsylvania in 1993. They estimated lost earnings, psychological
costs and the opportunity cost of victims’ time, in addition to the costs of
policing, incarceration and life insurance. Their estimate for Pennsylvania of
$14.2 billion equalled $329.8 billion when extrapolated to the full USA
population; equivalent to 3.3% of the gross domestic product. Separately,
2 Costs of interpersonal
violence
4 The NCVS is conducted annually based on 100 000 interviews with crime victims. The NCVS does not
include incidents that occurred more than 6 months prior to the interview date.
THE ECONOMIC DIMENSIONS OF INTERPERSONAL VIOLENCE
14
Miller, Cohen & Wiersema (1996) estimated a total annual cost to victims of
personal crime in the USA, including domestic violence, sexual assault, rape and
child abuse, of $507 billion.This estimate, which included psychological costs of
pain and suffering, was equivalent to 6.5% of the gross domestic product.
In England and Wales, Brand & Price (2000) estimated total costs from crime of
$63.8 billion; 63%, or $40.2 billion, of this amount was attributable to violence
– including homicide, wounding and sexual assault. This tally included both
direct costs such as police and judicial system costs, and indirect costs that
included foregone output and physical and emotional costs.
Estimates from other high-income countries tended to be conservative. The
Australian Institute of Criminology (2001) reported national annual costs of
assault of $159 million, an estimate that did not include indirect costs and was
largely based on the costs of incarcerating offenders. Similarly, the cost of
homicide in Australia was calculated at $194 million per year, based on a cost
per homicide of $602 000 (Walker, 1997). Fanslow et al. (1997) calculated the
economic cost from homicide in New Zealand: when lost earnings, legal fees,
incarceration and policing were included, the cost per homicide was $829 000
for a total of $67.9 million.
At the global level, Pfizer (2001) estimated that crime and violence together cost
the equivalent of 5.0% of the gross national product of industrialized countries,
and as much as 14% of the gross national product of low-income countries. But
there were few documented estimates of the costs of violence in low- and
medium-income countries. Additionally, comparisons with high-income
countries were complicated by the fact that economic losses related to
productivity tended to be undervalued in lower income countries since these
losses were typically based on foregone wages and income. Phillips (1998)
calculated the cost of homicides in the Western Cape Metropolitan Area in
South Africa. Using a 4.0% rate to discount future productivity and opportunity
costs, the costs of homicide were $31.6 million. This worked out to an average
of $15 319 per homicide, sharply lower than the New Zealand estimate above of
$829 000.
Costs of violence in Latin America Table 1
Country % 1997 gross domestic
product lost due to violence
Brazil 10.5%
Colombia 24.7%
El Salvador 24.9%
Mexico 1.3%
Peru 5.1%
Venezuela 11.8%
Source: Buvinic, Morrison & Shifter (1999).
Definition of violence includes collective violence.
2. COSTS OF INTERPERSONAL VIOLENCE
15
The Inter-American Development Bank (IADB) has spearheaded efforts to
systematically document the costs of both interpersonal and collective violence
in the Americas. Central America has been particularly affected. In 1995, El
Salvador spent 6.0% of the gross national product to control violence (IADB,
2002). As part of the IADB’s work, Buvinic, Morrison & Shifter (1999) reported
estimates of economic losses due to social violence – including collective violence
– in a variety of countries (Table 1). These losses ranged from 5.1% of the gross
domestic product in Peru to 24.9% in El Salvador, which was still in the throes
of guerrilla war in the mid-1990s when the data were generated. The estimates
included lost earnings, the opportunity cost of time, policing, incarceration,
judicial costs, foregone investments in human capital and effects on investment5.
2.2 Child abuse and neglect
The extent of child abuse and neglect is difficult to gauge since much if not most
of it is unreported. The 2002 WRVH estimated that the rates of homicide of
children under 5 years of age was 2.2 per 100 000 for boys and 1.8 per 100 000
for girls in high-income countries. In low- and middle-income countries, the
corresponding rates were 6.1 and 5.1 per 100 000 for boys and girls,
respectively. In Africa, the rates were 17.9 per 100 000 for boys and 12.7 per
100 000 for girls.
Most of the estimates of the costs of child abuse available in the literature are
from the USA (Table 2). Fromm (2001) reviewed a variety of sources and
calculated an aggregate total of $94 billion in annual costs to the USA economy
resulting from child abuse – 1.0% of the gross domestic product. This total
included direct medical costs and the related costs of legal services, policing and
incarceration, as well as the value of indirect productivity losses, psychological
costs and future criminality. Hospitalization accounted for $3.0 billion, mental
health treatment costs for $425 million, and child welfare costs for $14.4 billion.
The largest single component of Fromm’s estimate was adult criminality related
to child abuse, for which he calculated an annual figure of $55.4 billion.
Using secondary sources, Courtney (1999) calculated a figure of $14 billion for
direct costs, including counselling and child welfare services resulting from child
abuse in the USA. Caldwell (1992) presented detailed estimates for Michigan,
totalling $1.0 billion. He included direct medical costs, lost tax revenue due to
premature death, special education, psychological and welfare services,
protective services, foster care, preventive services, and adult criminality and
subsequent incarceration related to child abuse. Separately, the United States
Department of Health and Human Services (2001) calculated the costs of child
abuse and maltreatment in Colorado to be $468 million – of which indirect costs
represented 53%.
There was a considerable range of estimates of individual-level treatment costs
for child abuse, depending on the types of costs included. On the high end,
5 The cost just of health expenditures related to violence as a percentage of gross domestic product was as
follows (Buvinic, Morrison & Shifter, 1999): Brazil 1.9%; Colombia 4.3%; El Salvador 4.3%; Mexico 1.3%;
Peru 1.5%;Venezuela 0.3%.
THE ECONOMIC DIMENSIONS OF INTERPERSONAL VIOLENCE
16
Irazuzta et al. (1997) calculated treatment costs per child of $42 518 based on
daily hospital charges of $6 317 in a paediatric intensive care unit in
Charlestown,West Virginia. Forjuoh (2000), working with discharge data from
acute care hospitals in Pennsylvania, calculated a mean hospitalization cost of
$18 103 per child abuse victim. Plotnick & Deppman (1999) calculated an
average cost of $12 028 for hospitalizing an abused child for one week.
Table 2
Study Study location Cost categories included Total annual costs
and population (indirect costs in italics) (2001 US$)
Peer reviewed articles and government studies
Courtney (1999) US. Maltreated children and their
families
Direct medical (including
counselling) $14.0 billion
Evasovich et al.
(1998)
US, Ohio (4 counties). Children
referred to burn unit for suspected
child abuse (n = 104)
Direct medical, legal fees
(court costs, fines)
$0.3 million;
$13 781 per child
Forjuoh (2000) US, Pennsylvania, 1995. Hospital
discharge data (n = 348 children) Direct medical $6.3 million;
$18 103 per child
Irazuzta et al. (1997) US,West Virginia, 1991-1994.
Paediatric ICU admissions (n = 13) Direct medical $0.6 million;
$42 518 per child
New & Berliner
(2000)
US,Washington state, 1994. Mental
health treatment costs compensated
by Crime Victims Compenstation
program (n = 608 children)
Direct medical $1.2 million;
$2 921 per child
Summers &
Molyneux (1992)
UK, 1990. Children hospitalized
(n = 181)
Direct medical – examinations
only $128,758 per child
US Department of
Health and Human
Services (2001)
US, Colorado, 1995
Direct medical, legal services,
incarceration, workers’
productivity, lost earnings and
opportunity cost of lost time
$468 million
Walker et al. (1999)
US,Washington state. Randomly
selected sample of women
(n =1 225) enrolled in an HMO,
42.8% maltreated as children
Direct medical $9.1 million;
$17 356 per child
Caldwell (1992) US, Michigan. Cases recorded by
Department of Social Services
Direct medical, incarceration,
policing (Protective Services),
lost earnings and opportunity
cost, lost investments in human
capital, psychological costs, other
non-monetary costs
$1.0 billion
Fromm (2001) US. Aggregated studies
Legal services, direct medical,
policing, incarceration, workers’
productivity, psychological costs,
other non-monetary costs
$94 billion
Studies conducted by advocacy groups
Costs of child abuse – selected studies
2. COSTS OF INTERPERSONAL VIOLENCE
17
Evasovich et al (1998) studied the costs of medical care and related legal fees for
children admitted to the burns unit at the Children’s Hospital Medical Center
in Akron in Ohio – finding costs totalling $13 781 per child. Public funds paid
for 65% of these costs. New & Berliner (2000) examined claims paid by the
Crime Victims Compensation programme in Washington state and calculated
mental health treatment costs of $2 921 per child, with counselling services
costing between $70 and $90 per hour and exclusive of hospitalization or other
health care costs.
Walker et al. (1999) studied health care costs for adult women in Washington
state who had been abused as children. They found that 42.8% of a randomly
selected sample of 1 225 women had been abused as children. The median
annual health care costs for these women were $108 greater than for women
who had not been abused.
There are few estimates available of the cost of child abuse internationally. A
study of abused children seen in the emergency room of a hospital in Liverpool,
England, found annual costs of $128 758 per child just for assessment and initial
treatment (Summers & Molyneux, 1992). Mendonca, Alves & Filho (2002)
measured hospital costs due to violence against children and adolescents in
Pernambuco state, Brazil. The mean cost of hospitalization for children and
adolescents was $184. Violence against children and adolescents in greater
metropolitan Recife, the principal city in the state, accounted for 65.1% of
hospital admissions and 77.9% of hospital costs.
2.3 Intimate partner violence
As with other types of interpersonal violence, the true extent of intimate partner
violence is unknown. Disparate surveys suggest there is a wide range in
prevalence, but the survey results are difficult to compare given cultural
differences and taboos in responding to questions. In Paraguay and the
Philippines, 10% of women surveyed reported being assaulted by an intimate
partner (Heise, Ellsberg & Gottemoeller, 1999), compared to 22% in the USA
(Tjaden & Thoennes, 2000), 29% in Canada, and 34% in Egypt (El-Zanaty et
al., 1996). Other s
