plospmedplmePLoS MedplosmedPLoS Medicine1549-12771549-1676Public Library of ScienceSan Francisco, USA10.1371/journal.pmed.005024308-PLME-PF-1451R2Policy ForumNon-Clinical MedicinePublic Health and EpidemiologyMedical consequences of war/conflictResearch MethodsPublic HealthEpidemiologyThe Dirty War Index: A Public Health and Human Rights Tool for Examining and
Monitoring Armed Conflict OutcomesPolicy ForumHicksMadelyn Hsiao-Rei*SpagatMichael* To whom correspondence should be addressed. E-mail: MJHHicks@aol.com
Madelyn Hsiao-Rei Hicks is an Honorary Lecturer with the Sections of Community Mental
Health and Cultural Psychiatry, Health Service and Population Research Department,
Institute of Psychiatry, King's College London, London, United Kingdom. Michael Spagat
is a Professor in the Department of Economics, Royal Holloway College, University of
London, Egham, United Kingdom.
The authors have declared that no competing interests exist.
12200816122008512e2432008 Hicks and SpagatThis is an open-access article distributed under the
terms of the Creative Commons Attribution License, which permits unrestricted use,
distribution, and reproduction in any medium, provided the original author and source are
credited.A New Tool for Measuring the Brutality of WarThe Dirty War Index: Statistical Issues, Feasibility, and
InterpretationRape in War Is Common, Devastating, and Too Often Ignored
Madelyn Hsiao-Rei Hicks and Michael Spagat introduce the "Dirty War Index," a public
health tool that identifies rates of undesirable or prohibited war outcomes inflicted on
populations during armed conflict.
The authors received no specific funding for this article.citationHicks MH-R, Spagat M (2008) The Dirty War Index: A public health and human
rights tool for examining and monitoring armed conflict outcomes. PLoS Med 5(12): e243.
doi:10.1371/journal.pmed.0050243
Documentation, analysis, and prevention of the harmful effects of armed conflict on
populations are established public health priorities [1–5]. Although public health research on
war is increasingly framed in human rights terms [6–13], general public health methods are
typically applied without direct links to laws of war. Laws of war are international
humanitarian laws and customary standards regarding the treatment of civilians and
combatants, mainly described in the four Geneva Conventions of 1949 and their Additional
Protocols I and II regarding international and civil conflicts [14]. With notable exceptions
[11,15–17], absolute numbers are usually
reported (e.g., number of persons killed), without systematic description of the
proportional effects of armed conflict, thereby limiting the utility of findings and scope
of interpretation.
In this paper, we introduce the “Dirty War Index” (DWI): a data-driven
public health tool based on laws of war that systematically identifies rates of particularly
undesirable or prohibited, i.e., “dirty,” war outcomes inflicted on
populations during armed conflict (e.g., civilian death, child injury, or torture). DWIs are
explicitly linked to international humanitarian law to make public health outcomes directly
relevant to prevention, monitoring, and humanitarian intervention for the moderation of
war's effects. After choosing the particular outcome to be measured, a DWI is calculated as:
Summary Points
War, a major public health problem, is a situation where the interests of public
health, human rights, and humanitarian law intersect.
The DWI is a data-driven public health tool that identifies rates of particularly
undesirable or prohibited, i.e., “dirty,” outcomes inflicted on
populations during war (e.g., civilian death, child injury, or torture).
A DWI is calculated as: (Number of “dirty,” i.e., undesirable
or prohibited cases/Total number of cases) × 100.
DWIs are designed for direct, easy translation of war's public health outcomes into
the human rights, policy, and interdisciplinary work needed to address war's
practice.
DWIs support monitoring, deterrence, and humanitarian intervention by explicit
links to international humanitarian laws and by exposing rates of unacceptable
combat outcomes (DWI values) from different weapons or combatant groups.
For example: In Table 1, we measure
the DWI ratio of “Number of civilians killed/Total number of civilians and
opponent combatants killed” using a casualty dataset for Colombia's civil conflict
[18]. Table 2 links this DWI to relevant laws of
war. DWI values of 99 for illegal paramilitaries, 46 for guerrillas, and 45 for government
forces show that paramilitaries are “dirtiest” in terms of proportion of
civilians constituting their victims of unopposed attacks (chi-square = 5,010,
degree of freedom [df] = 2, p < 0.001).
99% of paramilitary victims were civilians and only 1% were military
opponents. This finding, combined with the paramilitaries' methods (execution by close-range
gunfire in massacres), suggests intentional targeting of civilians that requires recognition
in Colombia's paramilitary demobilization, disarmament, and reintegration process
[19].
10.1371/journal.pmed.0050243.t001
Dirty War Index for Attacks by Actors in the Colombian Civil Conflict,
1988–2005: Civilian Versus Opponent Combatant Mortality
10.1371/journal.pmed.0050243.t002
DWIs Suggested for Measuring Rates of Undesirable or Prohibited Outcomes from
Aggression in Armed Conflict
As ratios, DWIs complement absolute numbers and lend themselves to comparisons over time,
between wars, between weapons, and between warring combatant groups to identify better
versus worse performers. Noncombatant wounded-to-killed ratios can provide evidence of war
crimes [16].
Proportional “atrocity statistics” [20] from a Darfur survey substantiated US
Secretary of State Colin Powell's declaration of genocide and the referral of Darfur's
situation to the International Criminal Court [20,21]. By facilitating clear, systematic comparisons, DWIs can help analyze
and expose how combatants engage in war and affect populations, thereby increasing the
accountability of military and political leaders. This paper describes the theoretical basis
and practical applications of the DWI, with brief examples from armed conflicts. More
detailed DWI analyses of specific conflicts are planned for future papers.
Calculating and Using DWIs
A DWI can be easily used and understood, facilitating interdisciplinary communication and
research on war's effects. DWIs can measure rates of undesirable outcomes from accepted
methods (e.g., civilian casualties from aerial bombing of military targets). They can also
measure rates of using prohibited, illegitimate methods (e.g., torture), and rates of
applying illegitimate methods to especially vulnerable populations (e.g., torturing
children) to describe rates of exceptional atrocity. However, the mere application of DWI
analysis to a combatant group does not indicate that it is “dirty”: a
DWI ratio simply identifies how often, if at all, the group is linked with the particular
undesirable outcome being measured, facilitating comparisons. To illustrate, we draw on data
from B'Tselem (http://www.btselem.org/english/statistics/Index.asp), a nongovernmental
organization that monitors casualties from the Israeli-Palestinian conflict. We apply a
“female mortality DWI” (Number of females killed/Total number killed) to
conflict-related killings from September 29, 2000 to April 30, 2007: Israeli security forces
killed 213 females among 4,057 Palestinians (DWI = 5). Palestinians killed 283
females among 705 Israeli civilians (DWI = 40). Palestinians killed 10 females
among 317 Palestinians (DWI = 3). Comparison of actors' DWIs shows significantly
higher discrimination of female from male targets by Israeli security forces and by
Palestinian actors when targeting Palestinians, and lower discrimination of female from male
targets when Palestinian actors target Israeli civilians (chi-square = 833, df
= 2, p < 0.001).
The best possible DWI value is 0, indicating that the objectionable outcome is identified
in no measured cases. The worst possible DWI value is 100, indicating that the objectionable
outcome is identified in 100% of measured cases. Any rate above 0 for prohibited
actions or war crimes is unacceptable, and eliminating violations is imperative. DWIs for
undesirable outcomes are less straightforward. The highly undesirable outcome of civilian
harm is not prohibited by laws of war if combatants do everything feasible to distinguish
between civilians and military targets (the principle of distinction), if they attempt to
minimize incidental harm to civilians, and if they intend to avoid harming civilians in
excess of anticipated military goals (the principle of proportionality) [1,22,23]. Civilian harm is also balanced against the “military
necessity” of objectives [24]. Though what is feasible, proportional, or necessary is highly
subjective [22–24], clearly the lowest possible rates should be sought for
undesirable outcomes such as “incidental” civilian death. High DWI
values for undesirable outcomes indicate extreme destruction, signal the need for close
scrutiny, and may suggest war crimes.
Tables 2 and 3 list specific DWIs, their pertinent laws of war, and
example calculations. Table 2 lists
DWIs for undesirable or prohibited aggression in armed conflict. DWIs can be analyzed by
demographic subgroup for indiscriminate warfare, disproportionate effects of targeting, or
particular vulnerability to weapons. For example, with “casualties”
defined as injuries or deaths, a “child casualty DWI” (Number of child
casualties/ Total number of casualties) applied to weapons-casualty data from Chechnya
[25] gives the
following child casualty ratios for different explosive devices: antitank landmines (34/223,
DWI = 15), antipersonnel landmines (223/1,004, DWI = 22), booby traps
(65/214, DWI = 30), and other unexploded ordnance (UXO) (255/892, DWI
= 29). DWIs indicate that in Chechnya, UXO and booby traps are more dangerous to
children than landmines and significantly “dirtier” in this respect
(chi-square = 25.0, df = 3, p < 0.001).
10.1371/journal.pmed.0050243.t003
DWIs Suggested for Measuring Rates of Unacceptable Endangerment in Armed
Conflict
Table 3 lists DWIs for unacceptable
endangerment in armed conflict [14,23,24]. To illustrate, we apply the last DWI listed,
“Destroying infrastructure essential for civilian survival (food, water,
hospitals),” to survey data from eastern Burma where the Burmese military junta is
in conflict with ethnic minority groups. The Burmese military regime destroyed or stole food
from 472 of 1,813 surveyed households [11]. The Burmese military's DWI of 26 indicates a
26% rate of committing the humanitarian violation of destroying civilian food
sources, associated in the study with significantly greater odds of household landmine
injury (perhaps due to foraging for food), child malnutrition, and death [11].
In Table 4 we analyze the Northern
Ireland conflict for two complementary DWIs: aggressive acts (killing civilians) and
endangerment to civilians (by not wearing uniforms). Combatants who blur distinctions
between themselves and civilians transfer their risk onto civilians [23,24]. Endangerment of noncombatants can be a
byproduct of a method, as when guerrilla forces hide “among the people,”
taking the battlefield to civilians [23,24,26]. Endangerment can also be a direct goal. As
described by Viet Cong leaders [27] and American soldiers [23] in the Vietnam War, Viet Cong forces trained
children to throw grenades at South Vietnamese and American soldiers, partly to provoke
opponents to shoot children and bring shame to themselves and their force. Child soldiers
are more often killed or injured than adult soldiers, being deployed at the front line, to
lay or clear mines, or as suicide bombers because they provoke less suspicion
[3,28,29]. To illustrate the issue of variable access to valid data for DWI
applications, precise data for calculating child solider DWIs (Table 3) may be difficult to obtain for some conflicts.
However, DWIs for using child soldier suicide bombers (Tables 2 and 3) could be highly accurate due to extensive media coverage of suicide
attacks.
10.1371/journal.pmed.0050243.t004
The Northern Ireland Conflict, 1969–2001: Complementary DWI Analyses for
Unacceptable Aggression and Endangerment by Actors
DWI analysis can use any data source (media reports, epidemiological surveys, coroners'
reports) as long as the data are adequately valid, accurate, and comprehensive. DWI analysis
can be applied to event-based data or to aggregated data covering, for example, a year, a
phase, or a whole conflict. Analysis of all DWIs supported by good data provides fuller
description of a conflict and combatant behavior. A qualitative understanding of a
conflict's nature and context is necessary for DWI application and interpretation. When
possible, analysis should recognize when combatants avoid inflicting dirty outcomes, i.e.,
“clean” combat. DWIs suggest valuable data for prospective inclusion in
conflict monitoring.
Considerations
When DWIs are used to compare combatant groups or methods, it should not be assumed that
those with the highest values are simply the dirtiest. Nor should it be assumed that lower
DWI values “don't count.” A group may have a low DWI for recorded
civilian mortality, but high DWIs for assassinating civilian leaders and disappearances.
Another group may have low DWIs generally, and a very low DWI for torturing prisoners, but
torture breaches the precepts of humanity utterly so that to have a measurable rate at all
is deplorable.
DWIs reflect, in part, local conditions. For example, the lethality of civilian injuries
reflects local treatment technology and access. It may therefore seem incorrect to compare
DWIs for civilian lethality when health services differ. Similarly, it may seem unfair to
compare child mortality DWIs between a conflict where children comprise a large proportion
of the population and so are more likely to be killed and a conflict where children are few.
However, researchers should not adjust for such factors when comparing DWIs across settings.
This is because actors in armed conflict know, or are morally obliged to know, local
resources and demographics and their implications for civilian harm. Combatants are obliged
to take proportionately more care not to kill children when waging war in a child-dense
population. Responsibility for dirty outcomes is not ameliorated by local conditions.
As for any conflict analysis [2,3,30], DWI selection, application, and
interpretation must recognize the potential, varied biases of data sources and of particular
DWIs. Conflicts are highly politicized, and combatants, supporters, and detractors have
always tried to manipulate reports of war outcomes. Combatants may attempt to construct more
favorable DWIs not only by decreasing dirty combat, but by concealing dirty outcomes, or by
misrepresenting or provoking opponents' dirty outcomes. For example, a group might attempt
to raise an opponent's child mortality DWI by using child soldiers or children as human
shields.
Some DWI outcomes, such as injuries, may tend to be under-reported [16]. War-associated rape may be
difficult to measure due to stigma and under-reporting, though substantial reports exist
[21,24,31,32]. Although
bias can affect DWI values, as ratios DWIs are relatively less affected by under- or
over-counting than absolute numbers. For example, if a population generally under-reports
war-related rape by 40%, this does not bias comparing rates between different
combatant groups.
DWIs, complemented by absolute numbers, can suggest strategic aspects of actors' methods.
For example, systematic civilian targeting is suggested by combined findings of: many events
killing or injuring civilians; high ratios of civilian versus combatant mortality; frequent
use of methods causing high civilian casualties; frequent use of methods causing high
civilian lethality; and high rates of civilian harm from methods that are inherently
“targeted” (handguns, machetes). Such proportional and numerical
findings on civilian casualties have been used as evidence in International Criminal Court
trials to establish systematic patterns indicating war crimes [33].
DWIs Measure Outcomes, Not Justifications or Intentions
DWIs focus on whether the practice of war is just (jus in bello) and
ignore whether the reason for war is just (jus ad bellum), separating two
logically distinct moral issues in war [23]. We focus on practical outcomes because
justifications for war are contested, are used to legitimize dirty combat, and can bias
examination of war's impact [24,27,34–36]. Combatants and their supporters
may believe or describe methods as just, whether the method is suicide bombing
[24,37,38] or the World War II targeting of civilians by Germany and by the Allies
with carpet bombing, fire-bombing, and atomic bombs directed at cities [23,24,26,27].
Although intentions affect combat outcomes, such as civilian mortality rates
[16,27,34,35], we separate
DWIs from intentionality for the following reasons. Intentions are contested, obscured, and
distorted [3,35]. Dirty outcomes can result
from malicious intent, beneficent intent, or recklessness (lack of intent to take due care).
Frequently, combatant violence that appears wanton, sadistic, or vengeful (e.g., rape,
mutilation) is mobilized by political actors for hidden strategic aims [24,27,35,39,40]. Combatants' intended effects may be disrupted
by targets or adversaries [41]. Individual combatant behavior reflects overriding goals and
sociocultural aspects of larger groups [34,35,38].
Accommodating intentions or justifications in DWIs would imply that good intentions or a
“just war” attenuate responsibility for bad outcomes; an implication
that is morally and legally refuted [36]. DWIs therefore only recognize the crucial matter of outcomes: the
killing, injury, or abuse of individuals and populations who should be protected from
war.
Potential Deterrent Effect of the Dirty War Index
We choose the term “Dirty War Index” for three reasons. First, it
unites moral, humanitarian, and scientific values inherent to most armed conflict research.
Second, it avoids euphemisms that sanitize descriptions of war-induced public harm
[12,13,27,42]. Third,
emotional and cultural implications of “dirty” versus
“clean” may heighten the sensitivity of combatant groups to the index,
increasing its potential deterrent effect. No nation or combatant group wants to be
considered “dirty” or described as dirtier than others.
Increased accountability can have a deterrent effect in armed conflict and encourages
adherence to international humanitarian law; an important element in preventing violence
towards noncombatants [1,24]. DWIs increase scrutiny and
accountability specifically for dirty war methods. DWIs are analogous to corruption and
bribery indices used by nongovernmental organizations and the World Bank to improve
international governance through public monitoring and ranking governments by corruption
[43,44]. In Better: A Surgeon's Notes on
Performance [45], Atul Gawande describes how systematic analysis of war casualties
reveals problems and suggests solutions, and how identifying exemplary performers can
improve general performance. The DWI is developed for systematic, data-driven identification
of relatively good versus bad performance, heightening its potential to stimulate positive
change.
Military and political leaders not only want to win wars. They also seek superior moral
authority [23]. Moral
authority has social currency, creating better access to material resources, support, and
security within local and international communities. To improve behavior in combatants and
politicians insufficiently motivated by altruism, harnessing such self-interest is crucial.
Exposure of atrocities through DWIs can put reputation, legitimacy, future resources, threat
of retaliation, or power itself at stake [24].
As comparative rates, DWIs evoke the potential for change. The possibility of becoming
“cleaner” may appeal to some offenders [24]. Actors may compete for better outcomes
relative to military opponents, relative to in-group political competitors, or relative to
themselves over time. A DWI's potency can be increased by engagement with social, cultural,
and religious values of actors and their communities: honor versus dishonor [24], gaining versus losing
“face,” shame versus pride, dignity versus humiliation [37,46], sacred versus profane [37], and valuing mercy and the
lives of innocents [47]. Terms other than “Dirty War Index,” e.g., the
“Dishonorable War Index,” could be used to greater effect in different
contexts.
War and its destruction trigger emotions and self-interests that can obscure analysis by
threatening us so that we revert to familiar prejudices, reactions, and cognitive
frameworks. Through a public health approach using valid, precise proportional rates as
outcomes, DWIs can help us and our audiences to detach from political biases and break
through psychological denial when considering actors or methods in war. DWIs can present
conflict data from a new perspective, thereby encouraging actors in war to reassess their
combat methods, accountability, and interests.
Linked Perspectives
This Policy Forum is further discussed in two PLoS Medicine
Perspectives:
Taback N (2008) The Dirty War Index: Statistical issues, feasibility, and
interpretation. PLoS Med 5(12): e248. doi:10.1371/journal.pmed.0050248
Sondorp E (2008) A new tool for measuring the brutality of war. PLoS Med 5(12): e249.
doi:10.1371/journal.pmed.0050249
We thank Hamit Dardagan of Iraq Body Count for suggesting the mistreatment of captured
combatants as Dirty War Indices and the term “Dishonorable War Index,”
and for his helpful comments.
Provenance: Not commissioned; externally peer reviewed
Abbreviationsdf
degree of freedom
DWI
Dirty War Index
UXO
unexploded ordnance
ReferencesKrugEGDahlbergLLMercyJAZwiABLozanoReditors2002World report on violence and health. World Health
Organization.Available: http://www.who.int/violence_injury_prevention/violence/world_report/en/.
Accessed 4 November 2008MurrayCJLKingGLopezADTomijimaNKrugEG2002Armed conflict as a public health problem.324346349MackAeditor2008Human Security Brief 2007. Human Security Report Project.Available: http://www.humansecuritybrief.info/access.html. Accessed 4 November
2008CouplandR2007Security, insecurity and health.85181184CobeyJCRaymondNA2001Antipersonnel land minds: A vector for human suffering.134421422AnnasGJ1998Human rights and health—The Universal Declaration of Human Rights
at 50.33917781781FlanaginA2000Human rights in the biomedical literature: The social responsibility of
medical journals.284618619LevyBSSidelVWeditors20082nd editionNew YorkOxford University Press486van OmmerenMSharmaBPrasainDPoudyalBN2002Addressing human rights violations: A public mental health perspective on
helping torture survivors in Nepal.Inde JongJeditorNew YorkKluwer Academic259281ThomsONTRonJ2007Public health, conflict and human rights: Toward a collaborative research
agenda.111MullanyLCRichardsAKLeeCISuwanvanichkijVMaungC2007Population-based survey methods to quantify associations between human
rights violations and health outcomes among internally displaced persons in eastern
Burma.61908914McDonnellSMBoltonPSunderlandNBellowsBWhiteM2004The role of the applied epidemiologist in armed conflict.14NathansonV2000Preventing and limiting suffering should conflict break out: The role of
the medical profession.839601615Available: http://www.icrc.org/Web/Eng/siteeng0.nsf/html/57JQQ5. Accessed 4 November
2008International Committee of the Red Cross2008International humanitarian law.Available: http://www.icrc.org/Web/Eng/siteeng0.nsf/htmlall/ihl?OpenDocument.
Accessed 4 November 2008CouplandRM2001Armed violence.73337CouplandRMMeddingsDM1999Mortality associated with use of weapons in armed conflicts, wartime
atrocities, and civilian mass shootings: Literature review.319407410TabackNCouplandR2005Towards collation and modeling of the global cost of armed violence on
civilians.211927RestrepoJSpagatMVargasJF2004The dynamics of the Colombian civil conflict: A new data
set.21396428SpagatM2006Colombia's paramilitary DDR: Quiet and tentative success.Available: http://www.cerac.org.co/pdf/UNDP_DDR_V1.pdf. Accessed 4 November
2008StrausS2006‘Atrocity statistics' and other lessons from
Darfur.InTottenSMarkusenEeditorsNew YorkRoutledge189195TottenSMarkusenEeditors2006New YorkRoutledge284SchmittMN2005Precision attack and international humanitarian law.87445466WalzerM1977New YorkBasic Books361SlimH2007LondonHurst & Company319BilukhaOOTsitsaevZIbragimovRAndersonMBrennanM2006Epidemiology of injuries and deaths from landmines and unexploded ordnance
in Chechnya, 1994 through 2005.296516518SmithR2005LondonAllen Lane, Penguin Group428GrossmanD1995New YorkBack Bay Books/Little, Brown and Company366Coalition to Stop the Use of Child Soldiers2008Child soldiers.Available: http://www.child-soldiers.org/childsoldiers. Accessed 4 November
2008McKayS2005Girls as “weapons of terror” in Northern Uganda and
Sierra Leonean rebel fighting forces.28385397RomeuJL2008Statistical thinking and data analysis: Enhancing human rights
work.InAsherJBanksD ScheurenFJeditorsNew YorkSpringer6585KolbeARHutsonRA2006Human rights abuse and other criminal violations in Port-au-Prince, Haiti:
A random survey of households.368864873Physicians for Human Rights2002War-related sexual violence in Sierra Leone: A population-based
assessment.Available: http://physiciansforhumanrights.org/library/report-sierraleone-2000.html.
Accessed 4 November 2008SpirerHFSeltzerW2008Obtaining evidence for the International Criminal Court using data and
quantitative analysis.InAsherJBanksD ScheurenFJeditorsNew YorkSpringer195226AppyCG2006LondonEbury Press574KalyvasSN2006New YorkCambridge University Press485BugnionF2002Just wars, wars of aggression and international humanitarian
law.84523546AtranS2006The moral logic and growth of suicide terrorism.29127147BloomA2006Dying to kill: Motivations for suicide terrorism.InPedahzurAeditorLondonRoutledge2553RichardsP2004Portsmouth (NH)Heinemann198LeitesNWolfCJr1970ChicagoMarkham174HarrisonM2006Bombers and bystanders in suicide attacks in Israel,
2000–2003.29187206ZwiAB2004How should the health community respond to violent political
conflict.1e14doi:10.1371/journal.pmed.0010014Transparency International2008About Transparency International.Available: http://www.transparency.org/about_us. Accessed 4 November 2008World Bank2007Governance matters 2007: Worldwide governance indicators,
1996–2006.Available: http://info.worldbank.org/governance/wgi/index.asp. Accessed 4 November
2008GawandeA2007LondonProfile Books273LindnerE2006Westport (CT)Praeger Publishers224al-OadahS2007A Ramadan letter to Osama bin Laden.Available: http://www.islamtoday.com/printmenice.cfm?cat_id=29&sub_cat_id=1521.
Accessed 4 November 2008CouplandRM2005Ballistic trauma, armed violence and international law.InMahoneyPFRyanJMBrooksAJSchwabCWeditors2nd editionLondonSpringer-Verlag122134International Committee of the Red Cross2003Children in war: Summary table of IHL provisions specifically applicable to
children.Available: http://www.icrc.org/web/eng/siteeng0.nsf/html/5fflj5/$file/ang03_04a_tableaudih_total_logo.pdf?openelement.
Accessed 4 November 2008SpiegelPBSalamaP2000War and mortality in Kosovo, 1998–99: An epidemiological
testimony.35522042209SuttonM2002An index of deaths from the conflict in Ireland. CAIN Web
Service.Available: http://cain.ulst.ac.uk/sutton/index.html. Accessed 6 November
2008