2007 (CDC, n.d.a). (CDC, 2007 and 1999 15–24 between aged allraces of women for death of leadingcause second the reportedly was homicide example, for States, United the In women. against committed violence homicidal of gravity the tounderestimate however, take, amis- be would It 2009). ABS, 2010; FBI, 2009; (INEGI, inAustralia cent and29 per States, United inthe cent 23per inMexico, victims homicide of cent 10per approximately constitute women while homicide, of victims the of percentage high adisproportionately represent generally Men conflict. armed andfollowing during women against committed violence sexual as well awoman—as of killing ‘femicide’—the on focuses It women. against directed violence armed of patterns global unpacks chapter This wanting. found are generalizations crude these Yet inspection closer on households. headsof single as andoften survivors, victims, as ways: indifferent affected are said, is often it Women, killed. andbe tokill likely most to be lence conceal complex patterns of victimization victimization of patterns complex conceal lence vio- andsexual femicide of estimates statistical Yet 2005). BJS, 2008; Secretariat, Declaration (Geneva partners or former current frequently are women against violence serious of trators perpe- the violence; andsexual homicide trating inperpe- involved commonly more also are Men T especially young men—are determined determined men—are young especially Men— terms. ingeneral described often of he demographics When the Victim is aWoman Chapter 1

Four armed violence are are violence armed challenge for development. for challenge aserious women—constitutes against violence especially violence—and armed broadly, More 1978). (Pearce, insociety women marginalizing further poverty’, of ‘feminization tothe contribute can violence of aculture that evidence is growing Yet there considered. rarely are market labour wider and workforce female the on violence and sexual femicide of implications the contrast, In 2006). Florquin, 2008; Secretariat, Declaration (Geneva economy formal inthe men wage-earning of tivity produc- lost the highlight violence armed of costs andeconomic social the examining assessments Many andcommunity. family inthe role former victim’s the reflect that repercussions reaching far- generates it effects; knock-on traumatic with isatragedy individual any of killing violent The pp.12–13). 2008, (MOWA, marriage forced including violence, andsexual physical, logical, psycho- of forms tosevere isattributed women amongAfghan suicideandself-immolation reported in increase sharp The misery. their toend order suicidein commit also relationships and violent abusive endure who women Many and bystanders. witnesses, includingchildren, others, of lives the taking also suicidewhile commit sometimes tors perpetra- casualties; indirect frequently also are there iskilled, awoman when Moreover, comes. out- leadtofatal cases insome that aggressions escalating of culmination the isoften It incident. anisolated as occurs seldom women against violence that given especially and suffering, 113 2 3 4 5 1 WHEN THE VICTIM IS A WOMAN 114 While femicide and sexual violence are prevalent text of widespread tolerance of sexual violence in the domestic sphere, this chapter finds that and a high level of impunity (Peterman, Palermo, women are exposed to many other forms of vio- and Bredenkamp, 2011).2 Indeed, a study conducted lence—from gang violence to robberies and stray by the World Health Organization in selected bullets. They are also often singled out for hate countries reveals that adolescent girls frequently crimes, particularly when they are categorized as experience the practice of forced sex initiation migrants and refugees (Freedman and Jamal, 2008, (WHO, 2002, p. 153); this finding suggests wide- pp. 13–19). Guatemala and Mexico’s Ciudad Juarez spread acceptance and impunity of violence exhibit staggering numbers of women victims and against women and girls. exceptional brutality. In these contexts, intimate Femicide is an important component of armed vio- partner violence accounts for a relatively small lence and includes violence in the domestic sphere,

ARMED VIOLENCE 2011 VIOLENCE ARMED proportion of femicides (Suarez and Jordan, 2007); such as that perpetrated by intimate partners and of many women are victims of the increasingly wide- strangers. This chapter seeks to disaggregate the spread violence related to organized crime and demographics of armed violence and capture the narco-trafficking, which affects the entire popu- ways in which women of different ages are at risk. lation (Molloy, 2010). Furthermore, according to Specifically, the chapter finds that: a group of Mexican NGOs:

impunity and government permissiveness, which In the 111 countries and territories under review, an annual average of 44,000 women

GLOBAL BURDEN BURDEN GLOBAL serves as a crude expression of institutional violence, have led to a multiplication in the became homicide victims in 2004–09. number of women murdered throughout the Roughly 66,000 women are violently killed country and this can be attributed to a lack of around the world each year, accounting for due diligence (RNOCDH, 2010, p. 4). approximately 17 per cent of total intentional homicides. In such areas, an exclusive focus on femicide may limit the understanding of a broader picture of On average, men are killed approximately five extreme violence, which reveals major ‘systemic times more frequently than women. failures’ (Eriksson Baaz and Stern, 2010, p. 12; Femicides generally occur in the domestic A UNIFIED APPROACH). sphere; the perpetrator is the current or former partner in just under half of the cases. The social, cultural, and political risk factors for femicide and sexual violence are widely debated. Countries featuring high homicide rates in Analysts frequently point to cultures of ‘machismo’ the male population also typically experience that can distort traditional gender roles and high femicide rates. encourage constraints on the freedom of girls High levels of femicide are frequently accom- and women, misogynist behaviour, and recurring panied—and in some cases generated by—a violence with impunity. In many war zones, rape high level of tolerance for violence against has been used as ‘a weapon of mass destruction’ women. (UNDP, 2008, p. 2). Assessments of sexual violence In countries where violence is widespread, in the Democratic Republic of the Congo indicate the rate of victimization of women reaches that mass rapes and atrocities committed against levels far above the average risk of domestic women during periods of fighting occur in a con- violence. post-conflict settings. post-conflict and HIV/AIDS inconflict contracting of risk the as well as witchcraft, with associated practices and lethal killings, ritual killings, ‘honour’ deaths, includingdowry women, against violence lethal and non- lethal of forms other examines section third The victims. female involving homicide of anddynamics characteristics the considers tion sec- second The used. andinstruments offender, and victim between relationship the incidence, includingthe statistics, available of basis the on femicide of anoverview provides section first The violence. andsexual shapingfemicide and risks settings particular the considers chapter This remains scarce. remains data comparable internationally information, sex-disaggregated tomaintaining committed are countries of number agrowing Although occurred. event agiven inwhich context the on information orcritical victim the of sex the record isfactorily sat- not may andfiles reports police is pursued, suchanalysis when Even data. homicide overall tosupplement information additional generate and toidentify required be may analysts settings, inmany available readily not issimply such data Yet since sex. by statistics homicide gregating disag- isby women against perpetrated violence lethal of extent the measuring of way useful One sex by homicide Disaggregating defines femicide broadly as ‘any homicide with a with homicide ‘any as broadly femicide defines chapter This scope. geographical possible largest the covering femicide on statistics national piles Violence Armed of Burden Global 2011 The In some countries that exhibit low homicide homicide low exhibit that countries some In similar to that of malevictims. of tothat similar is victims female of percentage the rates, 3

com- the following: including violence, andsexual femicide on tics statis- disaggregated ingenerating challenges numerous face Countries standards. national inter- of short falls toviolence—still in relation life—including andprivate public of areas in many statistics gender andreliable valid of production the improvements, widespread Notwithstanding and parliamentary representation (UN, 2010a). (UN, representation and parliamentary employment, schoolenrolment, population, to relating sectors inkey available increasingly are statistics sex-disaggregated Fortunately, offenders. to andrespond violence toprevent measures of effectiveness the toassess which against indicators includingbaseline women, against violence of incidents on information objective of provision the for calls context This law. national inter- under violence from free tolive right their of inrecognition allwomen, on afocus for allows definition wider The sex. their of because females of targeting the suchas intent, of interpretation narrow anexclusive, avoiding thus victim’, female the reliance on mixed-quality data from dif- from data mixed-quality on reliance the leads statistics inavailable detail of lack the in toinvest andresources capacity limited standards international agreed of absence the concepts, basic of under-development the mines validity (Alvazzi del Frate, 2010). Frate, del (Alvazzi validity mines under- data) andmortuary andhealth sources, justice criminal andother police data, survey and administrative (suchas sources ferent to under-diagnosis; outputs; qualitative on inanover-reliance results collection data comparability; reduces systems and coding collection; data limits andmethods definitions,

115 2 3 4 5 1 WHEN THE VICTIM IS A WOMAN 116 Investments in generating accurate sex-disaggre- Box 4.1 Unpacking femicide: what’s in a label? gated data in the crime and justice sectors appear 5 When it was coined by the feminist movement in the 1970s, the term to be growing. Several initiatives have already ‘femicide’ implied the killing of women specifically because they were started to generate a wealth of data, albeit not women. As such, it was intended to convey ‘the misogynous killing of always comparable across time and space.6 women by men’ and to capture ‘the proportion of female deaths that Owing to the absence of agreed definitions, many occurred due to gender-based causes’ (Radford and Russell, 1992, p. 3; institutions and researchers collect data and Bloom, 2008, p. 178). The overall femicide concept emerged as an expres- develop datasets based on incompatible working sion of the feminist movement to politicize and contest male violence against women. definitions. The term ‘femicide’ is a case in point (see Box 4.1). Some analysts describe femicide as Diana Russell, an architect of the femicide concept, argues that the notion ‘the proportion of female deaths occurring due th ARMED VIOLENCE 2011 VIOLENCE ARMED has been in use for centuries. She traces its origins to early 19 -century

of to gender-based causes’ (Bloom, 2008, p. 178). Britain, when it was used to describe ‘the killing of a woman’ (Russell, 2008, p. 3).4 She also acknowledges that the term later emerged as a symbol of the Others contend that femicide refers to gender- battle to emancipate women and free them from violence in the 20th century. disaggregated data on homicide or ‘murder of women’, as indicated by the former Special Rap- From the beginning, the idea of femicide was designed to account for a range of specific forms of violence. These include dowry and ‘honour’ kill- porteur on Violence against Women, Its Causes ings, intimate partner or spousal violence, murder with rape, the killing of and Consequences (UN, 2008, p. 18). prostitutes, female infanticide or selective abortion, and other deaths that,

GLOBAL BURDEN BURDEN GLOBAL This chapter draws on a combination of data according to forensic reports, occur as a result of women or girls being sources to generate a profile of femicide and targeted on the basis of their sex (for example, victims of a serial killer who specifically targets women) (Bloom, 2008, p. 178). sexual violence. These sources include reports and surveillance data from national statistical While the concept has drawn attention to the particular ways in which institutions, law enforcement and public health women are selectively targeted, the definition has become progressively agencies, and research institutions. The resulting diluted and confused. Indeed, it is now often conflated with a broader understanding of violence against women (such as any killing of a woman) GBAV 2011 femicide database entails the follow- and has thus lost much of its original political connotation. The broaden- ing three sets of data, covering 111 countries and ing of the definition may be connected to a growing interest in generating territories—56 per cent of the world’s female quantitative information of violence against women to facilitate compara- population—for the period 2004–09:7 bility across countries and jurisdictions.

Today, most of the related literature applies a broad interpretation of femi- Femicides in general (homicides with female cide. Yet there are also several qualitative studies and data collection victims): drawing on the general femicide initiatives assessing femicide in the stricter sense, examining the intent of database alluded to above, information was the perpetrator and thus separating femicides from other killings of women. checked for consistency and categorized These projects are especially common in Latin America and include, for according to the classification of countries example, the Registro de Feminicidio del Ministerio Público, hosted by the presented in Chapter Two (TRENDS AND Crime Observatory of the Ministerio Público in Peru; the Observatorio de 8 Muertes Violentas de Mujeres y Femicidio in Honduras; the Observatorio PATTERNS). The examination therefore de Femicidios in Argentina; the Observatorio Ciudadano Nacional del includes a total of 104 countries or territo- Feminicidio in Juarez, Mexico; and the Banco Datos Feminicidio for Latin ries, providing information on most world America, based in Chile. regions (with the exception of Middle and Western Africa). global level. global the at year every killed women 66,000 mately approxi- yields estimate The data. no with regions two the for rate global the andusing information missing with countries of populations cide tothe femi- of rates regional applying by calculated was women against violence lethal of extent global The femicide?of 2004 and 2009. and 2004 between review under andterritories countries 104 inthe reported were femicides 44,000 about of annualtotal anaverage findsthat analysis The Two. inChapter presented data of that than andcomparable reliable isless information sex-disaggregated of quality overall the ards, stand- international relevant of absence to the Due excluded. been thus have andthey women, of deaths conflict-related direct on information insufficient was There femicide. torecord structed con- ever largest the isamong database femicide 2011 GBAV the femicide, of burden overall the under-representing andthus exhaustive not While male homicide victims than female ones. female than victims male homicide more times five are there average, on means that, ratio this population; total 100,000 per killed men five vs. woman one of inarate results This 4.1). Figure (see year inagiven deaths violent of number total the of cent 17 per approximately sent

tal’ violence, may be used in some settings. insome used be may tal’ violence, or‘mari- ‘spousal’ suchas definitions, varying however, partners; intimate orformer current by committed tohomicides refers dataset IPV the territories. and countries 54 of sample asmaller yielded dataset this (IPV), violence intimate-partner to attributable femicides on information of femicides: Intimate partner-related sources for 24 countries andterritories. countries 24 for sources of avariety from information collates set firearms: with committed Femicides 11 Female victims thus appear torepre- appear thus victims Female 9 In principle, all data included in included alldata principle, In 10 What, then, is the global extent extent global isthe then, What, consisting consisting this data- this vs. male homicide victims per year, 2004–09 per victims malehomicide vs. Legend: Figure assessments of interventions. assessments and policy-making effective more for allow will data accurate that conviction the on based Africa across systems instatistical invested have cies agen- development of asmallnumber however, Goals, Development Millennium the of lishment estab- Sincethe PATTERNS). AND (TRENDS rates homicide high featuring those of including many countries, andAsian African most for unavailable is information andsex-disaggregated common are gaps major data shows, mapclearly the As isavailable. information reliable which for tries incoun- rates femicide specific reveals 4.1 Map Sources surveillance capacity can be established. be can capacity surveillance acomprehensive before taken be still must steps major developed, being are systems monitoring andnon-governmental public both that shows Twenty-two countries feature ‘low’ rates (1–2) rates ‘low’ feature countries Twenty-two population). female 100,000 (<1 per rates low’ ‘very exhibit 41countries findsthat It available. is data which for andterritories in 104countries population female 100,000 per femicide of sity inten- the of distribution the considers 4.2 Figure and Europe. Americas the from isavailable coverage best the Female victims (66,000; 17%) Male victims (330,000; 83%) 4.1 GBAV 2011 database and femicide database femicide and database 2011 : GBAV

Estimated average proportion of female proportion average Estimated 12 While evidence evidence While 13 To date, To date, 117 2 3 4 5 1 WHEN THE VICTIM IS A WOMAN 118 Map 4.1 Rates of female homicide victims per 100,000 female population, 2004–09 ARMED VIOLENCE 2011 VIOLENCE ARMED of

LEGEND: Per 100,00 women >6 3–6 2–3 GLOBAL BURDEN BURDEN GLOBAL 1–2 <1 No data

Source: GBAV 2011 femicide database

Figure 4.2 Countries and territories by rate of femicide per 100,000 while 16 countries have ‘medium’ rates (2–3). female population, 2004–09 The figure shows, however, that 13 countries suffer from ‘high’ rates (3–6) and another 12 Number of countries countries report a ‘very high’ intensity of femicide 90 (>6). All in all, 25 countries, or nearly one-quarter 80 of all reviewed countries, exhibit ‘high’ or ‘very 70 high’ rates of femicide—that is, rates above the 60 global average. The sum of femicides committed 50 in these 25 countries represents almost half 40 (47 per cent) of the total number of femicides in 30 the database. 20 At the regional level, the highest rates of femi- 10 cide are found in regions that also exhibit very

Very low (<1) Low (1–2) Medium (2–3) High (3–6) Very high (>6) high rates of homicide. Figure 4.3 compares the Rate of femicide per 100,000 female population average rates of total homicide and femicide Source: GBAV 2011 femicide database across countries pooled by region.14 Spreading Western, and Southern European countries. countries. European andSouthern Western, Northern, of majority inthe rates homicide overall than higher times up tofive are rates femicide the countries, these allof In femicide. of rates high extremely register andHonduras Guyana addition, In PATTERNS). AND (TRENDS ingeneral violence andlethal femicide inboth first ranks Salvador El Indeed, orabove. population female 100,000 per 10femicides of rates allreport Africa and South Jamaica, Guatemala, Salvador, El level, country the At homicide. of rates high tofeature tend also femicide of rates high exhibit that Countries Femicide bycountry rates Africa, order—Southern in descending namely— ranking, femicide the of top the at ture fea- Two also inChapter rates homicide highest the with regions five of out Four higher. are rates where larger appear that discrepancies shows tion popula- total the across femicides of number the (CHARACTERISTICS OF ARMED VIOLENCE). ARMED OF (CHARACTERISTICS 1:1 approaches therefore ratio victim to-female male- resulting The victims. the frequently are women inwhich violence ordomestic passion’ of ‘crimes of proportion includeahigher ably prob- ingeneral—homicides violent less thus are that rates—and homicide low exhibit that tries incoun- that be may explanation One population. total 100,000 per 0.4 only by rate femicide the exceeds that rate ahomicide with difference, est small- the records Europe Western isminor; rates andhomicide femicide between difference the (<1 100,000), per rates low very with regions In list. the of bottom the at appears Africa Northern femicide, of rates lower exhibit that regions Among ingeneral. homicides to respect with high disproportionally appear Europe inEastern rates femicide meanwhile, America; andCentral Caribbean, the America, 15 South South rate of 5.0 per 100,000. per 5.0 of rate ahigh has Belize 141,000, approximately of tion popula- andafemale homicide of victims female seven of count atotal with example, For rate. femicide high arelatively generate women of ings kill- afew even smallcountries, andother these In individuals. 500,000 than fewer of populations hometofemale andGuyana—are Belize, Bahamas, the as countries—such Some inAfrica. and one inAsia are three countries, remaining the Among Europe. inEastern andfour inNorthern three inEurope: located are rates femicide high or very high reporting countries Seven 4.4). Figure (see America inSouth andsix America, in Central four Caribbean, in the four Americas: inthe are (14) cent 50per than more rates, femicide high andvery high feature that 25 countries the Of Northern Africa Northern Southern Europe Asia Eastern Oceania Europe Northern Europe Western Asia Southern America Northern Eastern Africa Asia South-Eastern Asia Western Asia Central America Central Caribbean South America Eastern Europe Southern Africa 2004–09 region, by population, Figure Source Femicide Femicide GBAV 2011 database and femicide database (unweighted regional averages) regional (unweighted database femicide and database 2011 : GBAV 4.3 Total homicide

Average homicide and femicide rates per 100,000 total total 100,000 per rates andfemicide homicide Average 0 16 51015202530 40 Rate per 100,000 total population 54 50 45 35 119 2 3 4 5 1 WHEN THE VICTIM IS A WOMAN 120 Figure 4.4 Average femicide rates per 100,000 female population in 25 countries and territories with high and very high rates, 2004–09

El Salvador

Jamaica

Guatemala

South Africa

Russian Federation

Guyana

Honduras

Azerbaijan

Lesser Antilles Region ARMED VIOLENCE 2011 VIOLENCE ARMED

of Colombia

Bolivia Very high Bahamas

Lithuania High Belarus

Venezuela

GLOBAL BURDEN BURDEN GLOBAL Latvia

Belize

Kazakhstan

Moldova

Brazil

Kyrgyzstan

Ukraine

Ecuador

Dominican Republic

Estonia

0123456 7 8 91011 12 Rate per 100,000 female population

Source: GBAV 2011 femicide database

Countries in Latin America and the Caribbean on femicide at the sub-national level reveals even appear to share a particularly high incidence of higher rates than those recorded as national aver- femicide. A comparison reveals that not only young ages. In Mexico’s Ciudad Juarez, for example, the men, but also a disproportionately high number 2009 femicide rate was 19.1 per 100,000 female of women and girls, are dying in high numbers. population (see Box 4.2). In the state of Espirito Killings appear to be most frequent in specific Santo in Brazil, the government authorities docu- states and cities of these countries. Indeed, data mented a femicide rate of 10.9 per 100,000 in 2008.17 Source 5 10 15 20 25 30 Rate per 100,000 female population Figure Ciudad Juarez Ciudad Juarez 9319 95199 1995 1994 1993 level of savagery often confined to war zones. war to confined often savagery of level aparticular betrays killed, being before tortured are whom of many women, of executions brutal the Nevertheless, victimization. male of rate the in trend asimilar by ismirrored 2009 in observed peak the 2008, to 2007 from climb a radical Following 2009. 1993– during levels city and national the at femicide in Source: Small Arms Survey elaboration of Arroyo (2011) Arroyo of elaboration Survey Arms Small Source: serious human rights crisis. a constitutes 2009—and and 1990 between 669 counts issignificant—INEGI city this in killed girls and women of number The 2009). (INEGI, 2009 in population 100,000 per homicides 170 some corded re- It Mexico. in rates homicide highest the of one exhibits currently state Chihuahua in Juarez Ciudad the national average. above also were (3.7) Nayarit and (3.8), Michoacán (3.8), Tamaulipas Oaxaca (6.0), (4.5), (5.1), Sonora (7.1), Sinaloa Durango in (10.1). Rates Guerrero and (10.1), California Baja population), female 100,000 per (13.1 Chihuahua average: national the above well rates exhibited states three 2009 in violence; escalating by affected equally are Mexico of areas all Not 2009. by 100,000 per 3.5 to increase rapid a by 2007, in followed population female 100,000 per 2.0 of low ahistorical at rates femicide reported (INEGI)— yGeografía Estadística de Nacional Instituto Mexico—the of authority statistical national The crisis a humanrights B Femicide in Ciudad Juarez: Juarez: inCiudad Femicide ox 4.2 Small Arms Survey elaboration of Arroyo (2011) (2011) Arroyo of elaboration Survey Arms : Small 4.5

Trends in femicide per 100,000 female population in Ciudad Juarez and Mexico, 1993–2009 1993–2009 andMexico, Juarez inCiudad population female 100,000 per Trends infemicide Mexico 9719 1999 1998 1997 6 18 Figure 4.5 shows trends trends 4.5 shows Figure 0020 0220 0420 0620 082009 2008 2007 2006 2005 2004 2003 2002 2001 2000 Mexico, June 2009. © Teun Voeten/Panos Pictures Voeten/Panos ©Teun 2009. June Mexico, Juarez, Ciudad in women of disappearance and murder the Photo Armed police stand behind pink crosses marking marking crosses pink behind stand police Armed 121 2 3 4 5 1 WHEN THE VICTIM IS A WOMAN 122 As noted above, high femicide rates are often Fears of reprisal during the war partly explain the accompanied by high levels of tolerance to vio- silence regarding contemporary forms of violence lence against women among the wider population. and crime, including violence against women. Analysts claim that such behaviour is shaped by Certain members of the Salvadoran state had levels of gender inequality and norms that discrimi- employed local militia to perpetrate violence nate against the status of women—norms that are during the civil war, highlighting how conflict- often shared by women themselves. One widely related violence can affect the domestic and cited study reports that about one in four women private spheres. Thus: surveyed in 33 countries agreed that it was justi- despite important legislative changes, normative fiable to be hit or beaten for arguing or refusing to notions of appropriate behavior for men and have sex with one’s husband (UN, 2010a, p. 137). women still make violence ‘acceptable’ in certain ARMED VIOLENCE 2011 VIOLENCE ARMED

of Additional research has demonstrated that in contexts to the point that it is not always recog- settings with high rates of femicide the criminal nized as violence (Hume, 2008, p. 64). justice response may be substantially slower and less efficient than for homicide more generally. Trends in femicide Cases may not be investigated and consequently It is possible to examine time-series trends in not prosecuted, resulting in very low clearance femicide and violence against women for coun- and conviction rates. For example, a recent study 21

GLOBAL BURDEN BURDEN GLOBAL tries with available data. Attributing the value finds that between 2008 and 2010 in Honduras, 100 as a starting point for all countries in 2004 only 211 of 1,010 reported cases of femicide were allows several different patterns emerge. These heard in court and only 56 sentences were passed include countries experiencing rapid increases, (Sánchez, 2011, p. 40). Similar patterns have rapid decreases, fluctuating trends, and more emerged in other regions where femicide is poorly stable rates. Among 22 countries with trend investigated and countered.19 series available, only seven exceeded the starting Armed conflicts in particular can reconfigure point rate in 2008 or 2009; the countries are—in gender relations. Research tracking the popular descending order of femicide rates—Honduras, usage of violence in everyday speech in post- Mexico, Finland, El Salvador, Azerbaijan, Guatemala, conflict El Salvador identifies men as having and the Dominican Republic. In some countries— ‘more “right” to use violence than women’ such as Azerbaijan and Finland—rates that had (Hume, 2008, p. 66). Indeed, prolonged repres- initially climbed may have dropped towards the sion and everyday violence affect the lives of end of the observed period. The majority of coun- many long after a conflict ends. In particular, tries under review show stable or decreasing Hume finds that: rates. Figure 4.6 displays trends in femicide rates for a sample of five countries with data for the individuals and communities learned that silence entire 2004–09 period. was the only option when no one could be trusted and violence was an ever-present possibility. They testified to feeling afraid of the orejas Male versus female victims (informers), who were often neighbors or family Available data shows that there is a correlation members (Hume, 2008, p. 71). between the rates of homicides with male victims dor, order. inthat Salva- and El Guatemala, Colombia, Venezuela, in observed being differences largest the with increase, violence overall of levels the when pan. andJa- Norway, Switzerland, inAustria, such as rare, isrelatively homicide where case the also is It Darussalam. Brunei suchas populations, male large disproportionately with countries in case the ismostly This men. for observed tothat close very appears women for rate victimization rates—the homicide low presenting those countries—especially insome Indeed, and those with female victims. female with and those the population structure of eachcountry. of structure population the reflecting partly variation, considerable shows ineachcountry ratio victim homicide to-female male- the of inspection adeeper Nevertheless, tomalevictims. respect with rates homicide high show which Islands, Virgin US andthe Venezuela, inColombia, observed be can opposite the while rates, femicide high tively rela- feature andGuyana Federation Russian The victims. maleandfemale toboth respect cide with homi- of rates highest the exhibit and Guatemala Salvador El femicide. of rates high report also victims malehomicide of levels high featuring tries Source 50 100 150 200 250 300 Femicide rate (index year 2004=100) Figure El Salvador 2004 24 GBAV 2011 database : GBAV femicide This gender gap becomes more significant significant more becomes gap gender This 4.6

Guatemala Femicide trends in five selected countries, 2004–09 (basis 2004=100) (basis 2004–09 countries, selected infive trends Femicide Honduras Honduras 2005 22 Mexico In general, coun- general, In United States 23

2006 countries. Even if the proportion of women killed killed women of proportion ifthe Even countries. toother compared victims homicide female maleand of proportions higher have also rates homicide highest the exhibit that countries The review. under period the during highest the were rates femicide inwhich countries the also were they that given elsewhere, than countries inthese safer were women meanthat not does This ones. female of proportion the than greater times ten than more is victims malehomicide of proportion the Brazil, and Rico, Puerto Colombia, inVenezuela, example, For ismuchhigher. malevictims of proportion andthe ishigher, sexes between disparity the higher, are rates homicide Where killed. women of proportion andthe rates homicide between tion correla- isanegative There rates. high very with inthose cent and12.5 per rates high with tries 16.3 men: incoun- for cent per rates homicide of afraction just represent rates femicide cide rates, homi- high andvery high experiencing countries in that shows also 4.7 Yet 4.7). Figure Figure (see men for rates homicide than lower one-third just average, on are, rates femicide countries, these In smallest. appears andwomen men for rates between difference the population, total cide per homi- of rates low very feature that countries In 2007 20 2008 2009 123 2 3 4 5 1 WHEN THE VICTIM IS A WOMAN 124 Figure 4.7 Number of femicide victims for every 100 male homicide while female perpetrators represent a residual victims in 83 countries, 2004–09 portion.26 Since femicide often occurs in the family or in restricted circles close to the victim, the per- Number of femicides for every 100 male homicide victims petrator is likely to be found in most cases. Often 90 there is a previous connection with the victim, 80 most frequently at the family or intimate partner 70 level. The broad category of intimate partners 60 includes all sexual partners—such as current 50 and former spouses or partners—as well as other 40 close family members if involved in an intimate 30 relationship with the victim. Non-intimate part-

ARMED VIOLENCE 2011 VIOLENCE ARMED 20 ners include friends and acquaintances, as well of 10 as known strangers and family members.

Very low (<1) Low (1–3) Medium (3–10) High (10–30) Very high (>30) The perpetrator remains unknown if the investiga- 13 countries 30 countries 20 countries 13 countries 7 countries tion does not succeed in identifying the offender, Homicide rate per 100,000 male population preventing the case from being ‘solved’. This is Sources: GBAV 2011 database and femicide database frequently the case with respect to deaths occur- ring between armed groups, during armed con- GLOBAL BURDEN BURDEN GLOBAL may appear small with respect to that of male flicts, and as a result of robberies, gang activity, homicide victims within a country, the actual shootings in public places, and other incidents femicide rate remains high. A possible explana- of a similar nature. In addition, the capacity and tion for these sustained high rates is the involve- resources necessary for effective investigations may ment of organized crime and gangs in committing also have an impact on the number of ‘unsolved’ violent acts against women and girls, especially cases (CHARACTERISTICS OF ARMED VIOLENCE). femicide.25 This is not the case in the majority Some researchers contend that women who kill of countries, where femicides apparently occur themselves in the context of protracted violence predominantly within the family or in the context or aggression should be counted as victims of of an intimate relationship (CHARACTERISTICS femicide.27 Indeed, the category of ‘forced suicide’28 OF ARMED VIOLENCE). is the frequent outcome of strong social pressure and ‘honour’-based violence, such as reported in Iraqi Kurdistan (KWRW, 2008). The consequences Characteristics of femicide of structural repression and social stigmatization in such settings can lead to enduring physical In order to better understand the factors shaping harm and serious psychological hardship.29 femicide it is important to disaggregate the actors, causes, and circumstances shaping the killing of Another lethal scenario is the killing of a rape women. Specifically, the characteristics of the victim in order to restore the family ‘honour’. For perpetrator are a central component of this clas- example, in Libya women and girls who become sification system. It is widely accepted that male pregnant through rape run the risk of being mur- perpetrators comprise the vast majority of offenders dered by a family member in so-called ‘honour’ United Nations Assistance Mission in Afghanistan: in Mission Assistance Nations United the by reported As killed. and harassed, threatened, been have positions public in women Numerous roles. traditional their from away breaking as seen are who women target persistently groups Armed 2007). (Azarbaijani-Moghaddam, society civil astrong of absence the in change social meaningful about bring to struggling been since have agencies Both 2001. in Conference Bonn the at Affairs Women’s of Ministry the and Commission Rights Human Independent Afghanistan the of creation the by evidenced as rights, women’s regarding preoccupation international growing witnessed has period Taliban post- The impunity. and widespread discrimination, conflict, armed of decades by hampered been has girls and women Afghan of security personal The B school, especially daughters’ (MOWA, 2008, p.11). 2008, (MOWA, daughters’ especially school, to children their sending fear ‘parents attacks, such of aresult As destruction. total and burning, including alone, 2007 looting, their and February 2005 July between groups insurgent by schools on 192 attacks recorded Affairs Women’s of Ministry The violence: with targeted explicitly also are girls Young p.11). 2009, violence (UNAMA, domestic women dealing with police- ten of aunit leading was she because allegedly 2008, in September killed in Kandahar, was officer police female highest-ranking the Kakar, Malalai ox 4.3 flee the country (UNAMA, 2009, p.10). 2009, (UNAMA, country the flee to had one and murdered been have three country, their reconstructing in arole take to wanted who women Afghan of stories covering in interested anewspaper by 2005 in interviewed women high-profile five Of 2011, p. 15). p. 15). 2011, (UNESCO, of Afghanistan in province Helmand schools of 70% over of closure the in resulted have fears Security particular. in schools girls’ and general in infrastructure education attacked repeatedly have groups Insurgent Insecurity and impunity inAfghanistan andimpunity Insecurity 30 however. victim, the to rights few provide mechanisms These reports: Mission UN the As intercourse. sexual extra-marital of because prosecution including criminal victimization, further risks arape reports who woman Any cally. specifi- rape and general in women against violence of incidents to widespread factor contributing amajor isconsidered impunity of level high The 21). (p. husband the of member afamily by raped being of risk at cially espe- are marriages forced in girls young very that suggests research the Nevertheless, p.21). 2009, (UNAMA, marriage to relation in especially lence’, vio- ‘sexual or ‘rape’ as such terms use to difficult’ ‘extremely it found Mission Assistance UN the stan, women oninof violenceAfghani- research against suggesting significant undercounting. In the course reported, frequently less were rape of Incidents 19). 1–2, pp. 2006, (UNIFEM, were physical followed attack, by forced marriage violence of incidents common most The cent). (82 per cousin or son, father-in-law, ahusband, as such them, to close someone by and homes their within perpetrated were women against attacks most that highlights data the of analysis An respondents. 818 from cases 1,327 recorded project the 2005, June and 2003 January Between n.d.). Afghanistan, (UNIFEM women against violence on adatabase established Afghanistan, together with participating agencies, participating with together Afghanistan, UNIFEM 2006, In women. against directed lence vio- domestic extensive from suffers Afghanistan p.8). 2011, (UNESCO, boys 100 every for enrolled are girls 66 only Today, 2009, p. 25). or ajirga through level, community the at or privately case the resolve to attempt often families rape, of stigma social the and society Afghan Coupled with the conservative nature of shura shura [a council of elders] (UNAMA, [a council of elders] (UNAMA, [a gathering of tribal elders] elders] tribal of [a gathering 31

125 2 3 4 5 1 WHEN THE VICTIM IS A WOMAN 126 ARMED VIOLENCE 2011 VIOLENCE ARMED of GLOBAL BURDEN BURDEN GLOBAL © Ash Sweeting/Panos Pictures Sweeting/Panos © Ash assassinated violence. was Kakar domestice of victims from evidence gathers Academy, Police Photo Malalai Kakar, the first woman to graduate from Kandahar Kandahar from graduate to woman first the Kakar, Malalai by the Taliban in 2008. these cases to: attributes largely Commission 4.3).Box The (see inAfghanistan andwomen girls among infrequent isnot self-immolation Commission, Rights Human Independent Afghanistan to the according example, For consequences. lasting ates gener- andgirls women of repression place, take not may suicideandfemicide forced where Even 1998; Faqir, (Ruggi, 2001). ‘honour’ family the of indefence killed be may victim rape a that confirm studies 2011). Other (Harter, killings includes suicide as a key category among seven: among category akey suicideas includes andZimbabwe— Zambia, Swaziland, Africa, South Africa—Botswana, in Southern countries infive out carried astudy from femicide of A classification assumed. often than complex more suicide are and femicide between connections the Indeed, pp. 9–10). 2010, (Psytel, sample inthe women the of cent per 42 of death cause the suicideas identifies violence spousal of context inthe femicides and homicides of a sample considers that study Union European 1996). Arecent and Flitcraft, (Stark relationship abusive their of termination the orafter during point some at suicide attempt one least madeat violence ofdomestic tims vic- of cent and40per 35 between that reveals States United inthe undertaken Research flict. con- orarmed violence homicidal chronic with confronted tocountries limited suicide isnot and violence domestic between correlation The price) and family problems (AIHRC, 2004, p.32). 2004, (AIHRC, problems family and price) as such practices customary war, of 25 years of impact psychological the rights, women’s of awareness societal of lack the tices, prac- discriminatory other as well as marriages forced marriage, premature marriage and multiple Tuyana (bride 127 2 3 4 5 1 WHEN THE VICTIM IS A WOMAN 128 intimate femicide involving women killed by IPV-related femicides current and former partners; Femicide is often linked to situations involving femicide by other (male) family members; intimate partner violence, between either spouses sexual femicide such as rape followed by or partners. An intimate partner may be defined murder; as a person with whom the victim had a physical intimate relationship, either at the time of the witch femicides; femicide or in the past. In this context, it is irrel- ritualistic femicides involving women and girls evant whether they were ever married or lived ritually killed to cut out their genital organs; together. For this reason, the term ‘intimate partner women killed by thieves or robbers; and violence’ is more apt than ‘spousal’ or ‘marital’ Photo T A victim of domestic violence takes suicides by women experiencing violence, violence. All these categories, however, may be ARMED VIOLENCE 2011 VIOLENCE ARMED

of shelter at a safe house in including cases in which there is strong evi- captured in the wider definition of ‘domestic vio- Nevada County, California, dence that women kill themselves to escape lence’. As noted in Chapter Three, the proportion August 2010. © Rich Pedroncelli/ intolerable levels of violence (Watts, Osam, of homicides occurring in the domestic sphere is AP Photo and Win, 2001, p. 91). different for males and females and is not the same GLOBAL BURDEN BURDEN GLOBAL 100,000 population, 2008 per andrate sex by States, United the in partner anintimate by committed the victim. the over power exercise or dominate, coerce, to violence use who siblings other and parents by committed to include acts couples butextends ex- or couples to limited isnot which violence, domestic of apart considered be may IPV lence. vio- severe more to serious less from escalating often incident, asingle than rather a pattern as represented isfrequently IPV 2002). (WHO, couples same-sex in other the against partner one by and men against women by violence of cases to refers also it women; against men by committed violence to limited isnot IPV partners. dating and partners, live-in spouses, former and current to refers generally partner’ ‘Intimate Figure 2005). (BJS, male victims of with cent homicides by anper intimatefive partner, comparedonly to are perpetrated femicides of cent per 30 about Furthermore, is80. men for rate the while is 430, women for rate the population, 100,000 per 260 is rate total the while average; the to respect with victims IPV the of share large a disproportionately represent women 4.8, Figure in shown as Indeed, Figure 4.8). (see partner intimate an by committed crime of victims be to men than likely more times five than more are States United the in women tics, Statis- Justice of Bureau US the to According IPV. of risk at group largest the represent Women femicide. to factor contributing amajor as violence partner intimate identify studies Many B aevcis Femalevictims Male victims Intimate partner violence Intimate partner ox 4.4 Homicide rate per 100 4.8 32

Victims of violent crime crime of violent Victims 100 200 300 400 500 ,000 malepopulation Malta Poland Sweden France Portugal Cyprus Denmark Latvia Italy Austria Serbia Finland Puerto Rico Switzerland Republic Czech Luxembourg Croatia New Zealand UK (England, Wales) Guatemala Canada Belgium Romania Hungary Ecuador Republic Dominican Peru United States Greece Norway Slovakia UK (Scotland) Australia Nicaragua Bulgaria Region Lesser Antilles Panama Chile Uruguay Slovenia Netherlands Argentina Estonia Germany Russian Federation Paraguay Spain Costa Rica Honduras Ireland Mexico Lithuania Colombia El Salvador Figure Source 2004–09 country, per femicides IPV-related of percentage estimated GBAV 2011 database : GBAV femicide 4.9

Total femicide rates per 100,000 female population and population female 100,000 per rates Total femicide 0 1020304050607080900 Total femiciderate per 100,000 f 5 Percentage ofIPV-related femicides 10 emale population (based on average) Percentage of IPV-related femicides femicides IPV-related of Percentage Total femicide rate rate femicide Total 15 20 100 25 129 2 3 4 5 1 WHEN THE VICTIM IS A WOMAN 130 in all countries. Statistically, women run a much being physically abused by the same perpetrator higher risk of being killed by their partners than (McFarlane et al., 1999; Campbell et al., 2003). do men (WHO, 2002; UNIFEM, 2006; see Box 4.4). Understanding these and other risk factors is critical when considering measures to prevent Figure 4.9 shows total rates of femicide and the and reduce femicide. estimated percentage of IPV-related femicides based on the sample of 54 countries and territo- As noted above, the total number of femicide ries for which data is available. There is a weak victims may be higher if suicides and indirect negative correlation between the two indicators casualties are also included. For example, chil- (–0.431, n=54), reflecting that the higher the rate dren, relatives, or other witnesses may be killed of femicide, the lower the proportion of IPV-related as a result of IPV. A study carried out in the 27 femicide. This corroborates the conclusion that countries of the European Union analyses 3,413 ARMED VIOLENCE 2011 VIOLENCE ARMED

of higher rates of victimization of women may accom- deaths resulting from domestic and spousal con- pany widespread violence rather than domestic flicts in 2006, taking into account both male and or intimate partner violence. In countries where female victims (Psytel, 2010, p. 9; see Figure 4.10). femicide rates are high, women also run a higher The findings are instructive: risk of becoming targets of violence outside the private sphere. Incidents of femicide were most common, accounting for 41 per cent of all cases. Violence committed by current and former intimate GLOBAL BURDEN BURDEN GLOBAL partners represents a risk across all countries and Some 8 per cent of the victims were men killed cultures. Strikingly, the vast majority of women by their female partners. killed in the context of IPV had previous experi- Women who committed suicide after enduring ence of domestic violence or stalking, including long-lasting domestic or spousal violence represent 30 per cent of the sample. Figure 4.10 Deaths of men and women as a In 16 per cent of the cases, perpetrators killed consequence of domestic violence in the themselves in ‘homicide–suicide’ scenarios. European Union, 2006 Collateral victims who happened to be involved Legend: in domestic or spousal incidents represent Male homicide victims (272; 8%) Male collateral homicide victims (186; 5%) 5 per cent of the sample. Male suicides after femicide (536; 16%) Femicide (1,409; 41%) In some cases—known as ‘extended suicide’, Female suicides after violence (1,010; 30%) ‘familicide’, or ‘homicide–suicide’—the perpetrator may kill his or her entire family before commit- ting suicide. The vast majority of these incidents involve a male perpetrator who commits ‘femicide– suicide’ or, in some cases, ‘femicide and attempted suicide’; they tend to take place in the home and guns represent the instrument of choice (Auchter, 2010; Liem and Nieuwbeerta, 2010). Homicide– suicide represents an important component of Source: Small Arms Survey elaboration of Psytel (2010) lethal domestic violence. Research on this issue arm is more likely to be used to threaten and injure andinjure tothreaten used tobe likely ismore arm afire- indeed, andgirls; women against violence lethal facilitate also may home inthe weapon a of presence the 2009). Likewise, and Glass, Webster, (Campbell, afirearm with threatened being reported hadpreviously femicide of victims Many use. actual of isa predictor firearms displaying p.17). 1998, (Meloy, Nevertheless, victims’ injure than rather andintimidate control ‘to them and use weapons own may stalkers Specifically, stalking. of case inthe present frequently also are Weapons offenders. ormore one with tocomply victims tocoerce used be may Firearms 2000). Azrael, and (Hemenway victims to shoot than rather tothreaten merely used often more are Firearms 2008). al., et (Mathews higher suicideisalso commit will femicide of perpetrator the that probability the isused, arm afire- when inIPV. Moreover, especially women, against inviolence role animportant play firearms Nevertheless, frequent. isless infemicide arms fire- of use the men, involving in homicides used frequently are firearms While p.5). 2008, Secretariat, Declaration (Geneva afirearm with committed are homicides of cent per 60 approximately that mates Violence Armed of Burden Global 2008 The in femicides offirearms use The 2010). andMathews, Jewkes, Abrahams, 2010; al., et (Roberts murder the of aweek suicide within commit femicide of perpetrators of 19 cent per mately approxi- that highlighting countries, in high-income found tothose similar patterns identified has Africa inSouth Yet research countries. middle-income and low- from isavailable muchinformation Not 2003). andLehtia, Kivivuoria 2008; (Oberwittler, society of strata lower inthe than middleclass the among frequent ismore incident of type this where States, United andthe inEurope advanced is most esti- by rate of femicide per 100,000 female population. population. female 100,000 per femicide of rate by firearms with committed femicides of distribution average the shows 4.11 Figure afirearm. of use the involved in2004–09 perpetrated femicides of cent per 60 than andHonduras—more dor, Guatemala, Salva- El Colombia, Brazil, as rates—such homicide high with incountries contrast, of way By VIOLENCE). ARMED OF 3.7, CHARACTERISTICS Figure (see firearms of use rare the with correspond quently fre- rates homicide Low firearms. with committed femicides of percentage andthe rates femicide between correlation isadirect there available, is data adequate which for countries 24 the In (RCMP, 2010). period asix-year over killed women of number the of astudy on based tragically, ending disputes domestic of risk the reduced has inhouseholds tofirearms access limiting policy gun that demonstrated has study p.7). ACanadian 2011, (Hemenway, intruders from home the toprotect than members family Source 10 20 30 40 50 60 70 80 90 100 Percentage offemicidescommitted with afirearm 2004–09 rate, femicide by countries, 24 Figure GBAV 2011 database : GBAV femicide Very low(<1) 6 countries 4.11

Average percentage of femicides committed with firearms in firearms with committed femicides of percentage Average 7 countries Low (1–2) Femicide rate per 100,000 femalepopulation Medium (2–3) 4 countries 2 countries High (3–6) Very high(>6) 5 countries 131 2 3 4 5 1 WHEN THE VICTIM IS A WOMAN 132 Figure 4.11 also reveals that the use of firearms is the national average for both male and female much more frequent in countries that exhibit very victims. In 2007–09, the percentage of femicides high femicide rates. Among all countries under committed with firearms increased dramatically, review, an average of one-third of all femicides reaching 82 per cent in 2009 and thus nearing were committed with firearms. the percentage observed in homicides with male Figure 4.12 highlights the difference between the victims (89 per cent). use of firearms in homicides with male vs. female In the United States, data from the Centers for victims in Mexico on the whole and in Ciudad Juarez Disease Control and Prevention shows that firearms from 1993 to 2009. On average, approximately were used in approximately half of all recorded 60 per cent of homicides involving a male victim femicides in 1999–2007 (CDC, n.d.a; see Figure were committed with a firearm; in contrast, fewer 4.13). Most of the femicides by firearm occurred ARMED VIOLENCE 2011 VIOLENCE ARMED than 40 per cent of femicides involved the use of a

of in the age group of 15–24-year-olds, in which they gun. Since 2005, there has been a marked increase account for 59 per cent of the cases. This suggests in the proportion of homicides committed using that firearm use may be linked to the age of the a firearm, with more than two-thirds of male vic- victim. Firearms were also used to kill more than tims killed by firearms in 2009. one-third of the victims aged 5–9 and almost half The extremely high levels of violence in Ciudad of those aged 10–14. Indeed, ‘children aged 5 to Juarez, capital of Chihuahua state, are reflected 14 years in the United States have 11 times the

GLOBAL BURDEN BURDEN GLOBAL in the very high proportion of homicides commit- likelihood of being killed accidentally with a gun ted with firearms, which is generally higher than compared with similarly aged children in other

Figure 4.12 Percentage of male and female homicide victims killed by firearm, Mexico and Ciudad Juarez, 1993–2009 % of male homicide victims killed with a firearm in Mexico % of female homicide victims killed with a firearm in Mexico % of male homicide victims killed with a firearm in Ciudad Juarez % of female homicide victims killed with a firearm in Ciudad Juarez

Percentage of homicides committed with a firearm

100

90

80

70

60

50

40

30

20

10

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

Source: Small Arms Survey elaboration based on INEGI (2009) across Asia and Africa—and among diasporas— among andAfrica—and Asia across countries different inmany occur practices related andother These 4.5). cide (HRCP, Box see 2011; homi- as seriousness same the with punished or procedures justice tocriminal subject be not may deaths anddowry killings ‘honour’ Specifically, level. statistical national andeven aglobal at ture tocap- difficult be may that violence non-lethal and lethal of examples considers section This 2000). 1995; 1994; (UN, in2000 andSecurity’ Peace and 1325 ‘Women on Resolution Council Security inUN culminating Declaration, Beijing and the Women against Discrimination of Forms All of tion Elimina- the on Convention the with continued it 1993); (UN, Women against Violence of Elimination the on Declaration the with started only process slow.Asteady quite been has violence from them andtoprotect women for rights equal tee toguaran- debate aninternational of development The misbehaviour. awoman’s perceived sanction to violence of use the foresees that inplace tion legisla- have still countries Some and condoned. istolerated andgirls women against violence texts, con- some In systems. reporting of accessibility poor andduetothe toreport victims of reluctance the of a undocumented as result goes violence frequently such isthat fact The women. against and violence femicide of andbreadth extent the underestimates above presented data the that doubt islittle There women against Hidden oflethal forms violence p. 358).2003, andCasteel, Moracco, (Coyne-Beasley, a firearm with behaviour reckless of aresult as often dence, insidearesi- place took frequently most firearm a by killed girls involving incidents Furthermore, p.2). 2011, (Hemenway, countries’ developed have damaged the reputation of the household, household, the of reputation the damaged have to judged behaviour of arange for punishment as malerelatives by eachyear killed cases some andin toviolence, exposed abused, are girls and women 5,000 around that estimates Fund Population ‘honour’. UN damaged family the The or customs social on infringed allegedly they because relatives by orkilled toviolence jected sub- being women of examples myriad are There 62–63). 2010b, paras. (UN, killings’ orwitchcraft violence femicide, domestic honour killings, concerning ‘whether andkillings, violence gender-based linkbetween the on research more encourages Executions or Arbitrary Summary Extrajudicial, on Rapporteur Special the report, inhis2010 gap this Acknowledging acrime. constitutes even circumstances certain suicideunder tocommit them orforcing women killing that consider not may tions jurisdic- 2008). Some Watch, (Gendercide Americas andthe inEurope countries incertain as well as 10 20 30 40 50 60 70 80 90 100 Percentage offemicidescommitted with afirearm 1999–2007 group, age by States, United Figure Source Small Arms Survey elaboration of CDC (n.d.a) CDC of elaboration Survey Arms : Small 4.13 1–4

Percentage of femicides committed with firearms in the inthe firearms with committed femicides of Percentage 5–9 1–41–42–435–44 25–34 15–24 10–14 Age group 133 2 3 4 5 1 WHEN THE VICTIM IS A WOMAN 134 clan, or tribe (UNFPA, 2003). Women who demon- Box 4.5 Dowry deaths in India strate intolerance towards traditional restrictions, The home of the in-laws in South Asia can be a particularly dangerous especially as regards the choice of friends or part- environment for women and girls. Survey data suggests that the vast majority ners and sexual behaviour, are especially at risk. of reported murders of women in India are incidents of husbands killing their wives, with 85 per cent of female victims killed in their own home (Mohanty Another scenario in which women and children et al., 2004; UNFPA, 2003). may be killed is when they are believed to be Marriages involving dowry and other demands from the husband are not practising witchcraft. According to a 2002 report only a risk factor for domestic violence, but also a direct cause of violent by Radhika Coomaraswamy, the then Special death for women. While dowry payment is illegal in India, it remains com- mon practice, with related disputes or violent demands for more money Rapporteur on Violence against Women, Its frequently leading to the death of brides or brides-to-be (Ash, 2003). Causes and Consequences, the practice is found

ARMED VIOLENCE 2011 VIOLENCE ARMED Of India’s reported 32,369 homicide victims in 2009, around one-quarter were ‘mainly in Asian and African communities’ (UN, of female (NCRB, n.d.b). ‘Dowry deaths’ under Section 304B of the Indian Penal 2002, p. 16). Code are recorded separately.33 In 2009, 8,383 cases were filed under that heading, which may include more cases of violent deaths of women and While victims can be male or female, the majority girls than those recorded under the homicide category. Taken together, of witch killings across societies appear to target dowry killings may represent up to 26 per cent of violent deaths and more women. Reports of some 50 ritual killings in Papua than 50 per cent all violent deaths of women and girls in India in 2009.34 New Guinea in 2009, for example, provide accounts The number of police-recorded dowry killings in India has risen by 25 per of young and old victims, mostly women and girls cent since 1999 (NCRB, n.d.b, p. 58). Whether the rise is due to an increase GLOBAL BURDEN BURDEN GLOBAL in police and law enforcement attention to the issue or a real underlying (AI, 2009; Parry, 2009). In Tanzania, up to 1,000 increase is unclear. The latter explanation is supported by a retrospective persons are reportedly killed every year based hospital study of female homicide victims, which estimates that 30 per on allegations of witchcraft; the majority of the cent of those deaths are dowry-related (Mohanty et al., 2004, p. 153). victims are women above the age of 50 (HAI, 2008, Source: Malby (2011) p. 7). According to a study carried out in Ghana, being aware of the status of women and girls in a society is crucial to understanding witch-related femicides (Adinkrah, 2004). The Special Rappor- teur also points out that witch killing is often the result of highly unequal gender relations in a society (UN, 2002, p. 17).

Research suggests that witch murders are often linked to poverty and situations of economic despair. One study carried out in Tanzania assesses the number of witch murders in connection with extreme rainfalls that lead to droughts or floods. It finds that the link is significant: ‘There are twice

Photo W A newlywed stands in a shelter for victims of dowry violence next to a poster highlighting the economic demands some new brides face from their husbands and in-laws, New Delhi, June 2004. © Elizabeth Dalziel/AP Photo stoning, orbeating. stoning, includingburning, forms, ritual take may Killings ways. of inavariety killed are witches Alleged suffering: economic of result the be well may they that gest sug- inIndia killings witch of reports Similarly, p.1153). 2005, (Miguel, years’ inother as fall rain- extreme of inyears murders witch many as grammes and interventions. andinterventions. grammes pro- prevention of effectiveness the and limits HIV/AIDS of spread tothe contributes assistance andmedical information of Alack them. against violence further use may who partner, the of cion suspi- raise would care medical for asking which in relationships inabusive live Many incidents. orreport help seek rarely violence of Victims problem. the of extent the toquantify difficult 2010). isextremely It Moreno, andGarcía (Dunkle because of the likelihood of tissue laceration transmission virus of chances the increases sex forced Organization, Health World tothe According virus. the contracting of risk high at are violence sexual of victims andwomen transmission HIV of majormode isthe sex HIV/AIDS. Unprotected with infection isthrough killed unintentionally and intentionally are women that way Another are abducted and killed in order to sell their organs. their tosell inorder andkilled abducted are victims suchscenarios, 2009). In (Kelly, in organs trafficking for covers as acting rituals witchcraft of risk tothe attention called have reports eral the problems (UN, 2002, p.17). 2002, (UN, problems the iscausing that spirit evil the of rid be to death, through elimination their suggests and ‘dayans’ or witches be to women or awoman declares usually [...]He isapproached. doctor magic local the calamity, isanatural there if even or failure, isacrop there if or die, cattle or family, the in isadeath there if or water, drinking of a lack is there if or illnesses, from suffer people When 35 In the past few years, sev- years, few past the In 36

enced physical violence since they turned 15 turned they since violence physical enced experi- have they say who women of proportion 2008–09 The Moreno, 2010). García and (Dunkle status HIV their to due violence experience to likely more are HIV with women living In addition, decision-making. sexual healthy assert or infection HIV from selves them- protect to adequately women’s ability hamper can violence of threat the and Violence 2009). (Machera, partners their by abuse of forms many to subjected are women HIV-positive that reveals Kenya in out carried study A recent infection. HIV of source isamajor partner, male the by initiated usually violence, sexual spousal that reality: along-hidden confirm ings find- These 2009). NACC, (Kenya partnerships regular and unions existing within sex erosexual het- to attributed be can incidence HIV national of cent per 44.1 that finds study Arecent ences. such andviolence experi- silent to about remain domestic rationalize even and tolerate, accept, to socialized are women elsewhere, as Kenya, In high. remained women for risk the subsided, violence election post- the after Even (COVAW, 2008). rape gang of incidence elevated the to due people placed dis- 350,000 approximately the among serious extremely was situation the Specifically, others. than transmission HIV/AIDS and violence sexual to vulnerable more were women of groups Some 2008). (Holmes, assistance ing seek- victims rape of number the in surge a rapid with confronted soon were Hospitals rape. mass of in suddenly, waves ushering erupted Violence 2008. early and 2007 December of violence post-election the during HIV/AIDS contracting of risk high especially at were Kenya in girls and women that is evidence There pandemic. HIV/AIDS the of vector akey also but problem, health apublic only isnot women against Violence B ox 4.6 Kenya Demographic and Health Survey Survey Health and Demographic Kenya HIV and sexual violence inKenya violence andsexual HIV documents a sizeable reduction in the the in reduction asizeable documents XX

135 2 3 4 5 1 WHEN THE VICTIM IS A WOMAN 136 years old (KNBS, 2010). The survey finds a reduction from 49 per cent in 2003 to 39 per cent in 2008–09. Even so, at least 12 per cent of women aged 15–49 reported that their first sexual intercourse was forced against their will. Moreover, overall, one in five Kenyan women (21 per cent) expe- rienced sexual violence. Analysis across provinces indicates that the two provinces with the highest proportion of women experiencing physical violence (Nyanza and Western provinces) also featured the highest pro- portion of women experiencing sexual violence.

Among surveyed women who had been married at least once, sexual vio- lence was reportedly perpetrated mainly by current and former husbands and partners. Those who never married said that violence was committed ARMED VIOLENCE 2011 VIOLENCE ARMED

of mainly by boyfriends, although almost one in five never-married women (19 per cent) was violated by a friend or acquaintance and almost as many by a stranger (17 per cent). The likelihood of experiencing either physical or sexual violence increases with the age of the women. That said, women with secondary or higher education and those in the top two wealth quintiles are less likely to experience sexual violence than other women.

GLOBAL BURDEN BURDEN GLOBAL In the vast majority of cases, sexual violence is perpetrated by persons known to the victims (Machera, 2009). Indeed, strangers commit only 6 per cent of recorded sexual violence. About 37 per cent of women who experienced sexual violence reported current husbands or partners as the perpetrators, followed by current or former boyfriends (16 per cent) and former husbands or partners (13 per cent). Women who have experienced both physical and sexual violence are more likely to seek help than those who experienced only one or the other (KNBS, 2010). Older women are more likely than younger women to seek help to stop the violence.

To reduce sexual violence the Kenyan government has enacted the Sexual Offences Act No. 3 of 2007 (KNBS, 2010). The law has been lauded as a move in the right direction, although its implementation remains slow. For example, the cases brought by women and girls who were raped during the post-election violence in 2007–08 have not yet been addressed by the courts. Mechanisms for retrieving evidence and tracking down perpetrators are generally substandard. Consequently, the law has not yet succeeded in deterring rapists.

Source: Machera (2011)

Photo X A woman stands outside an office dealing with child abuse and violence against women at a police station in Mombasa, Kenya, 2005. © Sven Torfinn/Panos Pictures 137 2 3 4 5 1 WHEN THE VICTIM IS A WOMAN 138 The level of brutality has been heightened through of women and pervasive gender inequality. In the systematic use of rape as a weapon of war these places, women and girls cannot enjoy a safe and perpetrated by soldiers who carry the HIV/ or secure lifestyle. The chapter calls attention to AIDS virus. Indeed, the widespread and system- incidents occurring inside as well as outside the atic targeting of civilians and the use of rape is a domestic sphere. Women are vulnerable to vio- striking aspect of recent armed conflicts in Sub- lence committed by strangers, but more frequently Saharan Africa (Elbe, 2002, p. 168; see Box 4.6). they are unsafe in their own homes. In other parts of the world, rape has also been A key conclusion, then, is that there is a need to used as a systematic weapon of terror leading to enhance reporting and analysis of data on lethal the spread of HIV/AIDS. During the military rule and non-lethal violence against women—both cross- between 1991 and 1994 in Haiti, for example, nationally and sub-nationally. This goal could be

ARMED VIOLENCE 2011 VIOLENCE ARMED women were raped because of the alleged politi- achieved by undertaking steps such as those pro- of cal activities of their husbands. The perpetrators posed by the Geneva Declaration study on Tackling were reportedly ‘police, soldiers and criminal Violence against Women: From Knowledge to Prac- gangs operating with impunity’ (Bastick, Grimm, tical Initiatives (Milliken, Gilgen, and Lazarevic, and Kunz, 2001, p. 79). Twenty per cent of police 2011). Practical recommendations include: officers reportedly suffered from HIV/AIDS in Haiti at the time. As a consequence, in 2001 Haiti had supporting international initiatives to track the highest prevalence of HIV/AIDS in the Western violence against women globally; GLOBAL BURDEN BURDEN GLOBAL hemisphere (p. 79). the promotion of field-based research on mapping violence against women, including surveys to measure violence committed by Conclusion intimate partners and strangers; and The critical role of collecting data on lethal vio- the development of costing tools to improve lence against women cannot be overstated. And estimates of the effects of violence against while the evidence base is growing to demonstrate women on development, including direct and the scale and distribution of femicide and violence indirect costs of violence against women. against women, this chapter has underlined criti- cal information gaps, especially across Africa and Asia. The fact is that the quantity and quality of Abbreviations data on femicide are very poor and characterized INEGI Instituto Nacional de Estadística y Geografía (Mexico) by incomplete geographical coverage. Reliable IPV Intimate partner violence and valid information on violence according to sex, age, relationship to the perpetrator, and instrument used is crucial to designing effective violence prevention and reduction strategies. Endnotes In settings where reported levels of violence are 1 Figures based on a report generated at CDC (n.d.b). high, femicide levels are also likely to be high. 2 The findings are based on a 2007 survey that estimates that between 1.7 and 1.8 million women were raped in Similarly, these environments are likely to be their lifetime in the Democratic Republic of the Congo characterized by the systematic discrimination (Peterman, Palermo, and Bredenkamp, 2011). 3 See Small Arms (2006)foranassessmentSurvey of injury 13 For example, the World Bank recently provided funding 12 For methods of calculation, seetheonline methodological 11 Femicide rates represent anaverage over theperiod 10 Thefact that this dataset contains asmaller numberof 9 As inChapter Two, several smaller Caribbean islands have 8 For 7 See,forexample, thecompilation of data at UNECE (n.d.) deta See,forexample, UNECE (n.d.). 6 5 Russell alsonotes that certain dictionaries define‘femicide’ 4 Thedatapresent 3 collective violence. due to‘external’ cause light thescale anddistribution of mortality andmorbidity preventionontheAfricaninjury continent and tohigh- of individuals andinstitutions for Africa. The Initiativ surveillance systems fr prevention surveillance systems inAfrica. see Zavala andHazen (2009) about theroleof injury for therestructuring of theNigeri annexe at www.genevadeclaration.org. gaps indata series. and lowpoints as well as thedistortions resultingfrom 2004–09. This ‘smoothing’ of data reduces extremehigh the information,higher collection methods indifferent co due totheuseofFurthermore, varying definitions ordata is extremely scarce at theint the capacity orresources tocompile such information, which on thecircumstances of homicide. Many countries may lack the difficulty inherent incapturing additional information countries than does theonef the Maldives,Monaco, andQatar. from theanal countries that nofemale reported victims areexcluded and theGrenadines, and Trinidad and Tobago. The five nica, Grenada, St. Kitts andNevis, St. Lucia, St. Vincent they areAnguilla, Antiguaand been groupedtogether as the‘Lesser Antilles Region’; declaration.org. see theonline methodological annexe at www.geneva and that presentedin etSanmartín al.(2010). as ‘thekilling of awoman’ (Russell, 2008,p. 3). Union (UNECE, n.d.). statistics, covering 29 countrie Economic Comm 44 countries et (Sanmartín al.,2010);c)theUnitedNations carried out by theQueen Sofia Center in Spain, covering age andsex (Bhalla et al.,2011);b)astudy onfemicide public health data on violent deaths disaggregated by University,Harvard coverin database developedby theHomicide Advisory Groupat obtained fromthefollowing international sources: a)the ils onpopulationdataand ysis; thecou ission forEu ed inthischapter hasla e wasfou s, includinginterpersonal and om the Injury PreventionInitiative om theInjury ntries areAndorra, g 96countries andproviding ernational comparative level. or general femicide reflects rope databa s, mostly fromtheEuropean involved in violence and Barbuda, Barbados, Domi- nded in1997asa untries, the more detailed untries, themore detailed risk of incomparability. a Bureau of Statistics; regional c rgely been rgely been se ongender lassific Liechtenstein, network ations, ations, 1 Anotherchallenge f 21 Inordertocompare trend 20 See,forexample, Suar 19 Seethe2009 18 Adetailed breakdown of femicide rates at thestate and 17 Seetheonline methodological annexe at www.geneva 16 15 The anal 4 Regional homicide rates presentedheredonot correspond 14 4 Thesecountries exhibit someof thelowest homicide 24 Thepopulation sex ratio varies significantly across coun- 23 Whilethereis of adearth homicide data onwomenin 22 each country over theperiod2004–09. in the year 2004. Lines showperc corresponding tothenumberof femicides inthecountry terns over fromanormalized timestarting value of 100, interference of thedifferent levels,thechart shows pat- pp. 21–23). in thecountry andespecially inCiudad Juarez (UN,2011, Mexico respondtocontinued violence against women Causes andConsequences, Rashida Manjoo, that Special on Rapporteur Violence against Women, Its (SRE, 2009, p. 1);seealsother disappearances andkillings of womeninCiudad Juarez Human Rights holdingMexico responsible forunsolved trend data is frequently availab which limitsthepotential base includes fewcountriesvery with complete timeseries, represented by timeseries. The (2010). Ministry of Justice in Waiselfisz (2011). municipal levels inBrazil is presentedby theBrazilian declaration.org. Southern Africa. AND PATTERNS). calculated based onalarger numberof countries (TRENDS ter Two, which alsoincludedir to rates of violent deaths perregioninFigure2.3 inChap- small numberof cases. rates, inwhich there may belarge fluctuations duetothe AND PATTERNS). Switzerland (0.83), Norway (0.69), andJapan (0.45) (TRENDS rates intheworld: Austria (0.68 per100,000 population), (92) (UNdata, n.d.) (86 menper100women),Belarus (87), andCape Verde with alarger female population areLatvia andEstonia men per100women,respectively. Incontrast, countries of male population, with ratios of 146, 135, 121, and106 Arabia, andBrunei Darussalam have astrongprevalence female population. For example, Kuwait, Bahrain, Saudi larger differences, showing anexcess of eithermale or women for2010(UNdata, n.d.). Some countries have tries. The global ratio is estimated at 101.7 menper100 also limited. many countries, relevant information onmale victims is ysis inc decision oftheInt ludes datafromonlyin onecountry or statistical ez andJordan(2007) andRNOCDH oftrendsan s across countries without the ect conflict deaths andare le forcountries with low eiterated requests of the analysis GBAV 2011femicide data- er-American Court of entages of change for alysis. Fu of femicideis rthermore, rthermore, 139 2 3 4 5 1 WHEN THE VICTIM IS A WOMAN 140 25 See, for example, Suarez and Jordan (2007) on the involve- ment of organized crime in femicide in Guatemala. Bibliography Abrahams, Naeemah, Rachel Jewkes, and Shanaaz Mathews. 26 Statistically, female authors of femicide, either in same- 2010. ‘Guns and Gender-based Violence in South Africa.’ sex couples or in other circumstances, represent a very small portion of the total, for example less than 1 per cent South African Medical Journal. September, pp. 586–88. in the United States. See Glass et al. (2004). 27 Sexual abuse is a strong predictor of suicide; see, for example, McFarlane and Malecha (2005). ABS (Australian Bureau of Statistics). 2009. ‘4510.0—Recorded Crime: Victims, Australia, 2009.’ 28 The definition of ‘forced suicide’ is commonly used and has been adopted in the context of violence against women its Causes and Consequences. See UN (2008). Adinkrah, Mensah. 2004. ‘Witchcraft Accusations and Female 29 Any form of violence against women is likely to be reiter- Homicide Victimization in Contemporary Ghana.’ Violence ated and to have a long-lasting effect on the victim. This against Women, Vol. 10, No. 4, pp. 325–56.

ARMED VIOLENCE 2011 VIOLENCE ARMED generates psychological effects ranging from low self- ncjrs.gov/App/publications/Abstract.aspx?id=205819>

of esteem to the development of aggressive or criminal AI (Amnesty International). 2009. ‘Papua New Guinea: Authori- behaviour and, very frequently, self-inflicted violence. ties Must Act Now to Prevent More Witch Hunts.’ When the violence comes from the surrounding community See, for example, Ho (2008), Dubow, Huesmann, and Boxer (2009), and Leslie (2000) on the long-term psychological AIHRC (Afghanistan Independent Human Rights Commission). effects of violence against women in conflict settings. 2004. Annual Report.

GLOBAL BURDEN BURDEN GLOBAL 31 The agencies are the Ministries of Women’s Affairs, Interior, Alvazzi del Frate, Anna. 2010. ‘Crime and Criminal Justice Statis- Health, Education, and Justice, as well as the courts, tics Challenges.’ In Stefan Harrendorf, Markku Heiskanen, women’s shuras (councils of elders), provincial councils, and Steven Malby, eds. International Statistics on Crime the Afghanistan Independent Human Rights Commission, and Justice. Helsinki: European Institute for Crime Preven- referral centres, Save the Children, legal aid providers, tion and Control (HEUNI) and United Nations Office on and the UN Assistance Mission in Afghanistan. Drugs and Crime, pp. 167–75. 32 See Catalano et al. (2009). Victimization rates are per Arroyo, Mario. 2011. Femicides in Mexico and Ciudad Juarez. 100,000 persons age 12 or older. The difference between Unpublished background paper. Geneva: Small Arms Survey. male and female intimate partner victimization rates is Ash, Lucy. 2003. ‘India’s Dowry Deaths.’ BBC News. 16 July. significant at the 95 per cent confidence level. ‘where the death of a woman is caused by any burns or bodily injury or occurs otherwise than under normal cir- Auchter, Bernie. 2010. ‘Men Who Murder Their Families: What the cumstances within seven years of her marriage and it is Research Tells Us.’ NIJ Journal, No. 266. June, pp. 10–12. shown that soon before her death she was subjected to cruelty or harassment by her husband or any relative of Azarbaijani-Moghaddam, Sippi. 2007. ‘On Living with Negative her husband for, or in connection with, any demand for Peace and a Half-Built State: Gender and Human Rights.’ dowry, such death shall be called “dowry death” and such International Peacekeeping, Vol. 14, No. 1, pp. 127–42. husband or relative shall be deemed to have caused her death’ (India, 1860). Bastick, Megan, Karin Grimm, and Rahel Kunz. 2001. Sexual Violence in Armed Conflict: Global Overview and Implica- 34 NCRB (n.d.a; n.d.b). tions for the Security Sector. Geneva: Geneva Centre for 35 See Watts, Osam, and Win (2001, p. 91) about the stoning the Democratic Control of Armed Forces. of accused witches in Zimbabwe. See Schnoebelen (2009) for details on the beating and burning of so-called witches Bhalla, Kavi, et al. 2011. Country-level Estimates of Homicides in Angola. from Death Registration Data. Unpublished background paper. Geneva: Small Arms Survey. 36 In the last few years, the connection between trafficking in women and girls and trafficking in human organs has BJS (Bureau of Justice Statistics). 2005. ‘Homicide Trends in the moved up on the international agenda. See Pearson (2004) U.S.: Intimate Homicide.’ Geneva: World Health Organization andUNAIDS. Violence against Women andHIV/AIDS: What Works? and Family Psychology Review on Children’s Psychosocial Adjustment.’ Impact of ExposuretoPersi Cognitive-Ecological Framew April, pp. 355–60. Archives ofPediatrics &Adolescent Medicine, Vol.157. 2003. ‘Adolescent NGO/Disp-VAW_RapidAssess_COVAW_2008.pdf> Kisii, January–February 2008. IDPs Camps inWestern Kenya—Kisumu, Kakamega and Rapid Assessment of the Situation ofWomen andGirls in Violent Death Reporting System (NVDRS).’ ment of Justice. Victims ofViolence Bureau ofJustice Statistics—Selected Findings:Female org/cgi/content/abstract/93/7/1089> No. 7. July, pp. 1089–97. Chapel Hill, NC:MEASURE Evaluation. A Compendium ofMonitoring

American Journal ofPublic Health ding Causes ofDe Violence against Women andGirls: Control andPrevention).n.d.a. ‘National . Washington, DC:United States Depart- Femicide: APopulation-Based Study.’ a GarcíaMoreno.2010. gov/content/pub/pdf/fvv.pdf> gov/content/pub/pdf/cv08.pdf> Webster,and org/cgi/reprint/157/4/355.pdf> stent Ethnic–Political Violence ork forUnderstanding the

and Evaluation Indicators , Vol. 12,No.2,pp. 113–26. assets/file/Resources/ ath Reports, ath Reports, ehealth/publications/ rs/nvdrsDisplay.jsp> Nancy G Vol. 24,No.4, Clinical Child Addressing lass. 1999–2007.’ Kenya: , Vol.93, . Elbe, Stefan. 2002. ‘HIV/AIDS andtheChanging Landscape of War ECOSOC (Economic and Social Council). 2004. Resolution 2004/22 FBI (Federal Bureau of Investigation). 2010. ‘ExpandedHomicide Faqir, Fadia. 2001.‘Intrafamily Femicide inDefence of Honour: Eriksson Baaz, Maria and Maria Stern. 2010. Human Organs. Africa Institute. American Journal ofLifestyle Medicine BBC. 14June. ngos/HAITanzania41.pdf> armssurvey.org/fileadmin/docs/A-Yearbook/2006/en/ Oxford University Press, pp. 189–213. e_against_migr ov/ucr/cius200 Third World Quarterly ating andPunishing Trafficking in ide.org/case_honour.html> ant_women_35725562 a-portal.org/smash/record. . Copenhagen: Euro- , Vol. 27, No.2,pp. 159–77. GlobalBurden ofArmed orld-africa-13760895> Matters:Assessing the 9/offenses/expanded_ . SIDAWorking Paper Violence against . The Complexity of , Vol. 22,No.1, 0.pdf> 141 2 3 4 5 1 WHEN THE VICTIM IS A WOMAN 142 — and Deborah Azrael. 2000. ‘The Relative Frequency of Offensive Liem, Marieke and Paul Nieuwbeerta. 2010. ‘Homicide Followed and Defensive Gun Use: Results of a National Survey.’ by Suicide: A Comparison with Homicide and Suicide.’ Violence and Victims, Vol. 15, pp. 257–72. Suicide and Life-Threatening Behavior, Vol. 40, No. 2. Ho, Joyce. 2008. ‘Community Violence Exposure of Southeast Violence, Vol. 23, No. 1, pp. 136–46. Machera, Mumbi. 2009. Linkages between Violence against Women and HIV/AIDS among WOFAK Beneficiaries. Nairobi: Holmes, Stephanie. 2008. ‘Gang Rape Spirals in Violent Kenya.’ Women Fighting AIDS in Kenya and ActionAid International. BBC News. 23 January. Study_by_AAK_2009.pdf> HRCP (Human Rights Commission of Pakistan). 2011. State of —. 2011. ‘Gender-based Violence in Kenya.’ Unpublished back- Human Rights in 2010. Lahore: HRCP. April. ground paper. Geneva: Small Arms Survey. Malby, Steven. 2011. ‘Dowry Deaths in India.’ Unpublished back- Hume, Mo. 2008. ‘The Myths of Violence: Gender, Conflict, and ground paper. Geneva: Small Arms Survey. ARMED VIOLENCE 2011 VIOLENCE ARMED Community in El Salvador.’ Latin American Perspectives, of Vol. 35, No. 5, pp. 59–76. Mathews, Shanaaz, et al. 2008. ‘Intimate Femicide–suicide in South Africa: A Cross-sectional Study.’ Bulletin of the World Independent (UK). 2008. ‘Women Who Took on the Taliban—and Health Organization, Vol. 86, pp. 552–58. McFarlane, Judith and Ann Malecha. 2005. ‘Sexual Assault among Intimates: Frequency, Consequences & Treatments.’ India. 1860. ‘Section 304 B in the Indian Penal Code, 1860.’ Washington, DC: United States Department of Justice. INEGI (Instituto Nacional de Estadística y Geografía). 2009.

GLOBAL BURDEN BURDEN GLOBAL McFarlane, Judith, et al. 1999. ‘Stalking and Intimate Partner Exchange of Data on Homicides in Mexico. Unpublished Femicide.’ Homicide Studies, Vol. 3, No. 4. November, background paper. Geneva: Small Arms Survey. pp. 300–16. Kelly, Annie. 2009. ‘Child Sacrifice and Ritual Murders Rise in ed. The Psychology of Stalking: Clinical and Forensic Kenya NACC (National AIDS Control Council). 2009. Kenya: HIV Perspectives. San Diego: Academic Press, pp. 2–27. Prevention Response and Modes of Transmission Analysis. onepage&q&f=false> Kivivuoria, Janne and Martti Lehtia. 2003. ‘Homicide Followed Miguel, Edward. 2005. ‘Poverty and Witch Killing.’ Review of by Suicide in Finland: Trend and Social Locus.’ Journal of Economic Studies, Vol. 72, pp. 1153–72. Vol. 4, No. 2, pp. 223–36. Milliken, Jennifer, Elisabeth Gilgen, and Jasna Lazarevic. 2011. KNBS (Kenya National Bureau of Statistics). 2010. Kenya Demo- Tackling Violence against Women: From Knowledge to graphic and Health Survey 2008–09. Calverton, MD: KNBS Practical Initiatives. Geneva Declaration Working Paper. and ICF Macro. Geneva: Geneva Declaration Secretariat. KWRW (Kurdish Women’s Rights Watch). 2008. ‘Reports of Honour- based Murders and Suspected Murders.’ April. Mohanty, Manoj Kumar, et al. 2004. ‘Victimologic Study of Female Homicide.’ Legal Medicine, Vol. 6, No. 3, pp. 151–56. Leslie, Helen. 2000. ‘Conceptualising and Addressing the Mental Molloy, Molly. 2010. ‘Juárez Murders: Impunity Regardless of Health Impacts of Gender Roles in Conflict and Peace- Gender.’ Grassroots Press. 12 May. making.’ Development Bulletin, No. 53, pp. 65–69. Afghanistan: Baseline Statistics on Gender. Kabul: MOWA RNOCDH (Red Nacional deOrganismos Civiles de Derechos RCMP (Royal Canadian Mounted Police). 2010. Radford, Jill Psytel. 2010. Estimation delamortalitéliéeaux violences conju- Peterman, Amber, Tia Palermo Pearson, Elaine. 2004. Pearce, Diana. 1978. ‘TheFeminization of Poverty: Women, Work Parry, Richard Lloyd.2009. ‘“Witches” Put toDeath inPapua Oberwittler, Dietrich. 2008. —. n.d.b. ‘Table 3.3: Age Group Wise Victims of Murder(Section NCRB (National Criminal Records Bureau of India).n.d.a. ‘Table Pursuant to Article40of the I Consideration as Part ofits Fifth Periodical Report onMexico ‘All Rights for All’) to the UNHumanRights Committee for (National Network ofHumanRights Civil Organizations Humanos ‘Todos los Derechos para Todas y Todos’ the Red NacionaldeOrganismos Civiles deDerechos Information Presented by Organizations FormingPart of Humanos ‘Todos los Derechos para Todas y Todos’). 2010. 2007/DAP-1/140. scientifique. Rapport gales enEurope: ‘IPVEU Mortality Journal ofPublic Health Women intheDemocratic Republic of Congo.’ ‘Estimates andDeterminants of Sexual Violence against 2004.pdf> 13 February. 302 IPC)during2009.’ Amounting toMurder(C.H.)during2009.’ Afghanistan. Femicide: ThePolitics The Times(UK). . February. RCMP Canadian American Vol.11, Eschborn: UN (UnitedNations). 1993. Jordan.2007.Suarez, JulieandMarty Stark, Evan andAnneFlitcraft. 1996. SRE (Secretaría deRelaciones Small Arms Survey. 2006. Schnoebelen, Jill. 2009. José,etSanmartín, al.,eds.2010. Sánchez, Jessica. 2011.InformeFinaldeFemicidios enHonduras Russell, Diana. 2008. ‘Femicide: Politiciz Ruggi, Suzanne. 1998. ‘Commodifying HonorinFemale Sexuality: Roberts, Kerry, et al.2010. ‘Homicide–Suicide inDurban, South yearbook/small-arms-survey-2006.html> RighttoLive/ThreethousandandCounting,AReportonViolence USA. General deDerechos Humanos y Democracia. informativo No.160. 24December. Mexico City: Dirección Humanos: Agenda Internacional bodies/hrc/docs/ngos/RNOCDH_mexico98.pdf> Civil andPolitical Rights. for Refugees. Vol. 16. Washington, DC.May. Meeting on Strengthening Paper presentedat theInter pp. 12–15. Honor Killings inPalestine.’ pp. 877–99. Africa.’ Journal ofInterpersonalViolence, Witchcraft Allegations, RefugeeProtection Small Arms Survey 2006: Unfinished Declaration on the Elimination of emala Hum Exteriores). 2009. ‘Derechos e against Women inGuatemala. . Thousand Oaks,London, CA, sh/law/eliminationvaw.htm> Three Thousand andCount- nasofia.es/publicaciones. Women at Risk:Domestic Nations ing theKilling of Females.’ deMexico.’ Boletín an RightsC High Comm issioner . Tegucigalpa: Vol. 25, No.5, . Research ommission/ , No.206, 143 2 3 4 5 1 WHEN THE VICTIM IS A WOMAN 144 —. 1994. Convention on the Elimination of All Forms of Discrimi- UNDP (United Nations Development Programme). 2008. ‘Sexual nation against Women (CEDAW). A/RES/49/164. Violence against Women and Children in Armed Conflict.’ Background document presented at the 2008 Parliamentary —. 1995. Report of the Fourth World Conference on Women— Hearing at the United Nations. New York, 20–21 November. Beijing, 4–15 September 1995. A/CONF.177/20. —. 2000. Security Council Resolution 1325 on ‘Women and peace UNECE (United Nations Economic Commission for Europe). n.d. and security’. S/RES/1325 of 31 October 2000. ‘UNECE Statistical Database: Gender Statistics–Crime and Violence.’ Violence against Women, Its Causes and Consequences, Ms. Radhika Coomaraswamy, on ‘Cultural Practices in the UNESCO (United Nations Education, Scientific and Cultural Family that are Violent towards Women.’ E/CN.4/2002/83 Organization). 2011. The Hidden Crisis: Armed Conflict ARMED VIOLENCE 2011 VIOLENCE ARMED and Education. Paris: UNESCO.

of of 31 January 2002. Geneva: UN. Women in South Asia: A Regional Analysis. Kathmandu: —. 2008. Promotion and Protection of All Human Rights, Civil, UNFPA. Right to Development: Report of the Special Rapporteur on UNIFEM (United Nations Development Fund for Women). 2006. Violence against Women, Its Causes and Consequences, Uncounted and Discounted: A Secondary Data Research Yakin Ertürk—Indicators on Violence against Women and Project on Violence against Women in Afghanistan. State Response. A/HRC/7/6 of 29 January 2008. Geneva: UN. Kabul: UNIFEM. GLOBAL BURDEN BURDEN GLOBAL ESA/STAT/SER.K/19. New York: Department of Economic Waiselfisz, Julio Jacobo. 2011. ‘Caderno Complementar 2: and Social Affairs, UN. Homicídios de Mulheres no Brazil.’ In Mapa da Violência and Instituto Sangari. Summary or Arbitrary Executions, Philip Alston. A/HRC/14/24. Watts, Charlotte, Susanna Osam, and Everjoice Win. 2001. Geneva: UN. Roberta Harmes, eds., pp. 89 –99. —. 2011. Report of the Special Rapporteur on Violence against WHO (World Health Organization). 2002. World Report on Violence Women, Its Causes and Consequences, Rashida Manjoo— and Health. Geneva: World Health Organization. Addendum: Communications to and from Governments. Zavala, Diego and Jennifer Hazen. 2009. Understanding Violence: A/HRC/17/26/Add.1 of 18 May 2011. Geneva: UN. The Role of Injury Surveillance Systems in Africa. Geneva: UNAMA (United Nations Assistance Mission in Afghanistan). 2009. Silence Is Violence: End the Abuse of Women in Afghanistan. Kabul and Geneva: UNAMA and Office of the United Nations High Commissioner for Human Rights. UNdata. n.d. ‘Population Sex Ratio (Males per 100 Females).’ New York: United Nations Statistics Division.