takes two further steps, both of which have have which of both steps, further two takes edition This (homicide). violence interpersonal and conflict from death violent of levels regional andsub- andglobal violence, armed of burden global overall the estimating for framework basic a out sets in2008, published edition, first The consequences. andits causes, its violence, armed Armed Violence of Burden Global the of edition second This orcompare. tomeasure difficult are that ways in andsocieties communities of fabric economic and political, social, the andcorrode insecurity, and fear orinjury—spread indeath result not do they when violence—even armed of levels High humanbeings. fellow of orharming killing ate deliber- the involves it inthat isdistinct violence yet people’s lives, andaccidents—claim ters, disas- natural anddiseases, including epidemics tragedies— other Many violence. and sexual famine, tosickness, exposed and communities, homes their toleave forced been have others less count- while lives, their orlost injuries suffered have inMexico—millions war drug tothe Lagos and Angeles inLos violence street and from d’Ivoire, orCôte inLibya conflicts contemporary to Africa inSub-Saharan colonialstruggles the from andAfghanistan, inIraq interventions the A of settings—from the 20 the settings—from of avariety In world. the around people of millions directly affected the lives of hundreds of of hundreds of lives the affected directly manifestations—has and contemporary rmed violence Chapter A Unified Approach to Violence Armed report takes a unified view of of view aunified takes report —in both its historical historical its both —in th

-century world wars to wars world -century One armed groups, and persistently high levels of of levels high and persistently groups, armed andnon-state gangs, crime, organized national trans- involving networks regional of growth the wars, of dimensions Economic violence. motivated and economically motivated politically andbetween andcrime, conflict armed between line the blurred has violence armed porary contem- of nature changing the that noted have analysts Numerous contexts. of range in awide andappears forms many takes violence Armed violence. of armed and distribution scale inthe variations national and regional of comparisons fine-grained more allows also It estimate. nuanced amore provide andto 2004–07) than rather (2004–09 period time alonger at tolltolook death global the of 2011 Violence Armed of Burden the isthat analysis of level andthe ments, instru- data, of refinements these of result One with each other. interact may violence of forms different and how contexts, indifferent ismanifest violence how of understanding our anddeepening refining further for basis the provides thus It violence. armed of forms andinterpersonal criminal, conflict, between todistinguishing opposed as allsources, from death violent of overview aggregate first the topresent sources multiple from data available andanalyses synthesizes it Second, victimization. violent for estimates level national- comparable in’topresent ‘zooms it First, implications. andprogramming policy important revisits estimates estimates revisits Global Global 11 2 3 4 5 1 A UNIFIED APPROACH 12 interpersonal violence—whether in conflict and sources of violence and insecurity before post-conflict situations or in settings that have focusing on specific drivers or manifestations not experienced war—make clear that armed of violence. violence is a complex phenomenon to untangle. Drawing sharp boundaries around the organiza- The following chapters explore different elements tion, nature, and purpose of different violent of these general arguments. Chapter Two unpacks acts is unhelpful in developing responses to the national- and regional-level data on rates and diverse manifestations of violence around the levels of armed violence around the world, focusing world. Following the usage in the first Global on the 58 most violence-affected states, all with Burden of Armed Violence report, this volume an overall violent death rate exceeding 10 per defines armed violence generally as ‘the inten- 100,000. It shows that violence in non-conflict ARMED VIOLENCE 2011 VIOLENCE ARMED tional use of illegitimate force (actual or threat- settings (intentional homicide) is responsible for of ened) with arms or explosives, against a person, the vast majority of killings (slightly more than group, community, or state, that undermines 75 per cent of all deaths), while conflict-related people-centred security and/or sustainable violence accounts for only 10 per cent of all vio- development’ (Geneva Declaration Secretariat, lent deaths. The remaining deaths are attributed 2008, p. 2).1 to killings during legal interventions (4 per cent) and unintentional homicide (just over 10 per cent).2 This chapter presents an overview of the main

GLOBAL BURDEN BURDEN GLOBAL Chapter Three focuses on violent victimization and themes of the Global Burden of Armed Violence the instruments of violence involved in so-called 2011, focusing in particular on the reasons for— non-conflict settings. It finds that gang- and organ- and challenges to—adopting a unified approach ized crime-related homicides are highly concen- to contemporary armed violence. Its key conclu- trated in Central and South America, and that sions are that: deaths related to robbery are higher in countries with pronounced inequalities. It also points towards The intensity and location of conflict and non- the critical nexus that may exist between high conflict armed violence has changed signifi- homicide rates, a high proportion of homicides cantly over recent decades. committed with firearms, and a low proportion of Conventional typologies of armed violence cases solved by law enforcement. Chapter Four based on the context, intention, and type of examines global patterns of violence against actor have limitations for both research (data women. The final chapter studies the links between collection) and policy-making (prevention and armed violence and development by considering reduction programmes). the impact of armed violence on progress towards The boundaries between political, criminal, achievement of the Millennium Development and intimate or gender-based violence have Goals and other development indicators. It finds become increasingly blurred, as revealed in that countries with high and very high levels the cases of Iraq, Mexico, and Somalia. of violence are concentrated in the low human Effective violence prevention and reduction development and low-income categories, and that programmes and policies need to start with a there is a persistent link between poverty, armed unified assessment of the scope, scale, and violence, and development. gender-based violence. and and interpersonal violence, motivated cally criminal and economi- organized violence, terrorist and state, tointer-communal, war with ciated asso- violence large-scale the from range These actor. violent the of orintentions organization of level tothe according violence and criminal conflict of forms different todelineate attempt that andcategorizations typologies of range awide lies distinctions general these Beyond disputes. interpersonal or example) for robberies, during (murders gain economic impersonal by ismotivated it whether of 2008), regardless andWelch, (Riedel law by sanctioned isnot that violence armed of use considerations. topolitical subordinate are that ends particular toachieve way inacalibrated force and uses organized, ishighly means, other by struggles political of continuation is the forms) various its (in ‘war’ towhich according logic, Clausewitzian a follows violence Conflict acts. violent the behind, motivations of, and the organization of level the capture distinctions These violence. and motivated) andbetween violence, vidual) organized between drawn are distinctions common most two The perpetrator. the of intentions underlying orthe context the as tosuchfactors according categorized be can that types distinct of series a as treated isconventionally violence Armed Why perspective? aunified Figure 1.1 The ‘macro’ distinction is between is between distinction ‘macro’ The 1.1 Figure in presented violence armed of picture general tothe rise gives classification of sort This orrape. assault, kidnapping, extortion, violence, gang terrorism, such as acts violent specific into disaggregated be can (collective) and (collective) 3 criminal Criminal violence is simply the the issimply violence Criminal (economically motivated) motivated) (economically 4 Each of these categories categories these of Each interpersonal conflict (politically (indi- 1,000 victims 1,000 than more the for case the was as groups, nized orga- loosely by inmassacres ‘combat’ formal of outside killed often more are violence, organized of contemporary of victims proportion significant widely different estimates of the burden of conflict conflict of burden the of estimates different widely andleadto conflict of victims of majority the for Violence Armed of Burden Global the of edition inthis briefly only with dealt are conflict—which from deaths non-violent Indirect conflict. armed of aconsequence as populations civilian by suffered mortality orexcess deaths violent) (non- indirect the as well as groups), and armed states by victimized ordirectly zones, conflict in trapped (either andcivilians combatants both of deaths violent the through violence armed of burden global tothe contributes violence Conflict consequences. lethal the counting of ways different the captures category last the involved; andmotives actors of types different the captures distinction second-order the violence; of organization of levels different (AI, 2009, p. 22; Foglesong and Stone, 2007, and Stone, p.18). p.22;Foglesong 2009, (AI, deaths violent of cent 13–43 per for account can they however, orNigeria, Jamaica suchas places in inallcountries; statistics inhomicide counted systematically not are killings judicial andpolice extra- example, For levels. different between ‘translation’ isanimperfect there violence, armed of incidence actual the of picture an adequate toprovide likely orless more are violence lethal counting of ways various the While forces. security state by killings or‘legal’ extrajudicial as well as (manslaughter), killing der, andunintentional ormur- homicide intentional of interms captured iscommonly violence ornon-conflict Criminal in2011. East Middle andthe Africa North of parts across visible was as demonstrators, unarmed against violence orinstate in 2008, 5 Civilian non-combatants, who form a form who non-combatants, Civilian 6 of post-election violence in Kenya inKenya violence post-election of —can account account —can 7

13 2 3 4 5 1 A UNIFIED APPROACH 14 Figure 1.1 A framework of armed violence categories

ORGANIZED (COLLECTIVE) VIOLENCE INTERPERSONAL VIOLENCE

LEVEL OF ORGANIZATION

Violent (State) violence Terrorist Communal Gang and Domestic and Inter-state and assaults and against attacks and and non-state organized crime intimate partner civil war armed civilians violence violence violence violence ARMED VIOLENCE 2011 VIOLENCE ARMED robberies of

LETHAL OUTCOMES GLOBAL BURDEN BURDEN GLOBAL CONFLICT DEATHS Victims of extrajudicial killings NON-CONFLICT DEATHS

CIVILIAN Battle-related deaths CONFLICT Civilian deaths DEATHS Victims of terrorism INTENTIONAL HOMICIDES

Victims of legal INDIRECT CONFLICT interventions DEATHS UNINTENTIONAL HOMICIDES

deaths. The Democratic Republic of the Congo is Report 2009/2010 revises these figures and estab- an interesting, yet controversial case. The Inter- lishes an estimate of up to 2.4 million deaths national Rescue Committee, for example, finds (HSRP, 2010, part II, p. 38). The first edition of that up to 5.4 million people died between 1998 the Global Burden of Armed Violence suggests a and 2007, of which only 10 per cent were victims global average ratio of four indirect deaths for of violence (Coghlan et al., 2006, p. 44; IRC, 2007, every direct (violent) death due to armed conflict, p. ii). On the other hand, the Human Security although this depends heavily on the nature of the to which Iraqis have been subjected. been have Iraqis to which violence’ ‘everyday tothe compared as especially victimization, violent for estimates leads tolow agenda— apolitical with group armed recognized a by claimed are that attacks deaths’—violent ‘conflict counting on focus anarrow example, for 2010). Iraq, In (Stepanova, actors same the by perpetrated andbe simultaneously, present be also can violence armed of forms Different 483). pp.480, 2003, (Kalyvas, resources and land over ordisputes feuds, orclan family rivalries, local tosettle violence touse tions motiva- own their with parties external co-opted actors local cases, both In bombed. be would targets these that so presence Taliban orAl-Qaeda alleged about USforces informed factions local inAfghanistan, villagers; local by provided lists name on based perpetrated were killings instance, for war, civil Guatemalan the During overlap. can motives how of examples striking provides denunciation of practice The situation. violent inone present be can andrespect) resources, revenge, (suchas motives andother ideological, identity-based, economic, political, different motives; andoverlapping multiple have can violence Yet armed boxes. neat the one) of only (and in one fits incident violent particular any that misleadingimpression the give they is that first The settings. indifferent violence to armed responses andprogrammatic practical effective todevelop ability our hinder distinctions sharp and classifications convenient why reasons four are There picture? this with iswrong what So p.32). 2008, Secretariat, Declaration (Geneva response humanitarian andthe conflict based conflict, non-state conflict, and one-sided andone-sided conflict, non-state conflict, based state- datasets—covering (UCDP) Program Data Conflict Uppsala combined The inIraq. alone 2006 for killed 27,000 civilians records example, for Project, Count Body Iraq the by provided 8 The data data The similar findings appear throughout the 20 the throughout appear findings similar 1927); (Abbott, waves’ ‘crime non-conflict into over spill can violence wartime that note War, analysts USCivil the as back Yet far as causes. underlying similar share ormay linked, be may violence of forms different inwhich ways the tosee difficult is it isthat result One andconditions. survivors, victims, perpetrators, of system aparticular within self-contained as violence armed of forms different treat categorizations andexclusive rigid Finally, 1.1. inBox alysed an- Somalia—is of coast the off the piracy rise of example— recent, more Asecond, p. 294). 2010, (Malaquias, violence protracted tothe a support provided trade diamond coldwar, the the of end the at smallarms of availability unprecedented an with combination In mining. diamond on based activity predatory more favoured that constraints andstrategic political faced in1992,UNITA process electoral failed the after resumed war civil the Once p.347). 2009, Beck, (Koloma arena coldwar inthe USsupport from benefited it such, as region; inthe expansion’ Russo–Cuban the ‘against struggling as itself defining tone, aadopted clear anti-communist group the 1970s, p. 2010, 296). Inthe (Malaquias, group ethnic largest Angola’s Ovimbundu, the to avoice providing movement a revolutionary as emerged first UNITA example, for Angola, In actors. violent of andcapabilities motives of ity fluid- the reflect that andlulls peaks with tempo ashifting following often time, over to another form one from change also can violence Armed Colombia, the Democratic Republic of the Congo, Congo, the of Republic Democratic the Colombia, andHerzegovina, Bosnia as in suchplaces women against violence of andlevels patterns the that torecognize begun have researchers recently, 1976). More andGartner, (Archer tury the same period. for intotal deaths 4,261 only violence—register 9 th cen- 15 2 3 4 5 1 A UNIFIED APPROACH 16 Box 1.1 Somali pirates: bandits and soldiers of convenience Photo Armed Somali pirates prepare a skiff in Hobyo for future attacks, Somalia, January 2010. © Mohamed Dahir/AFP Photo Armed violence has been pervasive in Somalia since the 1991 collapse of the state and the ousting of Siad Barre’s government. Thousands of people, especially civilians, have suffered directly and indirectly from the armed confrontations among the numerous factions at war in the country. The heavy human toll of the early years of conflict resulted both directly from armed violence and reprisals against civilians, and indirectly from the devastation of farmland, which brought starvation upon thousands of Somalis and displaced tens of thousands of people.

The violent collapse of the Somali state is also linked in complex ways to the increasing acts of piracy in the Gulf of Aden. Initially, with the disappearance of any state surveillance of Somalia’s maritime waters, fishermen took up ARMED VIOLENCE 2011 VIOLENCE ARMED arms to oppose illegal industrial fishing and international waste disposal of off the coast. These armed groups rapidly realized that unarmed commercial vessels represented a convenient opportunity for enhanced income gen- eration (Lennox, 2008, p. 9). Initial small-scale attacks quickly grew more sophisticated, peaking with the seizure of the Sirius Star, a tanker with an estimated USD 100 million worth of cargo, and the seizure of the Faina and its cargo of 33 Russian T-72 tanks, weapons, and ammunition (Balakrishnan, Rice, and Norton-Taylor, 2008; Höges, Klussmann, and Knaup, 2008).

The scope and nature of the attacks are also influenced by the availability GLOBAL BURDEN BURDEN GLOBAL of weapons and the environment of pervasive insecurity. The International Maritime Bureau finds that the 213 reported attacks by Somali pirates in 2010 was double the figure for 2008 (and four times higher than the 2007 figure). Furthermore, Somali pirates perpetrated 48 per cent of worldwide reported attacks. Indeed, for the years 2009 and 2010, Somali pirates were responsible for an average of 80 per cent of all attempted attacks, and an average of 25 per cent of successful acts of piracy. Overall, they were responsible for half of the attempted and actual attacks worldwide (IMB, 2010; 2011).10 The Somali pirates were also more violent than their peers: although they account for only half of the attacks worldwide, Somali pirates accounted for 86 per cent of the hostage-taking, all of the deaths, and 78 per cent of the attacks involving guns in 2010 (IMB, 2010; 2011).

Piracy is by definition a violent act serving private economic interests, but in practice Somali pirates are entangled in local conflict dynamics.11 Though pirates have kept some distance from the ongoing civil conflict, recent reports show that these groups have bolstered their armaments, and that local government officials as well as opposing militias are increasingly relying on the pirates’ firepower and strength for carrying out protective and predatory tasks (Gettleman, 2010).

The involvement of piracy—commonly associated with international organ- ized crime—in the Somali conflict illustrates how blurry the distinction between criminally motivated and ideologically motivated violence has become. Though Somali pirates have so far not been directly involved in killings in the ongoing conflict, the fact that they are linked to the warring factions represents a serious threat to safety and security in the region. 17 2 3 4 5 1 A UNIFIED APPROACH 18 El Salvador, or Iraq may be conditioned by the arresting youths suspected of gang membership’ experiences of war and deeply entrenched con- (Rodgers, Muggah, and Stevenson, 2008, p. 16; flict dynamics.12 And the high suicide rates among Bateson 2009, p. 7). Yet data in El Salvador, for members of the US armed forces in the past example, shows that only around 13 per cent of decade are linked to wartime experiences in com- homicides in 2008 were attributed to gangs plex ways (USDoD, 2010). (IML, 2009, p. 70). The focus on gangs, however, means that other violent acts are left aside, such These four aspects—the multiple, simultaneous, as the 93 extrajudicial killings reported in 2006, and shifting motivations of violent actors, and the and the more recent escalation of drug-related links between different forms of violence—con- violence linked not to gangs, but to organized found simple classifications and policy responses. crime (Aguilera, 2008, p. 134). Facing criticism Rigid distinctions and categorizations lead to policy ARMED VIOLENCE 2011 VIOLENCE ARMED and lack of effect, heavy-handed approaches

of stovepipes, in which policies and programmes to have started to be combined with mano amiga deal with one sort of armed violence (gang violence, (Friendly Hand) and mano extendida (Extended for example, or conflict prevention strategies) are Hand) interventions, which focus more on incen- developed in a narrow fashion that disregards the tives for demobilizing gangs and the establish- way in which different forms of armed violence ment of stricter controls on small arms (Rodgers, can be closely linked. In Liberia, for example, Muggah, and Stevenson, 2008, p. 16). Finally, much of the post-war effort focused on more policy stovepipes can also lead to category GLOBAL BURDEN BURDEN GLOBAL traditional demobilization, disarmament, and errors, in which programmers misunderstand reintegration, as well as security sector reform. or mistake the form of or motivation behind the Post-war Liberia, however, faces other serious violence encountered, leading to inappropriate challenges; high unemployment, a large youth programmes or responses. population, and severe development needs all represent sources of discontent and risk factors for potential armed violence. To date, post-war security promotion efforts have not tackled Obstacles to a unified perspective these issues as a way of reducing the risk of the Achieving a unified perspective on armed more criminalized forms of violence (Small Arms violence is difficult. Policy-makers require Survey, 2011a, p. 2; 2011b). boundaries within which to structure practical Policy stovepipes also result in blind spots, in programmes. Public health practitioners, for which the most important types of violence may example, rely upon the World Health Organiza- be neglected due to the policy predispositions or tion’s epidemiological model, which focuses on orientations of major donors and stakeholders. risk and resilience factors while seeking to identify For example, strategies to deal with homicidal factors that can be addressed at the individual, violence in Central America have focused on community, and societal level (WHO, 2002). The violent gangs, neglecting the range of other vio- World Bank’s World Development Report 2011: lent acts and actors in the region. The initial Conflict, Security, and Development relies mainly response to violence in the region ‘can be char- on data relating to conflict, and only to a much acterized as enforcement-first’ measures, ranging lesser extent on data and analysis relating to from tougher prison sentences to ‘aggressively homicide and organized crime (World Bank, 2011). These challenges can be overcome through through overcome be can challenges These andconditions. contexts tospecific and adapted migrated successfully be can practices and best solutions promising which tolearn platforms in‘multi-stakeholder’ of actors range a wide together tobring policy-makers challenges goal this Furthermore, communities. and between within andinsecurity violence of patterns larger into fits interest particular their where on spective per- andgain focus, their widen predispositions, their aside toset communities research different requires It challenge. andananalytical practical a both isthus perspective aunified Achieving (IISS). Studies Strategic for Institute International and the Ploughshares, Project UCDP, suchas centres, research society and civil academic of aseries by isconducted conflict on research andCrime; Drugs on Office Nations United the suchas andorganizations nologists crimi- of realm isthe example, for homicide, on Research purposes. different andfor standings under- todifferent according violence armed of anddistribution scope the on data collects them eachof but base, andstatistical literature asolid produced have communities these years, the Over toanthropology. toeconomics relations international andfrom health, topublic nology crimi- from ranging disciplines, of array in awide isabundant women against andviolence lence, vio- gang crime, conflict, on Research disciplines. between communication little with violence, armed of manifestation oranother one understanding in specialize andanalysts researchers Similarly, erupted. has conflict’ ‘armed formal a inwhich contexts few those on mainly focuses community and peacebuilding prevention conflict the And programming. andreduction prevention violence other from inisolation works generally women against andviolence violence based gender- of issues on working community The 14 violence (Gilgen and Tracey, 2011, pp.38–39). andTracey, 2011, (Gilgen violence and reduce toprevent andstrategies issues on discussions andcatalyses community, the into information this brings hotspots, violence and crime identifies It sources. local from data and data crime police with hospitals from data links example, for Observatory, Crime Jamaican The processes. topolicy-making directly analysis and linkresearch which observatories, violence or systems monitoring violence armed effective argue that non-violent means of resolving con- resolving meansof non-violent that argue Others institutions. multilateral of efforts building andpeace- peacekeeping ofthe success steady but slow tothe linked are declines these that argue observers Some pp.148–51). 2005, Gleditsch, and Lacina 2011; Bank, (HSRP,war World 2010; major of end the toherald andeven war of costs human shrinking the to highlight analysts some led has This conflicts. of lethality overall the by measured as globe, the around conflict armed of intensity the as well as incidence inthe decline areported witnessed also have decades past The were these of All in2009. active conflicts armed state-based 36 with stabilized, has ongoing are and that states involve that conflicts armed of number the Sincethen, 1990s. early inthe peaking increased, steadily wars andcivil conflicts armed internal of number the while decades, recent over declined have states between warfare of forms Traditional decades. two past the over documented well been have violence armed of forms organized highly most the of andintensity scope, inlocation, Shifts violence in contemporary conflicts Neither war nor peace: armed side states (Pettersson and Themnér, 2010, p.16). 2010, andThemnér, (Pettersson side states out- from combatants involved them of and seven internal conflicts, conflicts, 15 19 2 3 4 5 1 A UNIFIED APPROACH 20 ARMED VIOLENCE 2011 VIOLENCE ARMED of GLOBAL BURDEN BURDEN GLOBAL

Photo Supporters of the opposition take part in post-election protests in Kibera slum, Nairobi, Kenya, January 2008. © Shaul Schwarz/Getty Images pp. 312, 342). 2008, (CIPEV, killed) persons (744 overall victims of numbers largest the suffered also deaths), which allfirearms-related of cent or47per (194 victims Valley Rift inthe highest were victims gunshot of numbers the allkillings); of cent per (35.7 Kenya across cases in405 death of cause the as identified were orgunshots firearms addition, p.237). In 2008, (CIPEV, in2007–08 violence post-electoral the during abuses sexual and other rapes perpetrated reportedly andpolice army Kenyan the p.200). Furthermore, 2008, (Mueller, game’ ethnic zero-sum in a‘winner-takes-all organized been have traditionally politics party where inanenvironment violence inthe pated Mungiki—partici- Taliban Kikuyu Luo andthe the as gangs—such aligned) (and ethnically Active estimated 350,000 persons (CIPEV, 2008, p.272). 2008, (CIPEV, persons 350,000 estimated an anddisplaced lives 1,000 than more claimed 2007 of elections disputed the after erupted that clashes violent the example, for in Kenya, 2008 In state. the of involvement direct the out with- population, the of segment aspecific target orthat eachother, confront that groups armed organized formally orless more of ing number increas- isthe captured easily isnot that conflicts armed contemporary of feature important One violence. tive collec- of outbreaks slideinto can countries how andof change, political of nature divisive the of examples recent two only are 2010 November of elections presidential contested d’Ivoire’s Côte after clashes armed 2011of andthe Spring Arab violent The tenuous. remains wide world- violence conflict-related inlethal drop the explanation, the Whatever interactions. violent replacing slowly are states indemocratic flicts or militias) as well as the Kenyan security forces. security Kenyan the as well as or militias) gangs, groups, ethnic (suchas groups non-state identified eight least at involved violence The 17 16

21 2 3 4 5 1 A UNIFIED APPROACH 22 This rise in armed group activity can be illustrated economies are all indicators of the complex web more generally in several ways. One way is to of violent actors involved in different forms of vio- track the number of non-state armed actors around lence at the local, national, and regional levels. the world, which, according to some estimates, The heightened activity among organized armed may reach more than 1,000 different groups, groups can also be illustrated by the number of large and small, around the globe (Carey and armed conflicts that do not involve governments. Mitchell, 2011, p. 1; IISS, 2009, pp. 465–74; Although the figures fluctuate widely from year 18 UCDP, 2011). Figure 1.2 illustrates that while the to year, the number of non-state armed conflicts overall number of international and civil wars has reached 35 in 2008—more or less the same as been stable or declining since the mid-1990s, the number of armed conflicts involving states the number of active pro-government militias (HSRP, 2010, ch. 11, pp. 5–22).19 Although mostly ARMED VIOLENCE 2011 VIOLENCE ARMED around the globe increased steadily from the of small-scale, this form of violence poses signifi- early 1980s through 1990s, only to decline since cant security and development challenges; the 2003. The increasingly complex power struggles 2008–08 post-election violence and political and the atomization of actors and political con- uncertainty in Kenya, for example, had a direct figurations have been noted in conflict settings impact on the economy.20 After several years of such as Darfur, Iraq, and Pakistan (Stepanova, high annual GDP growth (more than six per cent 2010). Predatory groups, counter-insurgency per year in 2006 and 2007), growth dropped to

GLOBAL BURDEN BURDEN GLOBAL operations undertaken by states, insurgent 1.6 and 2.6 per cent in 2008 and 2009, respec- actions, and criminal activities linked to war tively (World Bank, n.d.).

Figure 1.2 Active pro-government militias and the number of civil wars, 1981–2008 Number of pro-government militias (PGM) Number of civil wars Number of countries with at least one active PGM

Number

120

110

100

90

80

70

60

50

40

30

20

10

1981 1982 1983 1984 1985 19861987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Source: Carey and Mitchell (2011) than during a preceding armed conflict, as in as conflict, armed apreceding during than higher or even high as rates mortality to reach violence peacetime for uncommon isnot It 1.2). Box (see Iraq of case the by illustrated as wrong, are beliefs three these of ormore one contexts inmany But violence. andnon-conflict conflict of forms between links strong no are there that and settings, inconflict higher are victimization and violence of levels that improves, and safety andsecurity decline violence of levels overall terminate conflicts when that beliefs: three on based isgenerally violence andnon-conflict flict con- between divide andprogrammatic analytic sharp The settings. post-conflict in so-called occur can that violence armed of transformation the recognize must analysts conflicts, armed contemporary of nature changing tothe addition In p.13). 2010, p.225; UN, 2005, andStuddard, (Nitzschke agreements or peace cease-fires tonegotiate interventions targeted narrowly tolaunch difficult ismore it complex, more become them between links the and as increases, context inaparticular actors violent of andnumber type the As opposition). suppress (to orstate-led groups), armed (formal predatory and organized highly orresources), land over fight groups neighbouring (aswhen organized loosely and opportunistic be may violence The conflict. and insecurity of context overall tothe another or way in one linked are they but organization, formal of levels different have may groups the violence; of pockets of eruption the by terized charac- fire toabrush compared be can groups armed non-state involving conflicts Rather, gain. oreconomic change political ameanstoachieve as violence uses that goal andcommon ideology lar aparticu- with actor homogeneous aunitary, as group anarmed ideaof the question into calling andstrategies, goals collective defined easily less andhas stable isless group non-state A violent in many ways. in many linked be can violence andpost-conflict Conflict pp.37, andTobón, 2011, 46). (Restrepo police Guatemalan the by reported were cide victims homi- 6,498 year. per Yet alone, deaths in2009 3,508–5,800 into translates which killed, were people 119,300–200,000 war, anestimated civil 36-year the During Guatemala. contemporary interact in complex (and poorly understood) ways. understood) (and poorly incomplex interact and causes root share can violence of ifestations man- anddifferent blurred, become may motives andcriminal political time, Over p.141). 2010, Survey, (SmallArms situation security the of tion aggrava- toanoverall contribute also can settings inpost-war weaponry accessible Easily p. 308). 2009, andMuggah, (Rodgers time over existence groups’ these tosustain as well as level individual the at generation income for anoption become tohave known are activities p. 130). Criminal 2008, Secretariat, Declaration (Geneva violence from communities ortheir themselves to protect groups vigilante-like into themselves organize to people push also can uncertainty economic andsocio- insecurity heightened of situations to combatants former demobilized of return The conflict. the of settlement the after nal violence crimi- of continuation tothe also often but itself, conflict armed of duration tothe only not tribute con- economies towar related networks illegal importantly, Most andLiberia. Guatemala such as incountries andvigilantism lynching included has that justice andrough policing toinformal led have capacities policing effective of lack and the institutions state of breakdown the Similarly, inslumsorshantytowns. cleansing operations andsocial killings extrajudicial of incidence toahigher haveled andMozambique Salvador, El Colombia, suchas incountries conflicts during groups paramilitary of formation andthe sector 21 The militarization of the security security the of militarization The 23 2 3 4 5 1 A UNIFIED APPROACH 24 Box 1.2 How different forms of violence are linked in Iraq In post-invasion Iraq, the widely reported target- ing of non-combatants by insurgents, militias, and sectarian groups may have appeared cha- otic or random at first glance. Yet a closer look at underlying patterns and motives of violence suggests that this seemingly arbitrary or criminal violence may also serve ‘clear political purposes’ aligned with the political goals of these armed groups (Green and Ward, 2009, p. 612).

ARMED VIOLENCE 2011 VIOLENCE ARMED Such ‘dual-purpose violence’ in Iraq is charac- of teristic of both the politically motivated violence of insurgents and organized criminality since 2003. Individuals, often with a criminal back- ground, ‘prefinanced’ future insurgent activities by participating in the generalized looting shortly after the fall of the Ba’ath regime, and impover- ished looters targeted the homes of the political elite ‘in acts of political revenge but also to satisfy GLOBAL BURDEN BURDEN GLOBAL long accumulated material needs’ (Green and Ward, 2009, p. 618). Notorious organized crime figures have reportedly helped insurgent cells to fund their activities through kidnappings, bribery, and highway robberies (p. 619).

Violence against women has also been used to serve sectarian or political ends, demonstrating how common criminality, individual motives, and collective violence interact in Iraq. Under the Ba’ath regime the hijab (women’s traditional dress code) was forbidden and women’s rights were a ‘bargaining chip’ for the political elite. While advancing a rather progressive position in terms of women’s rights and participation in society and the workforce, the Ba’ath regime ‘accepted tribal practices’ in return for loyalty from local leaders (Green and Ward, 2009, p. 614).

Since 2003, traditional or sectarian practices against women have been used as part of politi- cal struggles. When a faction takes control of a

territory, the imposition of the veil or a strict dress Photo A woman stands next to the grave of her daughter, a Sunni code for women is usually among the first meas- who was killed by Shia militiamen, Baghdad, March 2008. ures announced as part of a wider ‘campaign of © Ghaith Abdul-Ahad/Getty Images cially through ‘honour’ killings and rape. and killings ‘honour’ through cially espe- violence, of levels high suffer and groups different among polarization increased of ments instru- and victims both are women case, Iraqi the In conflict. ongoing overall the of umbrella the under women against violence in engage to actors various for opportunities increased has Iraq in conflict The violence. for motives private and personal as well as political modate accom- acts violent individual how demonstrate women against violence of forms various These p.16). 2008, UNAMI, 2007; (USDoS, including self-immolation incidents, related ‘honour’- in burnt 150women and killed women Kurdistan Regional Government 56 reported the meanwhile, behaviour; un-Islamic allegedly their for dumped bodies their and killed were 2007,women In 57 Kurdistan. Iraqi of region northern the and Basra in concern particular of been have killings ‘Honour’ 2008). (Sarhan, murders sectarian like appear them make and killings out carry to gunmen professional hired had members family reported, lawyer Iraqi an cases, some 2007. 47in In versus murdered, been had 81women 2008, November to January From rise. the on were Basra of city southern the in ings UK-based the 2008, In p.615). 2009, Ward, and (Green opponents their upon harm and inflict intimidate maliciousdenunciationto have used conflict in families of members that reported been also has it but kin, their upon shame brought have who women denounce often members Family p.15). 2008, (UNAMI, militias different with affiliated are who guards university of attention the attracting avoid to order in families their by codes dress conservative adopt to pressed are students female many that reports Iraq for Mission Assistance Nations United The terms. sectarian in ‘dishonour’ define that groups armed by trated perpe- increasingly also are killings ‘Honour’ p.615). 2009, and Ward, (Green areas in militia-controlled terror’ Guardian reported that ‘honour’ kill- ‘honour’ that reported 25 2 3 4 5 1 A UNIFIED APPROACH 26 Neither peace nor war: violence in ward trend in Western Europe since the beginning of the early modern period. Homicide rates across post- and non-conflict settings Western Europe—in what are today Belgium, The incidence and severity of lethal violence in England, Germany, Italy, the Netherlands, Scan- non-conflict settings has also undergone signifi- dinavia, and Switzerland—declined by roughly cant changes over the past few decades, yet the half from the early 17th to the early 18th century, picture is more complex than that presented for and by the 19th century, they had dropped three armed conflict. The limited empirical data on to five times further (Eisner, 2001; Gurr, 1981; historical homicide rates reveals a regular down- Monkkonen, 2001).22 Although the exact timing

Figure3 1. Aggregated homicide rates in 13 selected Western European countries, 1970–2009 (base year=100) ARMED VIOLENCE 2011 VIOLENCE ARMED

of Change (basis year 1970=100)

200

180

160

140

120

100 GLOBAL BURDEN BURDEN GLOBAL

80

60

40

20

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986

Change (basis year 1987=100)

200

180

160

140

120

100

80

60

40

20

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

Note: The rate for the initial year is indexed at 100 to facilitate trend comparison. Source: Small Arms Survey elaboration based on UNODC (n.d.b.) one homicide per 100,000 persons. 100,000 per homicide one around of level isalow all13 countries for rate average the 1987. of 2009, rate For homicide initial the below rates Spain) of exhibit exception the (with review under 13 the countries of majority the 2009, in 1970. By than in1986 rate lower a have andSpain Finland, Austria, only rates; homicide recorded inthe increases severe suffer countries several 1970 and1986, Between worthy. note- also are trends level, country the At years. six injust cent 40per decreased rates homicide in the early 2000s. early in the todecline begin and eventually stabilize, then 1990s, early the until toincrease continue Rates trends. on amajorimpact have can cide defined orhomi- isrecorded data inhow small changes meanthat inEurope rates homicide low very the since caution some with considered tobe have figures these although andSwitzerland, in Norway seen are increases highest The 100,000. per cides homi- three around of arate reaching and 1986, 1970 between cent per to80 close by increased gradually rates homicide that reveal figures The trends. of comparability the to ensure year first the for to100 set are rates country All difficult. very andrates trends of analysis renders and observed countries the throughout rates of fluctuation inasignificant isvisible which in1987, homicide of recording statistical the in change duetoanapparent and 1987–2009, 1970–86 periods, two into isdivided data The Systems Justice Criminal of Operations the the by collected data on based states, 13 European in rates homicide aggregated the of evolution the 1.3 illustrates Figure significantly. fluctuate do rates homicide however, term, medium the In states. European within violence inlethal decline long-term the about doubt islittle there place, to place from varies decline the of and scope United Nations Survey of Crime Trends Crime of and Survey United Nations 23 Between 2003 and 2009, and2009, 2003 Between (CTS).

2000s, reporting decreased first but then stabi- then but first decreased reporting 2000s, the of beginning the In improving. simply ing was report- infact when increasing, were rates global though as seem madeit andUkraine) Estonia, Colombia, (suchas rates homicide high orvery high relatively with states from data more of inclusion progressive the consistently; reporting were 67 states about 1990s, the By period. in this under-represented especially was which Africa, Sub-Saharan except regions allother from vided pro- were samples though Europe, from were most reported; 50states around only example, for 1970s, early the In rates. global the changes radically states of andgroups states of (and exit) entry the andsince survey, inevery report states no almost since trends global togenerate used be cannot Yet data this Trends Surveys. 11 Crime of series its from 1970–2008 period the for 163 countries covering data published has andCrime Drugs on Office Nations United The analysis. trend for basis the as serve could that data historical reliable of isalack there level, global the At p.210). 2008, (Spierenburg, wounds andother gunshot from mortality reduce that interventions medical emergency improved as including suchfactors emerged, have drop the for explanations several 2007); Zimring, 2010; Tseloni al., et 2004; (Aeby, crime violent on inliterature recognized largely been also has mid-2000s inthe andagain mid-1980s sincethe rates homicide of trend downward The 2002). andDrass, (LaFree crimes tocommit ties opportuni- resulting andthe wealth increasing of manifestation a‘normal’ orconsidered in Europe, change andsocial modernization torapid variously attributed been has 1980s 1970s andearly the in trend upward The term. andshort medium the in violence lethal influence factors what about questions raise fluctuations these term, longer the over low remain rates European Although 27 2 3 4 5 1 A UNIFIED APPROACH 28 lized at a higher level, with around 95 countries There are several possible explanations for these reporting homicide data by 2005–06. different—and fluctuating—trends at the European and global level. Most analysts point to the pro- It is thus difficult to derive long-term trends from gressive development of modern state institutions, global averages based on partial and shifting and the expansion of the state’s practical monopoly data alone. Coverage has been comprehensive over the legitimate use of force through security enough to present trend data since 2004, reveal- institutions such as law enforcement and national ing an upward trend in global homicides in the defence as one explanation for declining homicide second half of the 2000s (see Figure 1.4). As (and overall crime) levels. Not only has the state noted above, Western European countries have been able to intensify its presence over its terri- a decreasing homicide rate over this period, but tory, but scientific progress and institutional in other parts of the world, entire regions were ARMED VIOLENCE 2011 VIOLENCE ARMED reorganization of police forces have also improved

of suffering from generally increasing rates in homi- its capacity to contain crime and apprehend per- cidal violence. Despite the rather small magnitude petrators (Spierenburg, 2008, pp. 169–70). The of the increase since 2004 (about 5 per cent), the low levels of violence in Europe also seem to reflect global impact in terms of human lives lost is sig- changing normative understandings of the legiti- nificant. The difference between the low point of mate use of violence and the importance of personal 2006 and the figure for 2009, drawing on data security.27 Violent practices that were commonplace from the Global Burden of Armed Violence data- a century or so ago—such as public executions, GLOBAL BURDEN BURDEN GLOBAL base, represents more than 54,000 additional torture, and lynchings—have today been stigma- homicide deaths (and an increase of 24,000 tized to the point of near extinction in the West, since 2004). although there are notable exceptions. Similarly, the everyday use of violence to resolve conflicts has Figure 1.4 Global homicide trends, 2004–09 been condemned, even in what was hitherto con-

Homicide rate per 100,000 population (index year 2004=100) sidered the ‘private sphere’, where intimate partner

120 violence existed beyond the reach of the state.

110 Behind these figures lies a more complex reality,

100 one in which lethal non-conflict violence unfolds

90 with different levels of organization. Although non-conflict lethal violence is generally counted as 80 ‘homicide’, it is often linked to highly organized 70 criminal activity or to different forms of ‘political 60 violence’ that either target political opponents or 50 civil servants such as mayors, teachers, or police 40 officers, or that seek to modify government poli- 30 cies. In these contexts, the term ‘homicide’ is a 20 slightly misleading term since it conjures up images 10 of individual inter-personal acts of violence.

2004 2005 2006 2007 2008 2009 One result is that the distinction between the Source: GBAV 2011 database activities of politically motivated armed groups power-sharing offers may be less effective as as effective less be may offers power-sharing suchas compromises political criminalized, become have groups armed where circumstances in implications: policy important has shift This groups. criminal organized of model business self-perpetuating the resemble activities their and fade goals ideological purpose, political ent appar- without violence predatory its maintains Army Resistance Lord’s Ugandan orthe financing, self- for need andthe opportunities increased of because activities criminal towards turn Colombia in FARC the suchas groups motivated politically p.17). ostensibly Yet 2009, when (Steenkamp, funds’ toraise activity incriminal engage allsides on protagonists political as actors andpolitical criminal between relationship knit isa‘close- result One efforts. war their sustain to income generate groups armed how of ples exam- classic are narcotics of and distribution production 1.3). Box The see 2007; Wennmann, p.234; 2010, (UNODC, andLiberia Colombia, andHerzegovina, Bosnia Afghanistan, such as inplaces effort war tofinancethe used been have ingoods, trafficking suchas activities, Criminal p.42). 2010, (Stepanova, goods luxury such as ends ductive unpro- for profits use andoften business their of expansion the on focus hand, other the on tions, organiza- criminal change; andsocial political toachieve troops) their for andcare weapons (tobuy resources tomobilize activities criminal use Rebels activities. these from arising profit the of use planned inthe lies maindifference the tooperate, money require both p. 210). Although 2007, (Cornell, profit’ ‘monetary pursuing latter andthe causes’ higher ‘self-defined pursuing first the idealtypes’, ‘opposing as seen been have groups armed motivated and economically politically Traditionally, spectrum. the of ends both from blurred has organizations and criminal the struggle against the drug traffickers. drug the against struggle the abandoning into authorities the andtocoerce tions institu- local isto weaken goal sinceone political, and criminal isboth cause andthe frequent, are officers andpolice mayors suchas servants civil and politicians of Killings territory. inits drugs on awar towage army the mobilized has state Mexican the while north, tothe narcotics of flow the over control for inabattle locked are cartels Drug 1.4). Box (see inMexico war drug the by denced evi- as violence, of types andpolitical criminal between boundaries the blurring for capacity amarked shown andhas violence of levels high by isaccompanied narcotics, illicit of trafficking the andespecially crime, Organized p.303). 2003, Kemp, 2010; (UNODC, security ornational regional, tointernational, athreat as recognized increasingly resources—are environmental goods, counterfeit smallarms, humans, drugs, trafficking—in of forms Different profit-seeking. remains motive main ifthe even andconsequences, implications political have can groups criminal organized of activities violent the angle, another from Seen phenomena. separate than rather acontinuum as of thought best are groups andpolitical criminal between interactions the that imply also developments 2008). These (Goodhand, operations criminal established their from profit that entrepreneurs’ ‘violence toremain interesting more findit groups some terrorist financing (Jojarth, 2009, p.8). 2009, (Jojarth, financing terrorist as such activities, illicit to destined but legally obtained and legal are that goods and goods; stolen or diamonds conflict as such ways, illicit in obtained or processed goods narcotics; as such goods, illicit covers ‘trafficking’ term The B ox 1. 3 Defining illicit trafficking illicit Defining 29 2 3 4 5 1 A UNIFIED APPROACH 30 Box 1.4 Mexico’s drug war Violence and insecurity in the wake of President Felipe Calderón’s war on drugs have reached extremely high levels in several regions of Mexico. Official esti- mates set the human cost of Mexico’s drug war at around 35,000 dead from 2006 to 2010 (Booth, 2010; Turbiville, 2010, p. 124; Mexico, n.d.a).28

Mexico’s overall levels of violence have long been steady—with an average violent death rate of 11.5 per 100,000 persons in 2004–09 (ICESI, n.d.). But this national rate, while demonstrating that most parts of the country are generally safe, masks the bitter reality that some cities and regions suffer from extraordinarily high levels of violence, higher than found in many war

ARMED VIOLENCE 2011 VIOLENCE ARMED zones (see Figure 2.9, TRENDS AND PATTERNS). of At the state level, Chihuahua—home of Ciudad Juarez—had a rate of organized crime-related deaths of 98.6 per 100,000 in 2009 (with an overall homicide rate of 108.0); it is followed by Durango, which has a rate of 43.5 per 100,000 (Mexico, n.d.a).29 By contrast, Mexico City had an overall murder rate of 8 per 100,000, and a drug-related homicide rate of 1.5 per 100,000. Yucatán and Puebla are among the states witnessing the lowest incidence of drug-related violence and overall homicide rates, with rates below 1 per 100,000 for drug- GLOBAL BURDEN BURDEN GLOBAL related violence and ranging from 2 to 7 for overall homicide rates. See Maps 1.1–1.3 for the distribution of organized crime-related violence in Mexico (ICESI, n.d.; Mexico, n.d.a).30

Not surprisingly, the violence is concentrated where the drug cartels are most active. Four groups reportedly control most of the drug trade and other illicit activities in Mexico, with their influence reaching far into the United States. Several splinter groups and factions have also become central players in Mexico’s drug war; these are the Sinaloa cartel, the Gulf cartel, the Juárez cartel, and the Tijuana cartel, but also the Zetas, the Beltrán-Leyva organization, and the cartel Pacifico Sur, among others (Bunker, 2010, p. 11; Stratfor Global Intelligence, 2010).31 As Maps 1.1–1.3 illustrate, violence in Mexico is highly concentrated, although it also fluctuates and spreads. Thus states not affected by drug-related violence in one year can have very high figures the next. For example, Tamaulipas recorded a total of 90 drug-related deaths in 2009, yet one year later this figure climbed to 1,209.

A war on two fronts is taking place in the regions where the cartels are active. On the one hand, they are in violent confrontation with each other over the control of the lucrative trafficking routes between the Andean regions and the United States, as well as other forms of revenue generation such as extortion, Photo Mexican soldiers burn marijuana seized during a clash kidnapping, and human trafficking (Bunker, 2010, p. 11). On the other hand, with drug traffickers in Tijuana, Mexico, October 2010. the Mexican government has openly declared war upon drug traffickers, with © Francisco Vega/AFP Photo record specifically targeted violence against state state against violence targeted specifically record hand, other the on Aggressions, forces. security and enforcement law with confrontations violent as well as groups criminal between battles gun both register Confrontations deaths. drug-related of level alesser for only account hand, other the on aggressions, and Confrontations body). the with isleft amessage when example, (for cartel a drug of isamember perpetrator suspected or known the when violence drug-related as coded also are they but members, cartel drug at directed typically are events These body. the of mutilation and torture, of traces gunshots, multiple as such violence, of levels high involving killings intentional captures which ‘executions’, as iscategorized violence recorded of majority vast the that reveals figure The violence. of types by 2009 in tribution by disaggregating the drug-related deaths dis- this shows 1.5 Figure p.12). 2009, Intelligence, Global violencemore (Stratfor episodes of overt even to way given have power of readjustments and struggles resulting the divided, and weakened been have cartels drug the While results. ambiguous Legend: 2009 Mexico, violence, of type by killings crime-related organized Figure Source Executions (8,906; 93%) (8,906; Executions 6%) (596; Confrontations 1%) (111; Aggressions Mexico (n.d.a) : Mexico 1.5

Distribution of victims of of victims of Distribution XX

Map 1.1 Map 1.2 L California California California population Per 100,00 EGEND 40.93 L population Per 100,00 Baja Baja Baja EGEND <3 3–10 10–20 <3 3–10 10–20 20–30 >30 : California California :

Baja Baja Organized crime-related homicide rates, Mexico, 2007 Mexico, rates, homicide crime-related Organized Sur Sur Organized crime-related homicide rates, Mexico, 2008 Mexico, rates, homicide crime-related Organized Sonora Sonora Sinaloa Sinaloa Sin aloa Chihuahua Chihuahua Durango Durango Michoacán Michoacán Michoac Michoac Guerrero Guerrero án án 63.04 31 2 3 4 5 1 A UNIFIED APPROACH 32 Map 1.3 Organized crime-related homicide rates, Mexico, 2009 Zetas, a non-family-based group that originated with defectors from the military elite forces. First active as the Gulf cartel’s enforcement arm, the unit became increasingly independent, sub-contracting to other groups, until it broke its relations with the Baja Gulf cartel and became an independent and fully California Sonora established actor among the other cartels in Mexico

Chihuahua 98.63 (Killebrew and Bernal, 2010, p. 21).32

Baja California Reports on several carefully planned and executed Sur 39.93 SinaloaSinaloa raids requiring high levels of intelligence, resources,

ARMED VIOLENCE 2011 VIOLENCE ARMED Durango 43.47 and weaponry have appeared over the past few of LEGEND: years. These attacks are frequently accompanied by Per 100,00 unusual and gruesome displays of violence, regu- population larly including decapitations, mutilations, mass >30 20–30 MichoacánMichoacán executions, and extreme torture (Quinones, 2009; 10–20 Guerrero González, 2009). 3–10 <3 Politicians, law enforcement agents, civil society GLOBAL BURDEN BURDEN GLOBAL members, and their families and friends suffer the consequences of criminal violence. For example, Source for Maps 1.1–1.3: Mexico (n.d.a) Carlos Reyes López, a policeman of the state of authorities, law enforcement agents, and security forces by drug cartel members, Tabasco, was shot with ten members of his family resulting in the death of the targeted person (Mexico, n.d.b). in February 2009 (Wilkinson, 2009). In another case in Michoacán, the public safety minister, Minerva In response to the war on drug traffickers, cartels have implemented bolder Bautista, escaped death in an attack that killed four and more professionalized violent tactics. Over the past decade, drug-related of her colleagues and bodyguards in April 2010 violence has ‘acquired an increasingly organized and paramilitary character’, (Wilkinson, 2010). In April 2011, the Mexican poet making a ‘transition from the gangsterism of traditional narco hit men to para- Javier Sicilia lost his son, who was killed together military terrorism with guerrilla tactics’ (Turbiville, 2010, p. 124). Occasional with six of his friends in what seemed to be ‘collateral large-scale firefights, targeted assassinations, and military-style raids are damage’ of the drug war (Miglierini, 2011). common manifestations of the improved tactical competence and effectiveness of the drug cartels. These strategic improvements in the cartels’ capacities are Low incomes and the ready availability of money reportedly linked to the sweeping Mexican military reforms of the 1990s. In from drug cartels are important contributing factors response to low-intensity guerrilla activities and growing drug trafficking vio- to Mexico’s spiral of violence. Members of the state lence at that time, the Mexican state trained new elite commando units and or federal police may earn between USD 350 and assault troops, which were deployed in Chiapas, Guerrero, Puebla, and other USD 1,000 per month. Yet when 93 police officers regions of the country. A lack of careful recruitment, low salaries, and continuous were arrested in June 2009 for charges of corruption efforts by criminal organization to recruit security officers led to a substantial in the Mexican state of Hidalgo, some officers had number of defections by military personnel, who brought their skills and capaci- revenues reaching as high as USD 225,000 per ties to their new employers (Turbiville, 2010, pp. 128–32). An example is the month (Nagle, 2010, p. 100). Sources 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 Total homicide victims Figure clear distinctions between various types of vio- of types various between clear distinctions todraw impossible isnearly it involved, motives and actors violent different many the of view In p.56). 2010, Stepanova, 2008; (Ellingwood, conflict amajorarmed of andintensity scale the resembles that violence of inasurge resulted and officers, police federal 5,000 and anadditional soldiers 40,000 about of deployment the involved inMexico cartels drug against fight The 2006. inDecember traffickers drug on war declared and crime organized on down tocrack chose Calderón Felipe President violence, drug-related and corruption increased of challenges the Facing p.87). 2010, (UNODC, cartels drug Colombian the of andfragmentation dismantling the following position of Mexican drug cartels during the 1990s, the strengthened andMexico America Central towards routes trafficking drug of shift The reveals. Mexico of case the as however, zones, conflict to limited not are trafficking drug of quences conse- violent 2010). The Feldab-Brown, p. 209; 2007, (Cornell, andMyanmar Colombia, stan, inAfghani- conflict torecurring linked commonly andare ratio’ ‘value-to-size ahigh have Drugs Female victims victims Female 2003 PDH (2011); UNDP (2007) UNDP (2011); : PDH 1.6

Overall homicide and femicide levels in Guatemala, 2003–09 2003–09 inGuatemala, levels andfemicide homicide Overall Total victims Total 2004 2005 stood, but as Figure 1.6 shows, the increase increase the shows, 1.6 Figure as but stood, under- poorly are killings these of and causes nature p.79). The 2011, (PDH, in2003 killed edly report- women 383 the double isalmost figure that assaulted; orsexually tortured been having after many in2009, killed were women 720 an estimated example, for Guatemala, In pp.60–61). 2011, Bank, (World andRwanda Guatemala, Congo, the of Republic Democratic the as insuchplaces explored tobe beginning women—is against and violence violence orcriminal) (political organized between suchlinkage—that One violence. of forms other between mechanisms andtransmission interactions possible the highlight also violence and political criminal organized between intersections The bloodshed. the spared civilians innocent are nor immune, not are officials and government soldiers, officers, police killings, gang-on-gang be may deaths the of most While forces. armed andthe personnel military of deployment the through involved isdirectly it means, coercive over and monopoly authority challenge to its andtheir cartels drug of expansion the counter to attempts state Mexican the As inMexico. lence 2006 2007 2008 Totalfemale victims 2009 800 600 900 200 400 300 500 700 100 33 2 3 4 5 1 A UNIFIED APPROACH 34 appears to track the overall rise in criminal and Conclusion: a unified approach to gang-related violence in the country, which implies some connection between these different forms armed violence of violence (PDH, 2011).33 Different forms of armed violence interact in ways that go well beyond the simple dichotomy In Rwanda, an estimated 350,000 women were between political and criminal violence, or between raped during the genocide, translating to about conflict and non-conflict violence. As noted in 8,972 rapes per 100,000 women for the whole this chapter, the boundaries between violence adult female population; the risk of being raped categories are blurry and overlap, and they can for a Tutsi woman during this period was around reinforce each other in a vicious circle in multiple 80 per cent (Bijleveld, Morssinkhof, and Smeulers, ways. Violence can be ‘dual-purpose’, as high- 2009, p. 219). No comparable or reliable figures

ARMED VIOLENCE 2011 VIOLENCE ARMED lighted in the Iraqi and Somali examples; moreover, are available for the Democratic Republic of the of it can serve both ‘individual and organizational Congo, but the incidence of rape—as well as other goals’ (Green and Ward, 2009, p. 611). The con- forms of sexual and gender-based violence—is widespread (HHI, 2009, pp. 7–9). The prevalence cept of dual-purpose violence seems to apply to of rape is also anecdotally linked to the dynamics many settings in which large-scale acts of organ- of the conflict, including the exploitation of min- ized political violence coexist with individual and eral resources (Mukengere Mukwege and Nangini, criminal forms of violence, and where rape, looting, trafficking, personal revenge, and other forms

GLOBAL BURDEN BURDEN GLOBAL 2009, p. 3). Although poorly understood, the link- ages are important in policy terms. As Eriksson of opportunistic and criminal violence can be Baaz and Stern point out: observed.

the specific, often exclusive, focus on sexual Recognizing the multiple, simultaneous, linked, violence [. . .] hampers our understanding of the and changing forms that armed violence takes relationship between sexual violence and other is one step towards a unified approach to armed (supposedly) ‘ungendered’ violence . . . These violence prevention and reduction. An additional forms of violence are, to a large extent, mani- step is captured by the idea of a ‘system of vio- festations of the same systemic failures and lence’, in which high levels of violence and crime mechanisms as those contributing to [sexual in post- and non-conflict settings can be attrib- and gender-based violence] (Eriksson Baaz and uted to a series of intersecting factors present in Stern, 2010, p. 12). any given setting (Richani, 2007, p. 4 5). These These linkages—between organized political and factors include: organized criminal violence and between conflict low coercive and distributive state capaci- and sexual violence—highlight the ways in which ties, measured in terms of law enforcement armed violence can assume multiple, simultaneous, capacities (impunity and attrition rate) and and shifting forms that vary from place to place. Understanding how different forms of violence are the government’s social expenditures; linked is crucial to assessing the global burden low opportunity costs of crime derived from a of armed violence, and to developing effective lack of law enforcement and a lack of revenue- armed violence prevention and reduction strate- generating activities (such as education and gies and policies. job opportunities); 7 Figure opportu- new as catalysts—such specific in which context the provide acrime committing for costs opportunity andlow opportunities, economic weak enforcement, law for capacities state of lack the when occurs This together. come factors these when emerge can violence’ of A ‘system Source particular contingencies, such as the repatria- the suchas contingencies, particular Includes the unregulated availability and distribution distribution and availability unregulated the Includes (Richani, 2007, pp.4–5). (Richani, andAfghanistan inIraq suchas interventions orinternational andMexico, America Central to Caribbean the from routes trafficking drug of shift andthe America, toCentral States United the from immigrants illegal of tion of small arms and light weapons, mines, explosive explosive mines, weapons, light and arms small of remnants of war, and factors affecting their supply their affecting factors and war, of remnants OECD (2009, p. 50) p. (2009, : OECD 1.

The armed violence ‘lens’ violence armed The ntuet Agents Instruments informal (traditional and cultural) norms, rules, rules, norms, and cultural) (traditional informal Both formal institutions of governance and and governance of institutions formal Both societies affected by armed violence armed by affected societies Individuals, communities, and communities, Individuals, Institutions and practices People cycles of violence (World Bank, 2011). Bank, (World violence of cycles Report 2011 Bank’s World inthe lighted allhigh- are factors andcontingent underlying 2007, These p.5). (Richani, violence’ perpetuates and institutionalizes ‘that relationship dynamic a form which entrepreneurs, andpolitical crime), andorganized (gangs groups criminal sector, rity secu- violence’, the of suchas specialists the are units maininteracting ‘the inwhich emerge can violence of asystem 2002). suchasituation, In (Richani, spark the diamonds—provide such as goods’ ‘conflict exploitable oreasily trafficking, todrug linked generation income illicit for nities Perpetrators of armed violence and motivations for for motivations and violence armed of Perpetrators as playing an important role in repeated inrepeated role animportant playing as acquisition and misuse of arms and ofacquisition misuse arms (demand factors) World Development Development World National Regional Local Global 35 2 3 4 5 1 A UNIFIED APPROACH 36 Seeing different forms of armed violence as dual- This chapter has highlighted the way in which purpose, or as part of a broader system, opens the many different contemporary manifestations new opportunities for evidence-based research of armed violence blur the line between conflict and policy-making in various settings, ranging and non-conflict contexts. These various mani- from Jamaica to Afghanistan, Kenya to Haiti, and festations call out for a unified approach to Venezuela to Nepal. Piracy and warlordism amid measuring and monitoring armed violence that civil war and state collapse (Somalia and Afghan- captures the wide variety of actors, contexts, istan), drug-related violence and its political motivations, and consequences. The next chap- implications (Central America), shadow networks ter takes up this challenge by taking a broader in war economies (West Africa), as well as sexual perspective and presenting comprehensive and gender-based violence during conflict (Demo- national-level data that reflects how these

ARMED VIOLENCE 2011 VIOLENCE ARMED cratic Republic of the Congo) can all be better different forms of lethal violence are distributed

of understood by taking into account the complex around the world. It highlights not only that con- interactions between different forms of violence. flict deaths are a relatively small part of the global burden of armed violence, but also that the One potentially useful framework for designing majority of the most violent places on earth are policies and programmes is the Organisation for not found in conflict zones. Subsequent chapters Economic Co-operation and Development’s armed focus on how the armed violence lens can be violence ‘lens’ (OECD, 2009, p. 50). The lens pro- employed to map global and regional differences

GLOBAL BURDEN BURDEN GLOBAL vides a flexible and unified framework for appre- in patterns of homicidal non-conflict violence hending the contexts, motives, and risk factors (CHARACTERISTICS OF ARMED VIOLENCE), on associated with armed violence (see Figure 1.7). the state of knowledge about violence against Its three legs provide different entry points for women (WHEN THE VICTIM IS A WOMAN), and armed violence prevention and reduction policies, on the negative links between armed violence focusing on the perpetrators of armed violence and development outcomes (MORE VIOLENCE, and their motives, the instruments of armed vio- LESS DEVELOPMENT). lence, and the wider institutional environment that enables or protects against armed violence. Although the lens is not in itself a policy or pro- grammatic tool, it does highlight that a variety Abbreviations of factors can or do come together in different IISS International Institute for Strategic Studies situations to create an enabling environment in FARC Fuerzas Armadas Revolucionarias de Colombia which violence can occur and escalate, in both UCDP Uppsala Conflict Data Program conflict and non-conflict contexts. It also avoids UNITA União Nacional Para a Independência Total de debates about the specific categories or types of Angola armed violence; what counts, in the end, is to be able to assemble evidence regarding the various enabling factors of armed violence in order to design policies and interventions that effectively Endnotes address the serious challenges posed by armed 1 The definition also focuses on the physical use of force and violence, and deliberately excludes such concepts as violence to the safety and well-being of individuals structural, cultural, and psychological violence, however and communities. important they may be in other contexts. According totheUppsala Conflict Data Program, inthe 8 Countries such as Argentina, Colombia, and Costa Rica 7 oftheCommiss int Thereport ion ofInquiry 6 5 See the Thesetypes of violence, though by nomeans exhaustive, 4 According to von Clausewitz’s dictum, ‘war is thecon- 3 Thesefigures donot includetheglobal burdenof indirect 2 force between to anincompatibility, but must result fromuseof armed In territory). related totheoverarchin counted only if they result fromuseof armedforce directly case of state-based conflicts, battle-relat defence. See Table 2.4(TRENDS ANDPATTERNS). they arec offence. Nordoes theUnited States includethem,since by police duringconfrontations orcommitment of an such as France andNigeria, donot count civilians killed include extrajudicial andpolice (CIPEV, 2008,pp. 308–09). victims, of which 11werechildren and78werewomen Election Violence finds that the violence caused 1,133 (Geneva Declaration Secretariat, 2008,pp. viii–ix). for measuring thedifferent of direct andindirect conflict deaths andmethodologies for 2008(WHO, 2011). global burdenis estimated does not includeself-directed 6–7). For thepurposeof this research, ‘armed violence’ Moser andRodgers (2005,p. 5),and WHO (2002, pp. sources, includingHazen andHorner(2007, pp. 56–61), draw onthetypologies of violence published by several and Münkler (2003). armed violence, see Brzoska (2007, of the‘conflict’ lens as way toassess organized political Clausewitz, 197 tinuation of politics by other(i.e. violent) means’ (von AND PATTERNS). and extrajudicial killings, see Tables 2.3 and2.4(TRENDS homicide counts anddeaths dueto legaltions interven PATTERNS). For adetailed discussion onunintentional the humanitarian response; seeBox 2.6 (TRENDS AND status of theaffectedpopulation, andtherapidity of tion displacement, thebaseline health andnutrition conflict toconflict, dependingontheseverity of popula- conflict deaths, althoughthis rate varies widely from hovers around roughly facilities, food,andclean water dueto violence probably malnourishment, andageneral lack of access tohealth conflict deaths stemmingfrom conflict-related deaths. The estim Global Burden ofArmed Violence lassified as non-state conflict, thedeaths are warring factions 6). Ontheimplicatio ‘justifiable homic four times thenumberof direct g incompatibility (government or impacts of armedconflicts at 782,000 victims worldwide violence (suicide), whose (such as inMexico, where preventable disease, killings, whileothers, ated burdenof indirect app.2C),Ka ns andlimitations ide’, ed deaths are fordefinitions o thePost as isself- not linked ldor ( 1999), 1999), 3 In2008,2,831 homicides wereregistered by theInstituto 13 Eriksson Baaz and Stern (2009); Greenand Ward (2009); 12 Piracy is definedas ‘any illegal acts of violence ordeten- 11 In2010, Somali pirates perpetrated 148attempts and 10 9 The data 16 The report oftheCommiss int Thereport ion ofInquiry 16 These seven—called ‘internationalized intrastate conflicts’ 15 4 See,forexample, UNODC (n.d.a) andthesources forthe 14 605 victims of non-state armed violence (best estimates). For 2006,UCDPcounts 3,656 battle-related deaths and Count Project websites, res (IML, 200 a significant share of thesecases may begang-related for 1,910 cases in2008, maras, organg-related violence. It shouldbenotedthat in El Salvador. Of these cases, 354wereattributedto de Medicina Legal—the Instit Hume (2008); Sanchez et al.(2011); Wood (2003; 2006). private aircraft’ (Anyuand Moki, 2009, p. 95). ends by thecreworpassenge tion, orany act of depred (IMB, 2010;2011). representing 50percent of a 2009,January Somali pirates have perpetrated 430 attacks, amounted to169 attempts and48actual attacks. Since 65 successful attacks, whilein2009, thesenumbers Sundberg, 2010). version 1.3 (Eck andHultman, 2007; Eck, Kreutz,and Conflict Database version 2.3 andOne-sided Violence UCDP (n.d.); data examples herestemfromNon-state Luminoso inPeru). See thecodebooks foreach dataset at al-Qaeda attack of September 11thor victims of the Sendero formally organized groupagain use of armedforce by thegovernment of astate orby a one-sided violence registered between1993 and2008).Inthecategory of 751 deaths that wereduetointer-cartel warfare were 600,000 (CIPEV, 2008,p. 293; 2010, UNGA, p. 14). indicates 350,000 the figuresfordi are prone;meanwhile, IISS li under-counting towhich global incident systems reporting 258 fromnon-state violence), illustrating thedegreeof base only records 445deaths (187 fromone-sided violence; Election Violence lists 1,133 deaths, whilethe UCDPdata- ments involved, not thelocation of thefighting. Uganda, andtheUnited State —were Afghanistan, Algeri logical Annexe. direct conflict deaths database intheonline Methodo- was r 9, p. 70). etrieved formtheUCDP splaced persons,the wher , killings arerecorded incases of eas theUnitedNatio the motive ation, committed forprivate pectively (UCDP, n.d.; IBC,n.d.). a, Iraq, Rwanda, Somalia, sts 1,500 deaths. Regarding ll piracy attacks intheworld ute forForensic Medicine— s. Names reflect thegovern- st civilians (such as the rs of aprivate shipora was ‘unknown’; Commiss ion report and theIraqBody o thePost ns reports ns reports 37 2 3 4 5 1 A UNIFIED APPROACH 38 17 For data on identified state and non-state actors involved 26 On the early efforts to compile global data on crime, see in Kenya’s post-election violence, see the datasets on non- UNCJIN (n.d.). state conflict and on one-sided violence at UCDP (n.d.). 27 This point is explored in Norbert Elias’ thesis of ‘civilizing 18 These databases count 283 pro-government armed groups processes’, which links the general decline of lethal vio- (Carey and Mitchell, 2011) and anywhere from 500 to more lence to increasing degrees of ‘self-control’, based on than 900 non-state armed groups (IISS, 2009; UCDP, n.d.). individuals’ more detached and rational understanding 19 See data at HSRP (n.d.). of the world around them (Elias, 1994). 20 The report of the Commission of Inquiry into the Post 28 Estimates of the numbers have been rising significantly Election Violence lists 1,133 deaths, while the UCDP data- as well. In 2009, Felipe Calderón estimated that around base only records 445 deaths (187 from one-sided violence; 9,000 narco-linked murders occurred between 2006 and 258 from non-state violence), illustrating the degree of April 2009, while other estimates place the figure at around undercounting to which global incident reporting systems 11,000 people killed by early 2010 (Turbiville, 2010, p. 124). are prone; meanwhile, IISS lists 1,500 deaths. Regarding In April 2010, the Washington Post cited a confidential the figures for displaced persons, the Commission report report that estimates the human cost of drug-related

ARMED VIOLENCE 2011 VIOLENCE ARMED indicates 350,000 whereas the United Nations reports violence at 22,000 victims since 2006 (Booth, 2010). of 600,000 (CIPEV, 2008, p. 293; UNGA, 2010, p. 14). By early 2011, official government data counted 35,000 deaths, of which more than 15,000 occurred in 2010 21 On the complexity and multiplicity of links between alone. See Mexico (n.d.a) for the dataset on organized conflict and post-conflict violence, see Aguilera (2008), crime-related homicide victims. The Washington Post Cornell (2007), Restrepo and Tobón (2011), Jojarth (2009), subsequently indicated that Mexican news media counts Killebrew and Bernal (2010), Nitzschke and Studdard had surpassed the 40,000 threshold during the first half (2005), Rodgers and Muggah (2009), Small Arms Survey of 2011 (Washington Post, 2011). (2011), and Steenkamp (2009). 29 Drug-related violence is fluid and spreads easily. In 2010, 22 The homicide rate in England dropped from about 23.0 Chihuahua state exhibited the worst concentration of th th

GLOBAL BURDEN BURDEN GLOBAL per 100,000 in the 13 and 14 centuries to 4.3 per violence, with a rate of 129 per 100,000, whereas Sinaloa 100,000 by the end of the 17th, and 0.8 per 100,000 by followed in second place with a rate of 68 per 100,000. the first half of the 20th century. In the Netherlands and Belgium, contemporaneous figures are 47.0, 9.2, and 1.7 30 These numbers stem from two different datasets. One per 100,000; in Germany and Switzerland, rates fell from covers organized crime-related deaths (Mexico, n.d.a); 43.0 per 100,000 to below 2.0 for the 20th century (Eisner, the second reflects overall homicide deaths in Mexico 2001, pp. 618–38; Gurr, 1981, pp. 295–353; Monkkonen, (ICESI, n.d.). 2001, pp. 5–26). 31 The Sinaloa and Gulf cartels alone count an estimated 23 Trends in homicide rates can be considered in the light of 100,000 foot soldiers throughout the region, including the findings of LaFree and Drass (2002), who identify their enforcers or gang contractors (Bunker, 2010, p. 13). ‘crime booms’ between the late 1950s and the late 1990s 32 Increasing violence in Tamaulipas state, for example, is in more than one-third of their sample, which includes 21 linked to the battles opposing the Zetas to their former European nations. They define ‘crime booms’ as taking ally, the Gulf cartel. The Zetas are also known to have place in countries with (i) increasing homicide rates, expanded operations into Guatemala, forming alliances (ii) individual average annual growth in these rates of at with local gangs (Beittel, 2011, p. 10). least 10 per cent for any three successive years, and 33 See also Chapter Four on patterns and the extent of female (iii) sustained changes in the direction of the homicide victimization and resulting policy implications (WHEN THE trends (LaFree and Drass, 2002, pp. 780–81). By eliminat- VICTIM IS A WOMAN). ing countries that show some increase at some point, and by including reverse trends to evaluate crime booms, the authors still find that the period in question witnessed sig- nificant crime booms, though these were more pronounced in industrializing countries than in industrialized ones. 24 Recent improvements in data collection for violent crime have reduced the statistical errors that may be responsi- ble for some previously unclear or fluctuating patterns in homicide trends. 25 See UNODC (n.d.b) for criminal justice data from 1970 to 2008. Carey, Sabine andNeil Mitchell. 2011. Bunker, 2010. Robert. ‘Strategic Threat: Narcos andNarcotics Brzoska, Michael. 2007. ‘Coll Booth, William. 2010. ‘Drug Violence Appears inMexico’s North- Bijleveld, Catrien, Aafke Morssink Beittel, June. 2011.Mexico’s Drug Trafficking Organizations: Source Bateson, Regina. 2009. ‘ThePolitical Consequences of Crime Balakrishnan, Angela, Xan Rice, andRichard Norton-Taylor. Archer, Gartner. DaneandRosemary Anyu, J.Ndumbeand Samuel Moki. 2009. ‘Africa: The Piracy Hot- AI (Amnesty International). 2009. Aguilera, Gabriel.2008. ‘Enfrentar la violencia con “mano dura”: Aeby, Marcelo. 2004. ‘Crime Trend Abbott, Edith. Bibliography Militias? Working paper. June. March, pp. 8–29. Overview.’ University Press, pp. 94–106. andInternational SecurityDisarmament, Peace Research Institute. Definition of ArmedConflict.’ In Stockholm International article/2010/04/20/AR2010042004961.html> Comparative Politics Workshop. 14April. Guardian. 19 November.

Cornell, Svantee. 2007. ‘Narcotics andArmedConflict: Interaction Coghlan, Benjamin, et al.2006. ‘Mortality intheDemocratic von Clausewitz, CIPEV (Commission of Psychogenetic Investigations Criminology, cide Rates in Theoretical Perspective.’ Violence: The Long-Term Dynamics of European Homi- and Practitioners.’ NewOrleans, 17 February. annual meetingof ‘Theory vs. Policy? Connecting Scholars UCDP Non-State Conflict Dataset.’ Paper presentedat the of Peace Research Civilians in War: Insights fromNewFatality Data.’ Armed Violence. Geneva: Geneva Declaration Secretariat. RWP07-019. Cambridge, Mass.: University. Harvard April. Dominican Republic of CrimeandViolence: Les Contribution ofCriminalJustice Systems to the Control Washington, DC:FDDPress, pp. 90–112. Pakistan Theater: Militan Gartenstein-Ross andClifford May, eds. Asia: Beyond Taliban Profits andAfghanistan.’ InDaveed and theNordic Africa Institute. Working Paper onGender-based Violence. Uppsala: SIDA Violence in the Democratic Republic ofCongo (DRC) Vol. 53, No.2,pp. 495–518. Forces intheCongo (DRC).’ Rape? Masculinity, Violence, and Sexuality intheArmed latimes.com/2008/jun/0 for Control.’ Carl. h, pp. 207–27. Los Angeles Times Vol. 41, No.4,pp. 618–38. 1976. Inquiry intothePost Inquiry . Nairobi:CIPEV. , Vol. 44,No.2,pp. 233–46. . KSG Faculty Research Paper Series, On War. Translated by Michael Howard Sundberg. 2010. ‘Introducing the t Islam, Security andStability 3/world/fg-border3> Stern. 2009. ‘Why Do Soldiers sons from Jamaica and the International Studies Quarterly Princeton University Press. . Oxford: Blackwell Publishers. . 3June. 31 December 2009. London: International Chamber of Gilgen, Elisabeth and Lauren Tracey. 2011. Armed Violence Moni- Commerce. January. toring Systems: Contributing Evidence to Programming. —. 2011. Piracy and Armed Robbery against Ships: Annual Report, Geneva: Geneva Declaration Secretariat. 1 January–31 December 2010. London: International González, Franciso. 2009. ‘Mexico’s Drug Wars Get Brutal.’ Chamber of Commerce. January. Current History, Vol. 108, No. 715, pp. 72–76. IML (Instituto de Medicina Legal). 2009. Epidemiología de los Goodhand, Jonathan. 2008. ‘Corrupting or Consolidating the homicidios en El Salvador: Período 2001–2008. San Peace? The Drugs Economy and Post-conflict Peacebuilding Salvador: Corte Suprema de Justicia. October. in Afghanistan.’ International Peacekeeping, Vol. 13, No. 3, IRC (International Rescue Committee). 2007. Mortality in the pp. 405–23. Democratic Republic of Congo: An Ongoing Crisis. Green, Penny and Tony Ward. 2009. ‘The Transformation of Vio-

ARMED VIOLENCE 2011 VIOLENCE ARMED lence in Iraq.’ British Journal of Criminology, Vol. 49, No. 5.

of September, pp. 609–27. Jojarth, Christine. 2009. Crime, War, and Global Trafficking. Gurr, Ted Robert. 1981. ‘Historical Trends in Violent Crime: Cambridge: Cambridge University Press. A Critical Review of the Evidence.’ Crime and Justice, Kaldor, Mary. 1999. New Wars and Old Wars: Organized Violence Vol. 3, pp. 295–353. in a Global Era. Stanford: Stanford University Press. Hazen, Jennifer and Jonas Horner. 2007. Small Arms, Armed Kalyvas, Stathis. 2003. ‘The Ontology of “Political Violence”: Violence, and Insecurity in Nigeria: The Niger Delta in Action and Identity in Civil Wars.’ Perspectives on Politics, Perspective. Occasional Paper 20. Geneva: Small Arms Vol. 1, No. 3. August, pp. 475–94. Survey. Kemp, Walter. 2003. ‘Breaking the Crime–Conflict Nexus: A Chal-

GLOBAL BURDEN BURDEN GLOBAL HHI (Harvard Humanitarian Initiative). 2009. Characterizing lenge for the OSCE.’ In Institute for Peace Research and Sexual Violence in the Democratic Republic of the Congo: Security Policy at the University of Hamburg, ed. OSCE Profiles of Violence, Community Responses, and Implica- Yearbook 2003. Baden-Baden: Nomos Verlagsgesellshaft, tions for the Protection of Women. Final Report for the pp. 301–18. Open Society Institute. Cambridge, Mass.: HHI. August. Killebrew, Bob and Jennifer Bernal. 2010. Crime Wars: Gangs, Höges, Clemens, Uwe Klussmann, and Horand Knaup. 2008. Cartels, and U.S. National Security. Washington, DC: ‘Pirates versus Weapons: Dealers Looking for the Good Center for a New American Security. Guys off the Somali Coast.’ 10 June. in the Angolan UNITA.’ Contemporary Security Policy, HSRP (Human Security Report Project). 2010. Human Security Vol. 30, No. 2, pp. 343–55. Report 2009/2010: The Causes of Peace and the Shrinking Lacina, Bethany and Nils Petter Gleditsch. 2005. ‘Monitoring Costs of War. Pre-publication text. Vancouver: HSRP. Trends in Global Combat: A New Dataset of Battle Deaths.’ —. n.d. ‘Graphs and Tables: Report 2009/2010.’ European Journal of Population, Vol. 21, Nos. 2–3, LaFree, Gary, and Kriss Drass. 2002. ‘Counting Crime Booms Hume, Mo. 2008. ‘The Myths of Violence: Gender, Conflict, and among Nations: Evidence for Homicide Victimization Community in El Salvador.’ Latin American Perspectives, Rates.’ Criminology, Vol. 40, No. 4, pp. 769–800. Vol. 35. No. 5, pp. 59–76. Lennox, Patrick. 2008. Contemporary Piracy off the Horn of IBC (Iraq Body Count). n.d. ‘Iraq Body Count.’ Website. Africa. Calgary: Canadian Defense and Foreign Affairs Institute. ICESI (Instituto ciudadano de estudios sobre la inseguridad— Malaquias, Assis. 2010. ‘UNITA’s Insurgency Lifecycle in Angola.’ Citizen Institute for Studies on Insecurity). n.d. ‘Estadísti- In Klejda Mulaj, ed. Violent Non-State Armed Actors in cas oficiales.’ Mexico. n.d.a. ‘Base de datos de fallecimientos ocurridos por IISS (International Institute for Strategic Studies). 2009. The presunta rivalidad delincuencial.’ Mexico City: Presiden- Military Balance 2009: The Annual Assessment of Global cia de la República. Accessed 23 June 2011. Riedel, Marc and Wayne Welsh. 2008. —. 2007. ‘Systems of Violence andtheirPolitical Economy inPost- Richani, Nazih. 2002. Systems ofViolence: ThePolitical Economy Quinones, Sam. 2009. ‘State of War.’ Pettersson, Therése andLotta Themnér, eds.2010. PDH (Procurador delos Derechos Humanos). 2011. OECD (Organisation forEconomic Co-operation andDevelopment). Nitzschke, Heiko andKaysie Studdard. 2005.‘TheLegacies of Nagle, Luz.2010. ‘Corruption of Po Monkkonen, Eric.2001.‘New Standards forHistorical Homicide Miglierini, Julian.2011.‘Mexico Poet Javier Sicilia Leads Anger at Mexico. n.d.b. ‘Base dedatos dehomicidios presuntamente Münkler, 2003. Herfried. ‘The Wars of the21 Mukengere Mukwege, Denis andCathy Nangini.2009. ‘Rape Mueller, Susanne. 2008. ‘ThePolitical Economy of Kenya’s Moser, Caroline andDennis Rodg versity Press. Causes, andPrevention Conflict Situations.’ Unpublished paper. March. of New York Press. of War andPeace inColombia. New York: State University Research, Uppsala University. Conflict 2009. Uppsala: of Department Peace andConflict Derechos Humanos. Humanos enGuatemala Circunstanciado—Tomo I: Situación delos Derechos Enabling Development.Paris: OECD. 2009. Conflict andFragility—Armed Violence Reduction: No. 2,pp. 222–39. and Peacebuilding.’ War Economies: Challenges andOptions forPeacemaking pp. 95–122. Border.’ Small Wars andInsurgencies and theJudiciary inMexico andComplicity Across the tional Review of the Red Cross Research.’ uk/news/world-latin-america-13141263> Drug Violence.’ BBC News.22April. 2011. tute, SIPRIY Violence.’ In Stockholm Internation London: I.B. Tauria &Co. Ltd. tion: Managing Insecurity in the Aftermath ofPeace Accords Polity Press. in Europe from the Middle Ages to the Present Arms Survey.’ Liberia. LiberiaArmed Violence Assessment. Geneva: Small Armed Violence Assessment. Geneva: Small Arms Survey. and Guns.Cambridge: Cambridge University Press. Honour Killings Rise.’ Guardian. 30November. Rape andOther Violence: Leave My Body Out of War. Colombian Armed Conflict, 2001–2009 of Sexual Violence against Women in the Context of the Small Arms Survey. Costs, andInterventions rary Security Policy, Vol. 30, No.2.August,pp. 301–17. Armed Groups: The Central American Case.’ tion Secretariat. and Geneva: Editorial Panamericana andGeneva Declara- cijada: Panorama deuna violencia transformada pp. 123–44. Small Wars andInsurgencies Tactical Forms of Cartel Violence and Their Underpinnings.’ pp. 37–60. International Security Vol. 7, No. 5,pp. 375–94. Decline inCrime Rates.’ 20100802_mexico_security_memo_aug_2_2010> 20 December. atfor.com/an Guatemala enlaencru- al Peace Research Insti- . Bogotá: Campaign can_drug_cartels_ Contempo- ations: sassinations: alysis/ . Cambridge: . Bogotá Liberia Liberia . ,

41 2 3 4 5 1 A UNIFIED APPROACH 42 UCDP (Uppsala Conflict Data Program). 2011. ‘UCDP Actor Data- Wennmann, Achim. 2007. ‘The Political Economy of Conflict set v. 2.1-2010.’ Uppsala: Uppsala University. Financing: A Comprehensive Approach Beyond Natural WHO (World Health Organization). 2002. World Report on Vio- —. n.d. ‘UCDP Data.’ lence and Health. Geneva: WHO. —. 2011. ‘Deaths Estimates for 2008 by Cause for WHO Member UN (United Nations). 2010. Second Generation Disarmament, States: By Sex.’ Accessed June 2011. Demobilization, and Reintegration (DDR) Practices in Peace on Challenges and Opportunities for UN Peacekeeping. Wilkinson, Tracy. 2009. ‘Police Officer, Ten Relatives Killed in New York: Department of Peacekeeping Operations, Attacks in Tabasco.’ Los Angeles Times. 16 February. United Nations. Human Rights Report: 1 January–30 June 2008. ARMED VIOLENCE 2011 VIOLENCE ARMED —. 2010. ‘Mexican State Security Minister Can’t Trust Her Own

of Branch). n.d. ‘Data Set of the 1st and 2nd United Nations Wood, Elisabeth Jean. 2003. Insurgent Collective Action and Civil World Crime Surveys.’ Vienna: United Nations Office in War in El Salvador. New York: Cambridge University Press. Vienna. —. 2006. ‘Variation of Sexual Violence During War.’ Politics & UNDP (United Nations Development Programme). 2007. Informe Society, Vol. 34, No. 3, pp. 307–41. estadístico sobre violencia en Guatemala. Guatemala World Bank. 2011. World Development Report 2011: Conflict, City: UNDP. GLOBAL BURDEN BURDEN GLOBAL Security, and Development. Washington, DC: World Bank. UNGA (United Nations General Assembly). 2010. National Report —. n.d. ‘Indicators: GDP Growth (Annual %).’ Accessed on Submitted in Accordance with Paragraph 15 (a) of the 24 May 2011. Annex to Human Rights Council Resolution 5/1: Kenya. A/HRC/WG.6/8/KEN/1 of 22 February. Zimring, Franklin. 2007. The Great American Crime Decline. UNODC (United Nations Office on Drugs and Crime). 2010. The New York: Oxford University Press. Globalization of Crime: A Transnational Crime Threat Assessment. Vienna: UNODC. June. —. n.d.a. ‘UNODC Homicide Statistics.’ —. n.d.b. ‘United Nations Surveys on Crime Trends and the Opera- tions of Criminal Justice Systems (CTS).’ USDoD (United States Department of Defense). 2010. The Chal- lenge and the Promise: Strengthening the Force, Preventing Suicide and Saving Lives. Final Report of the Department of Defense Task Force on the Prevention of Suicide by Members of the Armed Forces. Washington, DC: USDoD. USDoS (United States Department of State). 2007. Iraq. Washington, DC: Bureau of Democracy, Human Rights and Labor, USDoS. Washington Post. 2011. ‘Debate Mounts over Death Toll in Mexico’s Drug War, as Government Refuses to Release Figures.’ 2 June.