ARMED VIOLENCE Issue Brief Number 4 September 2014
Handgun Ownership and Armed Violence in the Western Balkans
Introduction by Gallup Europe in the countries and The Western Balkans is home to an territories of the Western Balkans in 2012. estimated 3.6–6.2 million registered In the 1990s, the countries and territo- The Small Arms Survey had the oppor- and unregistered firearms. ries of the Western Balkans1 experienced tunity to insert three questions relating At least 500,000 and up to 1.6 several transformations: a transition to firearms possession and armed vic- million households own firearms from socialism to liberal democracy, timization into the 2012 Gallup Balkan in the Western Balkans. widespread economic decline, and Monitor (see Box 1). Where necessary, Since 1995, the average homicide episodes of violent conflict in Bosnia the data from the survey has been rate in the region has decreased and Croatia (1991–95) and in Kosovo supplemented by information from drastically, stabilizing at around (1999) (see Map). Since the turn of the 21st century, the region has witnessed other sources, such as international 2.0 per 100,000 between 2007 and increasing political stability and socio- and national data, special reports, and 2010. Nevertheless, the homicide economic adjustment. Yet while the policy and academic research. The main rate is still significantly higher than threat of armed conflict in the region has findings are as follows: in other European regions, and decreased, levels of handgun owner- ship and armed violence remain high. Throughout the region, which has an overall population of about Map Western Balkans 25 million (UNDESA, n.d.), an estimated 3.6–6.2 million firearms are in civilian possession. The high prevalence of HUNGARY SLOVENIA civilian-held firearms has been linked Zagreb to the rate of violent crime, with the ROMANIA homicide rate in the Western Balkans CROATIA being higher than in the other coun- tries of Southern Europe as well as in Belgrade Western Europe (Alvazzi del Frate and BOSNIA AND Mugellini, 2012; Geneva Declaration HERZEGOVINA SERBIA Secretariat, 2011, p. 60). In addition, the high prevalence of firearms and Sarajevo violent crime in the region is linked to the activities of organized crime, which MONTENEGRO is largely perceived by both the inter- Adriatic Pristina national and the local population as Sea Podgoricodgorica Kosovo one of the primary sources of insecu- Skopje rity in the Western Balkans. These findings call for an analysis Tirana MACEDONIA of the dynamics of firearms possession ITALY and armed violence in the Western ALBANIA Balkans. This Issue Brief examines the historical aspects of firearms prolifera- GREECE tion in the region in order to frame the issue. It also presents the results of a 0km 100 nationwide household survey conducted
http://www.smallarmssurvey.org 1 homicides are more frequently based his most institutionalized form handguns; the same view was expressed committed with firearms. of social banditry on the Balkan notion by 16 per cent of respondents in Albania About 1.2 per cent of all survey of the hajduk (Bracewell, 2003, p. 22). and 15 per cent in Serbia. Elsewhere in respondents reported that a house- Despite the emphasis on the ‘home- the region, however, the tendency to hold member was held at gunpoint grown’ nature of the bandit (in both invoke tradition as a reason for hand- in the 12 months prior to the admin- abovementioned senses of the term), gun ownership was much lower (below istration of the survey. historical analysis suggests that it was 5 per cent) (SEESAC, 2006a, p. 13). not solely the product of the local cul- This Issue Brief is divided into three ture, but that it also reflects the prom- sections. The first section outlines the Conflict-related spread of firearms inent state-building practices of the cultural and historical factors that have During the 1990s, the Western Balkans Ottoman Empire, which dominated facilitated the spread of firearms among saw not only the War of Yugoslav the region until the 19th century. Unlike the population. The second section Secession (1991–95),3 but also general in Western Europe—where the state examines the post-conflict security political instability in the region. eventually consolidated its monopoly dynamics, the role of organized crime Albania was experiencing an economic over the use of force by suppressing in the proliferation of firearms, and the and political crisis that peaked in 1997, other forms of armed violence (Tilly, prevalence of registered and unregis- 1991)—the Ottoman consolidation of following the collapse of its banking tered firearms in the region. The section state power and authority was achieved sector, and Macedonia was witnessing ends with a focus on longitudinal by making bargains and deals with growing ethnic animosity along its trends in the homicide rate, as disag- local armed bands (Barkey, 1994). It borders (Fischer, 2010; Irwin, 2010). gregated by sex and firearms. Section was only in the 18th and 19th centuries, Following the signing of the Dayton three, which comprises the bulk of the when the Ottoman Empire began to Peace Accord in 1995, both Bosnia analysis, presents region-wide house- weaken and national liberation strug- and Croatia faced precarious peace hold survey data obtained from the gles swept across the Western Balkans, processes and post-conflict reconstruc- 2012 Gallup Balkan Monitor to generate that the image of the bandit, with all tion. The Federal Republic of Yugoslavia, an evidence-based understanding of of its trappings—including the style of which comprised Serbia and Monte- armed violence in the Western Balkans. dress and an emphasis on armaments— negro, underwent dramatic political Specifically, the section unpacks issues took on cultural salience (Bracewell, changes as Serbia’s Socialist Party lost relating to self-reported and perceived 2003, p. 24). Moreover, the period of power in 2000 and the push for inde- levels of handgun ownership, pro- insecurity that accompanied the emer- pendence intensified in Montenegro. vides estimates of household firearms gence of nation-states in the region The end of the 1990s also saw a more possession, and reflects on experiences led not only to the proliferation of assertive drive for independence in of armed violence. bandit groups, but also to the view that Kosovo and an increasing opposition weapons can guarantee personal, fam- to it from the Serbian government. Framing the small arms ily, and community security (SEESAC, This political crisis, in turn, led to a 2006a, pp. 4–7). 78-day NATO air campaign against problem in the Western While the notion of hajduk may thus the Serbian forces and government Balkans refer to an important cultural phenom- that paved the way for the emergence enon, it also reflects Ottoman power- of Kosovo as a UN protectorate under Cultural heritage of the hajduk sharing with local elites as well as a Security Council Resolution 1244 A common assumption regarding fire- certain degree of tolerance of local (ICG, 1999). arms in the Western Balkans is that the ‘bandits’ who posed no threat to That all of these transformations had region has a deep-rooted ‘gun culture’ Ottoman central power or colonial an impact on the levels of firearms that predisposes its populations not rule. While the socio-cultural image throughout the Western Balkans— only to carry firearms, but also to use of the bandit might shed light on the especially in terms of the large quanti- them. Yet there has been very little symbolic meaning of owning a gun ties of weapons outside of governmental analysis of this ‘gun culture’ in the and even condone its use in certain control—has been acknowledged by region.2 In academic literature, the use situations, it is not clear to what extent both academics and security analysts. of guns in the Western Balkans is often they actually shape the proliferation One of the primary reasons for the high traced to the socio-cultural category of firearms. Indeed, as argued in a numbers of firearms in the region has of the bandit—locally referred to by recent study, ‘culture’ and ‘tradition’ to do with the fact that before 1991, the terms such as hajduk, haidut, uskok, and are not primary reasons for gun own- Socialist Federal Republic of Yugoslavia klepht—which carries ‘connotations of ership in the region (SEESAC, 2006a). (SFRY)4 had the fourth-largest army— oppression, thirst for liberty, and heroic There are a few exceptions, however. the Yugoslav People’s Army—in Europe masculinity’ as well as ‘lawlessness, In Montenegro, about 22 per cent of and a matching military industrial primitivism, and violence’ (Bracewell, survey respondents stated that ‘tradi- complex (Anastasijevic, 2006, p. 10). 2003, p. 22). It has been argued that the tion’ was the reason why individuals The Yugoslav People’s Army was historian Eric Hobsbawm (1971) even within their neighbourhoods owned made up of two elements: the regular
2 Small Arms Survey Issue Brief Number 4 September 2014 ground forces, controlled by the federal 1944 until 1985, under the communist arms traded on the international weap- government in Belgrade, and the ter- regime of Enver Hoxha, a great deal ons market, precursors (ephedrine) ritorial defence force (Gow, 2003, p. 52).5 of emphasis was placed on creating a and synthetic drugs’ (EUROPOL, The territorial defence units were strong and well equipped military 2013, p. 12). This remains the case especially important for guarding the (Arsovska and Kostakos, 2008, p. 362). even though all UN Member States in country’s large stockpiles. When the This practice ensured that even after the Western Balkans have signed the wars in Croatia and later in Bosnia broke the fall of communism, large stock- United Nations Convention against out, these stockpiles were increasingly piles of weapons existed in the country. Transnational Organized Crime and placed under the control of the Yugo- Following the collapse of the pyramid the Protocol against the Illicit Manu- slav People’s Army and, by extension, schemes and subsequent fall of the facturing and Trafficking in Firearms the Serbian government (Griffiths, Albanian government in 1997, large- (UN, 2001; 2003). 2010, pp. 183–87).6 Some of these scale looting of military stockpiles Although the international com- weapons were allegedly diverted to took place. In fact, according to some munity has devoted attention to the paramilitary-cum-criminal groups estimates, up to 643,220 small arms issue of organized crime in the West- that were backed by the various local and light weapons were pillaged; ern Balkans, very little systematic governments and that came to play a only about 15 per cent of these were analysis of the phenomenon exists key role in the conflict. Following the subsequently recovered (Khakee and (Woodward, 2004, p. 225). Similarly, 1995 Dayton Peace Accords, some Florquin, 2003, p. 25). It is assumed while links between Balkan organized weapons trickled into the hands of that the local population was respon- crime and arms trafficking are often civilians and others found their way sible for much of the looting, which made explicit in the grey literature to organized crime groups, whose resulted in the widespread diffusion (Anastasijevic, 2008), evidence is scarce importance in the region was grow- of illicit weapons and ammunition. and fragmentary, especially if the state ing steadily (SEESAC, 2007, p. 3; Nevertheless, a significant portion— is implicated in criminal activities. Milosevska, 2009, pp. 6–7). about 150,000 firearms—is thought to What little information exists is mostly Another reason for the prolifera- have been smuggled across the border drawn from journalistic accounts and tion of firearms in the region was the into Kosovo and sold to various rebel official discourse, both of which portray result of smuggling channels operating groups, such as the Kosovo Liberation organized crime as having a hierarchical in Croatia and Bosnia during the wars. Army, the Albanian National Army, structure, comprising members who The disintegration of Yugoslavia and and the National Liberation Army tend to come from a single ethnic group the outbreak of war led to the imposi- (NLA) (Arsovska and Kostakos, 2008, (often referred to as the Albanian, tion of international sanctions and an p. 362; Khakee and Florquin, 2003, p. 25). Montenegrin, and Serbian mafias), arms embargo on all the republics of and monopolizing not only traditional the SFRY. These measures had little forms of criminal activity (such as effect on Serbia and Montenegro, which, Post-conflict security trafficking in drugs, arms, and human as mentioned above, had inherited dynamics beings), but also white-collar types of the Yugoslav People’s Army and its Arms trafficking and the spread crime (such as money laundering and stockpiles. Croatia and Bosnia, how- of organized crime in the Western bank and investment fraud). There is, ever, were hit hard by the embargo however, very little empirical support and resorted to smuggling of arms in Balkans for these assumptions. order to build up their armed forces. Organized crime in the Western Balkans What little evidence exists seems Although it is difficult to estimate the did not emerge as a by-product of the to suggest that organized crime in the precise quantity and value of weapons political and economic turmoil that Western Balkans, as in many other parts funnelled into Croatia during the characterized the region during the of Europe, is neither a single, large period, it is believed that between 1990s, as some analysts have argued;9 entity, nor is it cohesive, homogeneous, 1993 and 1995 about USD 308 million rather, it predated that period. However, hierarchical, or monopolizing (Paoli and worth of weapons were smuggled the outbreak of war in Bosnia and Fijnaut, 2004, p. 608; von Lampe, 2008). into the country (Hajdinjak, 2002, Croatia, the imposition of an arms Rather, it is small and fragmented— pp. 9–10).7 Similarly, during the same embargo and comprehensive economic characterized by loose horizontal period about USD 270 million worth of sanctions, and political changes in networks and ethnic heterogeneity weapons was smuggled into Bosnia— Albania, Kosovo, and Serbia at the turn (Antonopolus, 2008, p. 320; Carapic, with the last year of the conflict seeing of the century did provide opportuni- 2014; Mladenovic, 2012).10 Thus, despite the value of illicit arms imports in the ties for organized crime to proliferate. the fact that organized crime leaders war-torn country reaching $800 million The socio-political environment of the and members tend to present them- (pp. 10–11).8 past 14 years has done little to curb selves as ultra-nationalistic—portraying Another source of weapons in criminal activities. As a result, ‘the an image of a unified ethnic under- the region was the breakdown of the Western Balkans are not only a transit world within specific countries—they Albanian government in 1997. From region, but also a major source of fire- are actually guided by pragmatism
http://www.smallarmssurvey.org 3 and often collaborate with different of employment. The second reason for former Yugoslavia (Hajdinjak, 2002, ethnic groups that are based else- the outflow and proliferation of crimi- pp. 16–18). Indeed, partially in response where in the region (Carapic, 2014; nal groups from the Western Balkans, to the events taking place outside Mladenovic, 2012). As one analyst and from the former Yugoslavia in Macedonia’s borders, organized crime observes, this ethnic heterogeneity particular, can be attributed to the groups mushroomed in the country; reveals that ‘Yugoslavia has never state’s desire to extend its political at the same time, they stepped up their ceased to exist for the organized crime influence outside its borders—in terms involvement in various illegal activi- world in the Balkans’ (Mladenovic, of targeting so-called ‘political enemies’ ties, such as the smuggling of arms, 2012). This is most evident when it and by involving itself in illegal and petrol, and cigarettes to Serbia and comes to the smuggling of arms (espe- illicit markets (Anastasijevic, 2008; Montenegro, and trafficking in drugs cially prior to 1995) and trafficking Carapic, 2014; CSD, 2004). At this (CSD, 2004, pp. 81–82). The imposition in drugs, activities that depend on stage of its development, organized of sanctions also had an effect on the the close collaboration and overlap crime in the Western Balkans can thus spread of organized crime in Albania. between organized crime groups best be described as ‘foreign-based’, During the first few years of the 1990s, (Anastasijevic, 2006). in the sense that the groups commit organized crime in Albania was the Contemporary forms of organized crimes outside of their place of origin domain of relatively small groups of crime largely took place during the (von Lampe, 2005). related individuals involved in the 1970s and 1980s, when communist When war broke out in Bosnia and smuggling of cigarettes and petrol— rule still predominated in the region Croatia in the early 1990s, these foreign- especially to Montenegro. By 1997, (Carapic, 2014).11 Unlike elsewhere in based groups returned to their countries when pyramid schemes crashed, these the world, where organized crime has of origin in order to take advantage of groups had transformed, not only filled a political vacuum created by the new profit-making opportunities becoming more organized and cohe- state weakness or collapse, the forma- and to help (formally or informally) sive, but also taking on a prominent tion and evolution of organized crime with the war effort, thus effectively role in society and acting as a threat in the Western Balkans was closely tied transforming into paramilitary-cum- to the state (CSD, 2004, pp. 88–89; to the state. In the former Yugoslavia, criminal groups. These groups were Hysi, 2004, p. 542). for instance, organized crime partly especially important when it came to Given its strategic location, Kosovo formed as a result of the ‘open border the smuggling of weapons—whose has for decades been one of the major policy’ promoted by the then socialist availability, as noted above, varied distribution centres of narcotics enter- regime, which enabled a portion of among the warring sides. The avail- ing Europe via the Balkan route. Money the population to emigrate and seek ability of weapons affected not only obtained from the heroin trade appears employment in Western Europe. This the ability to wage war, but also the to have fuelled not only the organized policy had two broad side effects. On types of illegal markets in which organ- crime groups themselves, but also the the one hand, it led to an increase in ized crime groups became involved. emerging Kosovo state and its inde- the general standard of living in the For instance, in Croatia and Bosnia pendence movement (Hajdinjak, 2002, former Yugoslavia, as the diaspora organized crime groups were heavily p. 12; Judah, 2000, p. 321). This was wired remittances to the local popula- involved in the smuggling of arms into especially the case with the Kosovo tion and the pressure on the domestic their respective countries in order to Liberation Army, which reportedly Yugoslav labour market dropped due aid the state in building up its war- evolved from the Jashari clan, itself to the outflow of unemployed individu- fighting capacity; some of these weap- implicated in the drug trade and alleg- als to Western Europe (Carapic, 2014). ons were provided by their Serbian edly the beneficiary of a significant On the other hand, the policy led to counterparts (Hajdinjak, 2002). Given amount of weapons that were looted a so-called ‘crime export’ of individual the abundance of weapons under the from Albania in 1997 (Judah, 2000, lawbreakers and small criminal groups control of Serbian forces, the primary pp. 135–69; Pugh, 2004, p. 55). These to Western Europe, where they later focus of organized crime groups in organized crime groups continued to developed into larger organized crime Serbia and Montenegro was not arms collaborate throughout the 1990s. groups. This exodus can be further trafficking (although they did engage Yet by the turn of the century, sig- attributed to two interrelated processes. in it), but sanctions busting—especially nificant transformations in the nature First, in terms of financial opportunity, in terms of smuggling petrol, drugs, of organized crime had begun to take Western Europe was a much more cigarettes, foodstuffs, and other place across the region, with many attractive environment than the West- scarce goods. groups attempting to ‘legalize’ them- ern Balkans, where the society and These activities could not have selves (Albertini, 2011). On the one economy continued to be controlled by taken place without the proliferation hand, this was the result of repressive the state. In this sense, it was rational of organized crime in neighbouring police policies that tried to restrain for criminal groups to take advantage Macedonia, one of the most active vio- the spread of organized crime. On the of the open border policy and emi- lators of the arms embargo and sanc- other hand, however, it was the result grate with other individuals in search tions imposed on the republics of the of the political and economic transfor-
4 Small Arms Survey Issue Brief Number 4 September 2014 mations taking place in the region. an effect on the levels of insecurity and Serbia, 94 per cent of whom are men The desire to enter the European Union violence in society, especially with (Small Arms Survey, 2012b). and the liberation of markets in the respect to trafficking in narcotics The same survey found that there countries of the Western Balkans cre- (Geneva Declaration Secretariat, 2011, were about 100,000 registered firearms ated incentives for political leaders to p. 4). It is perhaps not surprising that in Montenegro. As in Macedonia and tackle corruption and the influence of a recent survey discovered that about Serbia, the majority of firearms in organized crime; meanwhile, the eco- 37 per cent of the population in the Montenegro were handguns and bolt- nomic opportunities that came along Western Balkans saw organized crime action firearms. Moreover, there are with increased European integration as affecting their lives on a daily basis.12 about 80,000 licensed firearms owners allowed organized crime groups in in Montenegro, most of whom are men the region to buy up or open legal Availability of firearms and levels and only about 500 of whom are state businesses. These ‘controversial busi- personnel. In addition, an estimated nessmen’—as many organized crime of violence in the Western Balkans 40,000–80,000 unregistered firearms leaders are referred to locally (Carapic, The Western Balkans is home to between are circulating in Montenegro (Small 2014)—are embedded in mainstream 3.6 and 6.2 million registered and unreg- Arms Survey, 2012c). society and present themselves as law- istered firearms (see Table 1). There is One important, albeit under- abiding citizens, although they continue very little available data on the types of researched, area when it comes to to be involved in white-collar crime, firearms in circulation. Nevertheless, a the proliferation of firearms in the such as investment fraud and health recent survey of small arms in Europe, Western Balkans relates to legislation— insurance fraud (von Lampe, 2005). conducted by the Small Arms Survey, that is, the changes in the legal and This is not to say that organized provides insight into the different types law enforcement approach to gun crime groups involved in the business of firearms circulating in the region. ownership and use since the collapse of smuggling are no longer present For instance, of the 157,000 registered of the SFRY. In the former Yugoslavia, in the region. Rather—if the various firearms in Macedonia, the majority obtaining a permit to possess and carry assassinations that rocked Belgrade were handguns, revolvers, and bolt- small arms was relatively easy: any in mid-2012 and the arrest of Naser action rifles and carbines (Small Arms person of legal age without a criminal Kelmendi, an alleged Balkan drug king- Survey, 2012a). Similarly, a survey on record could obtain a gun permit (SFRY, pin, in May 2013 in Kosovo are any small arms undertaken in Serbia in 1967). This lax policy did not translate indication—organized crime continues 2012 shows that there were up to 1.2 into higher rates of homicide. Although to be present and active (Glavonjic, million registered civilian firearms in data is hard to obtain, it is believed 2012; OCCRP, 2013). A full understand- the country, 90 per cent of which were that up to 1990, the SFRY homicide ing of the levels of armed violence in handguns (revolvers and pistols) and rate was relatively low, compared to the Western Balkans thus calls for rec- bolt-action weapons (rifles and car- both the European and global rates at ognition of the presence of organized bines). In addition, there were about the time. Moreover, the rate of solved crime, which has been found to have 874,000 licensed firearms owners in or cleared homicides was around 90 per cent (Simeunovic-Patic, 2003, p. 34). Since the 1990s, and especially follow- Table 1 Estimated numbers of registered and unregistered firearms, per country ing the signing of the Dayton Peace Country or territory Registered firearms Unregistered firearms Total Accord, many of the countries of the Western Balkans have implemented Low High Low High Low High tougher laws for obtaining and pos- Albania 70,000 70,000 200,000 210,000 270,000 280,000 sessing firearms. Bosnia and 350,000 350,000 150,000 750,000 500,000 1,100,000 A detailed analysis of the laws is Herzegovina beyond the scope of this Issue Brief and has been provided elsewhere.13 Croatia 370,000 380,000 150,000 600,000 520,000 980,000 However, with the exception of Serbia Kosovo 30,000 70,000 260,000 260,000 290,000 330,000 and Croatia, most of the legislation in Macedonia 160,000 160,000 160,000 450,000 320,000 610,000 the region is still in its early stages and untested (Grillot, 2010, p. 162). That Montenegro 90,000 100,000 40,000 90,000 130,000 190,000 being said, a recent study conducted Serbia 1,100,000 1,190,000 500,000 1,500,000 1,600,000 2,690,000 in Montenegro considers the changes in patterns of homicide following fire- Total 2,170,000 2,320,000 1,460,000 3,860,000 3,630,000 6,180,000 arms legislation passed in 2007, which Note: The values are rounded to the nearest ten thousand. stipulates gun owners may carry fire- Sources: Albania: Holtom et al. (2005, pp. 45–48); SEESAC (2006a, p. 11); Bosnia and Herzegovina: Azinovic, Bassuener, and Weber (2011, p. 53); arms in public only if they have a licence SEESAC (2006a, p. 11); Croatia: SEESAC (2006a, p. 11; 2006b, pp. 3, 5); Kosovo: Republic of Kosovo (2009, p. 47); SEESAC (2006a, p. 11); Macedonia: Grillot et al. (2004, pp. 17–19); Small Arms Survey (2012a); Montenegro: B92 (2012); Florquin and O’Neill Stoneman (2004, p. 4); Small Arms Survey to that effect. While the study finds (2012c); Serbia: Glavonjic and Heil (2012); SEESAC (2006a, p. 11); Small Arms Survey (2012b) that the implementation of stricter
http://www.smallarmssurvey.org 5 firearms policies was followed by a Table 2 Homicides, by sex of the victim, 2006–10 decrease in the rate of firearms-related Country* 2006 2007 2008 2009 2010 homicide in Montenegro, it shows that the number of knife-related homicides M W M W M W M W M W increased (Radojevic and Cukic, 2012, Albania 77 17 72 14 63 21 47 12 60 8 pp. 229–30). The dearth of comparable studies and data from other countries Croatia 38 29 49 23 47 35 30 28 31 28 of the Western Balkans precludes an Bosnia and 265 72 271 77 282 65 255 79 232 60 assessment of the relationship between Herzegovina handgun legislation, gun ownership, Macedonia 42 11 36 12 38 8 20 9 31 13 and violence. Such comparisons can be instructive; a recent study of data Montenegro 9 3 4 1 12 3 10 4 n/a n/a from 50 US states, for instance, finds Serbia 106 48 127 73 103 39 94 46 88 42 that the states with more stringent gun laws have lower levels of gun-related Average 89.5 30.0 93.2 33.3 90.8 28.5 76.0 29.7 88.4 30.2
deaths (Fleegler et al., 2013). Notes: Variations in gun legislation can M=Men; W=Women. thus be taken as an entry point for * No sex-disaggregated data was available for Kosovo. country-by-country comparison of Source: Small Arms Survey (2013a) gun availability and types of violence, at both the national level and the level countries in the region, data collected the rate is still higher than in other of the individual (van Kesteren, 2014, by the United Nations Office on Drugs European regions. Homicide rates in pp. 54–55). Although this is the focus and Crime reveals a general down- Western Europe, for instance, are about of the next two sections, it is important ward trend in homicides since the end 1.0 per 100,000 while the average homi- to mention that despite the historical of the wars in Bosnia and Croatia in cide rate in Southern Europe (excluding experience of war and the high num- 1995, and following the political the countries of the Western Balkans) bers of firearms in the post-conflict changes in Albania and Serbia at the is 1.5 per 100,000 (Geneva Declaration period, the countries of the region are turn of the century. Although the exact Secretariat, 2011, p. 60). When the homi- not characterized by high levels of con- scope and timing of the decline varies cides are disaggregated by sex, it ventional crime or violence (UNODC, across the countries, it appears that the becomes clear that between 2006 and 2008, p. 9), which is most obvious when average homicide rate in the region 2009 men were on average three times homicide data, the most readily avail- began to decrease drastically in 1995, more likely than women to be victims able indicator of levels of crime and stabilizing to around 2.0 per 100,000 (see Table 2). This rate is much lower armed violence in a country,14 is exam- between 2007 and 2010 (see Figure 1). than the global average, which indi- ined and compared to the global average. Although the level of intentional cates that men are killed approximately While there is a general lack of homicides has decreased, to a level five times more frequently than women, reliable historical information on all significantly below the global average,15 but much higher than the average in Western Europe, where the ratio approaches 1:1 (Geneva Declaration Figure 1 Trends in intentional homicides in the Western Balkans Secretariat, 2011, pp. 114, 119). Average Albania Croatia Bosnia and Herzegovina Kosovo Macedonia Montenegro Serbia Moreover, in the Western Balkan AVERAGE RATES PER 100,000 PERSONS RATES PER 100,000 PERSONS region an average of 44 per cent of 50 10 homicides are conducted with fire- arms, which is much higher than in
20 8 Southern and Western Europe, where the rates are 30.5 per cent and 35.1 per cent, respectively. The authors 15 6 conclude that the availability of guns in the region constitutes one factor in 10 4 understanding the level of homicide rates (Alvazzi del Frate and Mugellini, 5 2 2012, pp. 149–50). To assess how the availability of firearms affects homi- 0 0 cide rates, it is important to unpack 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 its interaction with other factors, such Note: The homicide rate per 100,000 persons in Albania for the year 1997 is 49.9. The methodology used to produce this graph is drawn from Alvazzi del Frate and Mugellini (2012, p. 145). as levels of firearms ownership, per-
Sources: Albania, Croatia, Macedonia, Montenegro, and Serbia (2000–01): UNODC (n.d.); Bosnia and Herzegovina: Bosnia and Herzegovina (2011, p. 56); ceptions of firearms in society, and Kosovo: KAS (n.d.); Serbia (2003–11): SORS (2012, p. 79) victimization rates.
6 Small Arms Survey Issue Brief Number 4 September 2014 Box 1 Data on firearms possession and armed victimization own a handgun?’—which represents self-reported or direct ownership This analysis utilizes interview data on handgun ownership and victimization in the Western Balkans, data—it can be deduced that approxi- as obtained from Gallup Europe and collected in 2012 by the Gallup Balkan Monitor. The Balkan Monitor, mately 7.55 per cent of households in a nationwide annual survey, was used because it offers the advantage of three-stage probability sampling, the Western Balkans owned at least one ensuring higher levels of representation by conducting face-to-face interviews in the respondent’s home, firearm. On the basis of the responses stratifying the data according to geo-economic regions and types of settlements (urban vs. rural), and to the question, ‘How widespread gathering information on non-institutionalized adults (15 years of age and older) living in the countries of the Western Balkans. The nationwide surveys were conducted in Albania, Bosnia and Herzegovina, Croatia, would you say handgun ownership Kosovo, Macedonia, Montenegro, and Serbia.16 The Small Arms Survey inserted three questions into the is in the town or area where you live? 2012 Balkan Monitor survey, two relating to issues of handgun ownership (direct and perceived) and one How many households have hand- relating to direct experiences of armed violence. guns?’—which is taken as an indirect The examination of handgun ownership and armed violence relies on the data provided by 7,106 respondents measure of ownership—it can be esti- (see Table 3). The data obtained from the Balkan Monitor is a subset of the original questionnaire—including mated that the proportion of house- the responses to the three questions provided by the Survey and relevant data required for their analysis. holds that own handguns is nearly three The three questions appeared in the section on ‘political activity in the region’, which included questions on times as great, at about 23.1 per cent. political events both in the respondent’s country and in neighbouring countries. Table 4 shows the results of both the Table 3 Breakdown of respondents by country or territory (n=7,106) direct question on self-reported owner- ship and perceptions of ownership in Country or territory Number of valid responses Response rate (%) one’s neighbourhood. Accordingly, Albania 1,029 72.0 the number of households that own Bosnia and Herzegovina 1,005 68.7 firearms in the Western Balkans can be estimated to be between about Croatia 1,000 65.9 530,000 and 1.62 million. This range Kosovo 1,024 74.0 differs markedly from the abovemen- tioned estimates for registered and Macedonia 1,025 66.4 unregistered firearms (3.6 to 6.2 million; Montenegro 1,000 52.1 see Table 1), calling into question the Serbia 1,023 64.4 widely held assumption that every household in the region owns a gun In terms of handgun ownership, the Small Arms Survey suggested a combination of two questions to assess (SEESAC, 2007, p. 3). how many firearms are in civilian hands. One deals with direct household ownership of firearms and the The discrepancy may be partly other looks at perceptions of firearms in the community. The exact wording of the question relating to explained by the fact that the figures direct handgun ownership was: Do you, or does anyone else in your household, own a handgun? Possible responses were: 1. Yes; 2. No; 8. Don’t know; or 9. Refuse to answer. The exact wording of the question presented in Table 4 do not take into relating to the perception of firearms in society was: How widespread would you say handgun ownership account the possibility that one house- is in the town or area where you live? How many households have handguns? The possible responses were: hold can possess more than one gun. 1. Very few; 2. Quite a few; 3. The majority; 4. Most/almost all; 8. Don’t know; or 9. Refuse to answer. Likewise, they do not account for the The Small Arms Survey’s third question concerned direct experiences of armed violence. The advantage various firearms that could be in official of this question, as opposed to classic follow-up questions regarding victimization experiences (such as state stockpiles, or that may be unreg- ‘If you have been a victim of crime. . . was a gun used?’), is that it is asked of the entire sample and not istered or illegally possessed by ordi- just the subset of individuals who have been victimized. The exact wording of the question as it appeared nary citizens or organized crime groups. in the questionnaire was: Have you personally, or has anyone in your household, been held at gunpoint during In this context, it should be noted that the past 12 months? The possible responses were: 1. Yes; 2. No; 8. Don’t know; or 9. Refuse to answer. The ‘don’t know’ and ‘refuse to answer’ replies were withdrawn from the computations for questions with missing a significant correlation has been shown values, so that only valid responses remained. to exist between higher levels of gun Results in this Issue Brief are based on data that has been weighted to make the samples as representa- ownership and higher firearm homi- tive as possible of national populations aged 15 or older. The weight variable takes into account sex, cide rates, as well as an increased geo-economic region, urban vs. rural population distribution, age, and education (primary/secondary/ risk of both suicide and homicide of higher education) variables. women (Bailey et al., 1997; Killias, Van Kesteren, and Rindlisbacher, 2001; Siegel, Ross, and King, 2013). An analysis of handgun ownership holds that own handguns. Despite the Estimates of household and armed victimization in the West- fact that data on this issue is often not ern Balkans is provided below. firearms ownership recorded or not available, it is still The distribution of gun availability possible to estimate household firearms in the region is key to understanding possession by using data obtained Self-reported handgun ownership the dynamics of armed violence more from national surveys. On the basis of When asked the question ‘Do you, or generally. A good indicator of the dis- responses to the question, ‘Do you, or does anyone else in your household, tribution is the percentage of house- does anyone else in your household, own a handgun?’, about 6.8 per cent
http://www.smallarmssurvey.org 7 Table 4 Estimates of household firearms possession across the Western Balkans
Country or territory Total households Direct (self-reported) Indirect (opinion about how Average in population many households own firearms in the area)
% of Low estimate % of High estimate Average Average households number households number number
Albania 722,600 6.50% 46,969 17.20% 124,287 11.90% 85,628
Bosnia and Herzegovina 1,207,098 4.60% 55,527 31.40% 379,029 18.00% 217,278
Croatia 1,535,635 4.30% 66,032 23.30% 357,803 13.80% 211,918
Kosovo 294,886 2.90% 8552 21.70% 63,990 12.30% 36,271
Macedonia 564,296 5.70% 32,165 25.10% 141,638 15.40% 86,902
Montenegro 194,795 11.80% 22,986 27.80% 54,153 19.80% 38,569
Serbia 2,497,187 11.90% 297,165 20.00% 499,437 16.00% 398,301
Total 7,016,497 7.55% 529,396 23.10% 1,620,337 15.32% 1,074,867
of all respondents attested to owning Shkoder Lake came in second and third, such low levels of handgun ownership a firearm. This is nearly 2 percentage respectively. Kosovo also exhibited also lends credence to this argument, points higher than the proportion of regional variation, with individuals especially since owning a gun in the individuals reporting handgun owner- from Serbian north and south-west area might be seen as a politically ship in the European Union (European Kosovo reporting the highest rate of sensitive issue. Finally, in Macedonia, Commission, 2013, p. 6). However, ownership (5.6%), followed by Pristina, respondents from the south-western country-level results show that this which accounted for 4.2 per cent of part of the country were significantly difference is even greater in some cases. reports. Moreover, in both Albania more likely to report that they or some- Serbia and Montenegro have by far the and Kosovo, residents of small towns one in their household had a gun (12.2 highest self-reported ownership rate: or villages were more likely to report per cent). Moreover, there was also a 11.9 and 11.8 per cent of household that they or someone in their house- difference between Albanian Macedonia respondents confirming that they own hold owned a handgun, compared to and north-western Macedonia, with handguns, respectively. Indeed, the respondents living in more urban set- the former reporting a rate of 1.2 per reported number of handgun owners in tlements. These findings echo study cent and the latter 4.8 per cent. both countries was about double that results showing that handgun owner- In terms of respondent characteris- of Albania (6.5%) and Macedonia (5.7%), ship is much lower in larger cities— tics, men in the region were 1.5 times triple the ownership rate of Croatia where the presence of state security (p<0.001) more likely than women to (4.3%) and Bosnia and Herzegovina forces is stronger—than in rural areas, report that they or someone in their (4.6%), and nearly four times higher where the absence of a state presence household owned a handgun. Croatia than in Kosovo (2.9%) (see Table 4). fosters the provision of security exhibited the greatest gendered differ- Despite the ongoing political insecu- through private means (Kleck and ence, with men 2.4 times (p<0.01) more rity in Kosovo, respondents were Kovandzic, 2009). likely to report that they or someone least likely to report that they or some- In Bosnia the highest number of in their household owned a gun than one within their household owned a self-reported firearms was present in women. Similarly, in Montenegro men handgun. Indeed, only 30 out of 1,023 the Cazin region, whereas residents were 1.7 times (p<0.01) more likely respondents from Kosovo—or 2.9 per of the southern Republika Srpska to report handgun ownership than cent—answered the question on hand- reported the lowest number of hand- women; in Serbia, they were 1.8 times gun ownership in the affirmative. guns. The most striking aspect of this (p<0.01) more likely to do so. Although While no regional differences finding is that the Cazin region is not statistically significant, differences emerged in terms of patterns of firearms famous for having been one of the also emerged in the other countries of ownership in Croatia, Montenegro, or most contested areas during the the Western Balkans. Interestingly, in Serbia, such variations were apparent Bosnian war, suggesting that motiva- Albania women were more likely than elsewhere. In Albania, for instance, the tions for handgun ownership may still men to report firearms ownership; the central region accounted for almost 61 be related to the country’s recent his- opposite was the case in Bosnia and per cent of all reported cases of hand- tory. Although it might not be obvious, Herzegovina, Kosovo, and Macedonia. gun ownership, while the southern the fact that respondents from the Figure 2 presents the responses to part of the country and the area around southern Republika Srpska reported the question, ‘Do you, or does anyone
8 Small Arms Survey Issue Brief Number 4 September 2014 Figure 2 Self-reported firearms ownership, by sex shows that respondents who were in Yes No Don't know a domestic partnership provided reli- PERCENTAGE OF RESPONDENTS ably different responses from all the 100 other respondents (p<0.000); mean- while, single respondents only differed from individuals who described them- 80 selves as widowed or in a domestic partnership (p<0.05). An analysis that 60 considers both marital status and sex (p< 0.05) reveals that men in domestic 40 partnerships reported higher levels of handgun ownership than all the other 20 respondents (p<0.05).
0 Perception of firearms in society Male Female Male Female Male Female Male Female Male Female Male Female Male Female ALBANIA BOSNIA AND CROATIA KOSOVO MACEDONIA MONTENEGRO SERBIA When asked, ‘How widespread would HERZEGOVINA you say handgun ownership is in the Note: The number of ‘yes’ responses for Kosovo was too small to make the difference between men (n=19) and women (n=12) statistically meaningful. town or area where you live? How Source: Small Arms Survey (2012d) many households have handguns?’, numerous respondents—namely one- else in your household, own a hand- ownership (Hill, Howell, and Driver, third to one-half of all survey partici- gun?’ The abovementioned association 1985, p. 549; Kleck et al., 2011). pants—replied with ‘don’t know’ or between sex and firearms possession Figure 3 reveals the distribution of ‘refuse to answer’ (see Figure 4). This is visible for Croatia, Montenegro, firearms ownership by marital status may be a reflection of the politically and Serbia. Yet this finding has to be of respondents. The majority of indi- sensitive nature of the subject of interpreted with caution, as these viduals who reported gun ownership handgun ownership in the Western samples may not be truly representa- were married (55.7 per cent). Together Balkans. In line with the reported level tive due to various constraints. For with respondents who had previously of handgun ownership, nearly half instance, women are probably over- been married (6.7 per cent) or were in the respondents (n=3,521 or 49.5 per represented in the sample because a domestic partnership (3.3 per cent), cent) stated that ‘very few’ or ‘quite a they were more likely to be at home married respondents comprise about few’ households within their commu- during the day, when the interviews two-thirds (65.7 per cent) of all indi- nities had handguns. Throughout the were conducted. Conversely, women viduals who reported gun ownership. region, a combined total of 17.8 per could be under-represented due to About one-third (33.4 per cent) of the cent of all respondents stated that ‘the cultural constraints, which may have respondents described themselves as majority’ or ‘almost all’ households in prevented male interviewers from single. Divorced or separated individ- their area had handguns. speaking freely to women (Pavesi, 2013). uals were the least likely to report that Apart from Albania and Croatia, a In all countries under review except they owned a handgun (less than 1 per respondent’s ethnic or religious group Albania, men were more likely to report cent). The different categories of marital was significantly correlated with the handgun ownership. These results may status appear to affect handgun own- perception of handgun ownership in be attributed to long-standing cultural ership in different ways (p< 0.000). one’s neighbourhood or area. In Bosnia ties between guns and masculinity as A disaggregation of these categories and Herzegovina, for instance, indi- well as the fact that men accounted viduals who identified themselves as for the vast majority of combatants Figure 3 Percentage of handgun ownership by Croat or Muslim were more likely to during the periods of conflict and that marital status (n=479) assert that ‘quite a few’ or ‘the majority’ they are more involved in informal of households had guns, compared to economic activities than women. In respondents who identified themselves this sense, one possible explanation as Serb. for the continued prevalence of fire- Reflecting the overall demographic arms in the region could be the direct makeup of Kosovo, the majority of experience of war by men and the respondents identified themselves as continued insecurity they face in the Albanian (90 per cent), followed by post-conflict environment. In any Serb and Muslim (about 7 and 2 per event, the findings presented here are cent, respectively). Of the ethnic Married (55.7%) Single/Never been married (33.4%) in agreement with studies that point Widowed (6.7%) Domestic partner (3.3%) Albanians, more than half (55 per to a relationship between male victimi- Divorced (0.6%) Separated (0.2%) cent) described their neighbourhoods zation (or a fear thereof) and handgun Source: Small Arms Survey (2012d) as having ‘very few’ or ‘quite a few’
http://www.smallarmssurvey.org 9 Figure 4 Perception of handgun ownership, country or territory tion. The positive correlation identified above—which shows men to be 1.5 Very few Quite a few The majority Most/almost all Don’t know Refuse to answer PERCENTAGE OF RESPONDENTS times more likely than women to report 100 firearms ownership—can be further examined by taking into account the indirect measure of firearms ownership, 80 that is, perceptions relating to the pres- ence of handguns in one’s neighbour- 60 hood. With the exception of Croatia, the indirect measure of firearms owner- 40 ship was significantly associated with sex throughout the region. For each 20 country, responses to the question, ‘How widespread would you say handgun
0 ownership is in the town or area where ALBANIA BOSNIA AND CROATIA KOSOVO MACEDONIA MONTENEGRO SERBIA you live?’ were disaggregated by sex. HERZEGOVINA The high incidence of ‘refuse to answer’ Source: Small Arms Survey (2012d) and ‘don’t know’ by both women and men is noteworthy; together, these guns, while 16 per cent reported that described their neighbourhoods as responses account for between one- ‘the majority’ held firearms. Almost having either ‘very few’ or ‘quite a third and nearly half of the total. When one-quarter of all ethnic Albanians few’ guns. these responses are disaggregated by refused to answer or said that they Reflecting the overall demographic sex, they reveal that women replied ‘did not know’. Conversely, nearly 60 makeup of Montenegro, the majority with ‘don’t know’ more often than per cent of the Kosovo Serbs (n=72) of respondents identified themselves men. In Bosnia, Kosovo, Macedonia, were reluctant to answer the question as Montenegrin (about 52 per cent), Montenegro, and Serbia, men were relating to the perception of firearms; followed by those who called them- more likely to refuse to answer the the majority of those who did reply selves Serb (about 39 per cent), and indirect question relating to firearms stated that there were ‘very few’ hand- Muslim (about 7 per cent). Of those who ownership compared to women, sug- guns in their neighbourhoods. These identified themselves as Montenegrin, gesting that this is a sensitive issue in findings reflect the political insecurity about 35 per cent responded that they these countries and territories. still felt in Kosovo and the differences either did not know about the level In addition, and despite the fact between the ethnic communities. of handgun ownership in the country that sex is not significantly correlated In Macedonia, the three major or refused to answer the question. An with perceptions of firearms owner- ethnic and religious groups sampled almost equal amount of people (about ship in Croatia, both men and women were Macedonian (59 per cent), Alba- 40 per cent) responded that there were overwhelmingly associated themselves nian (34 per cent), and Muslim (5 per ‘very few’ to ‘quite a few’ handguns with the ‘don’t know’ category, which cent). The other 2 per cent comprised in their neighbourhoods. Those who aligns with the low level of direct roughly equal proportions of individu- identified themselves as Serb or Muslim, reporting of firearms ownership (see als who identified themselves as Serbs, however, were more likely to report Table 4). On the one hand, this could Bosnians, Hungarians, and Yugoslavs.17 that there were ‘very few’ or ‘quite a be interpreted as an indication of low Among those who identified them- few’ firearms in their neighbourhoods levels of firearms ownership in Croatia selves as Macedonian, about 35 per (about 50 per cent). overall. The low levels of self-reported cent18 were reluctant to answer the In Serbia the vast majority (about armed victimization in the country question on perceptions of firearms 90 per cent) of respondents identified further support this interpretation in their neighbourhood. Nevertheless, themselves as Serb, with the remaining (see Figure 5). On the other hand, this 24 per cent of the Macedonians who 10 per cent comprising Hungarians, finding could also hint at the unwill- did answer the question said they Muslims, Croats, Bosnians, and ingness of the respondents to answer saw their neighbourhoods as having Montenegrins.19 Of those who identi- the question due to its sensitive nature, ‘very few’ guns. Similarly, those who fied themselves as Serb, about 65 per the history of war, or the link between identified themselves as Albanians or cent stated that there were ‘very few’ or these issues and levels of firearms- Muslim were reluctant to talk about ‘quite a few’ firearms in their neigh- related suicide in the country, for firearm levels in their neighbourhoods, bourhoods and about 21 per cent instance (Bosnar et al., 2005; Cengija with about 60 and 57 per cent report- reported that they did not know or et al., 2012). ing that they did not know or refusing refused to answer the question. That respondents in Bosnia and to answer the question, respectively. As mentioned previously, the sex- Herzegovina, Croatia, Kosovo, The Albanian and Muslim respond- disaggregated responses on firearms Montenegro, and Serbia—all of which ents who did answer the question ownership should be treated with cau- experienced armed conflict directly or
10 Small Arms Survey Issue Brief Number 4 September 2014 Figure 5 Percentage of household respondents who indicated that a household member had been and the distribution of victimization held at gunpoint, by country or territory (n=86) according to location suggest that levels
PERCENTAGE OF RESPONDENTS of firearms ownership and armed vic- 2.5 timization are not related. This point is supported by the fact that Albania and Kosovo have much lower response 2.0 rates to the question ‘Do you, or does anyone else in your household, own a 1.5 handgun?’ (see Table 4). Nevertheless, these findings have to be treated with 1.0 caution because the question dealing with experiences of armed victimiza- 0.5 tion only deals with ‘being held at gunpoint’ and does not inquire about 0 other forms of violence, such as suicide. ALBANIA KOSOVO SERBIA MACEDONIA MONTENEGRO BOSNIA AND CROATIA HERZEGOVINA Moreover, there was no follow-up
Source: Small Arms Survey (2012b) question inquiring about the circum- stances surrounding the incident of being held at gunpoint, making it dif- indirectly during the 1990s—tended analyses, which show that respondents ficult to ascertain which kind of victimi- to select the ‘quite a few’ category is who identified themselves as Albanian zation individuals were exposed to. also of note, as it indicates that they made up nearly 50 per cent of all house- In addition, the household survey perceive firearms ownership levels hold respondents who reported that did not provide a follow-up question as high more than a decade after the they or someone in their household to identify the sex of the victim who cessation of conflict. An exception to had been held at gunpoint in the pre- was held at gunpoint.21 That said, men this is Kosovo, where men were more vious year. Of the respondents who likely to report that there were ‘very reported such experiences, those who and women reported almost equal few’ guns in their neighbourhoods. identified themselves as Serbs were rates of armed victimization in their One explanation for this could be the the next largest group, representing households, which may not be surpris- contentious nature of the issue of fire- about one-quarter of the respondents.20 ing, given the overall low homicide rate arms ownership. By comparison, in Based on the regional distribution in the region (see Table 2). Evidence Albania and Macedonia, most answers of reported experiences of armed vio- suggest that in regions with low homi- for both sexes fall within the ‘very lence, victimization seems to be higher cide rates—such as the Western Balkans, few’ category. in rural areas than in urban areas. In but also Europe in general—there is fact, about 50 per cent of the respond- not only a lower rate of female homi- ents indicated that they or someone cide, but also a smaller gap between Patterns and victims of in their household had been held at the rates of male and female homicides armed violence gunpoint on a farm or in a small town, (GBAV, 2011, p. 45). Together, these The findings of the household survey compared to 40 per cent who stated findings imply that in the Western indicate that about 1.2 per cent of all that they experienced armed victimi- Balkans most armed violence is gender- respondents reported that a household zation in urban areas. At the national neutral, in the sense that it is equally member had been held at gunpoint in level, however, there appears to be likely to affect men and women. the 12 months prior to the administra- more variation, with respondents from Nevertheless, men are more likely— tion of the survey. Figure 5 highlights Albania, Bosnia and Herzegovina, and even if only by 0.3 per cent—to report the distribution of armed violence Kosovo following the regional trend; that they or someone in their house- experiences by country. If weighted meanwhile, Serbia and Croatia revealed hold has been held at gunpoint (see by population, experiences of armed higher levels of victimization in urban Table 5). violence are significantly more com- centres. Montenegro differed most mon in Albania and Kosovo, than in with an equal distribution of victimi- Table 5 Percentage of respondents who reported the rest of the region. In these two zation among rural and urban areas. that someone in their household was held at countries, on average, 2.3 per cent of The reason for the variation at the gunpoint during the previous 12 months, by sex respondents reported that someone national level may be linked to the of the respondent (n=7,106) in their household had been held at different levels of urbanization within gunpoint, more than twice the rate of countries, with Serbia and Croatia Sex Yes No Don’t know Serbia and Macedonia, and about three being two of the most urbanized in Male 1.4% 94.8% 3.8% times that of Montenegro, Bosnia and the region (Hajdu and Racz, 2011). Female 1.1% 95.0% 3.9% Herzegovina, and Croatia. This differ- At first glance, the above findings ence is emphasized further by statistical regarding national victimization rates Source: Small Arms Survey (2012d)
http://www.smallarmssurvey.org 11 Figure 6 Respondents who reported that they or someone in their household was held at regional experiences with war and gunpoint, by sex and age group socio-political insecurity led men to experience conflict and thus gain access Male Female PERCENTAGE OF RESPONDENTS to firearms (Nikolic-Ristanovic and Dokmanovic, 2006, p. 29). 50 Data on relationships between vic- tims and perpetrators in the region is 40 scarce, rendering analysis of domestic violence difficult. That being said, a 30 recent study has detailed the degree of armed victimization within the 20 home in Serbia. According to the data provided by the non-governmental
10 organization Women against Violence Network, between 2010 and 2012 an average of 29 women were killed in 0 15–25 26–35 36–45 46–55 56–65 ≥66 Serbia every year. Most of the women Note: The number of ‘yes’ responses for Kosovo was too small to make the difference between men (n=19) and women (n=12) statistically meaningful. were killed with a firearm and were Source: Small Arms Survey (2012d) older than 26 years of age, with the highest number of victims falling in This finding is further corroborated domestic violence, for instance, is a the 36–45 age category. Comparatively if the experience of victimization is widespread and serious problem in all few women in the 18–25 age category disaggregated by age and sex. As of the countries in the region and that were killed (an average of one per year), shown in Figure 5, individuals who it is primarily targeted against women suggesting that women are most likely fall within the 15–25 and 26–35 age (Nikolic-Ristanovic and Dokmanovic, to experience domestic violence once brackets were most likely to report 2006, pp. 21–32). Indeed, it is estimated they have reached the age of marriage that they or someone in their house- that experiences of armed violence in (Small Arms Survey, 2013b). This insight hold had been held at gunpoint. The the home make up between 70 and 90 is supported by data on the relation- figure also shows that young men, per cent of all cases of domestic violence ship between the perpetrators and especially 26–35-year-olds, are most (SEESAC, 2007, pp. 5–6). victims, which shows that women likely to report armed victimization. According to one study, two pri- who die in domestic violence in Serbia Again, given the history of armed con- mary factors facilitate the high levels of are most often killed by an intimate flict, the prevalence of organized crime, domestic violence in the region. First, partner—a husband, ex-husband, and levels of firearms in the region, this the economic effects of the transition partner, or ex-partner—followed by finding is not surprising. Furthermore, from socialist rule to liberal democracy other male members of the family (see it is in line with the established notion placed a strain on the internal dynam- Figure 7). that although young men usually ics of the family and weakened the The socio-economic transition and account for the majority of the perpe- social control of the state. Second, the historical experiences of armed conflict trators of armed violence, they are also 15 29 likely to be its victims—with – -year- Figure 7 Perpetrators of intimate partner femicides in Serbia, average for 2010–12 old men accounting for about half of the world’s firearm homicide victims, or PERCENTAGE OF RESPONDENTS 70,000–100,000 deaths annually (Bevan 12 and Florquin, 2006, p. 295; Geneva Declaration Secretariat, 2011, p. 113). 10 The aforementioned point is impor- tant as it underscores differences 8 between men’s and women’s experi-
ence of violence in the region. Again, 6 it is difficult to unpack this dimen- sion of armed victimization based on 4 the data from the household survey, since follow-up questions relating to the circumstances were not asked. 2 Nevertheless, recent studies have shown that different forms of violence 0 Husband Partner Ex-husband Son Ex-partner Nephew Son-in-law Father Father-in-law are experienced in varying ways across the region. Available data suggests that Source: Small Arms Survey (2013b)
12 Small Arms Survey Issue Brief Number 4 September 2014 are also thought to influence the level prevalent in the region, they might Federal Republic of Yugoslavia, and of other forms of violence, such as provide insight into the way firearms Slovenia as independent states), but also suicide. According to the European are used. on the international system, especially when it came to relations between major powers Health for All Database compiled by Second, more research should be (the European Union, the Russian Federa- the World Health Organization (WHO, devoted to the link between organized tion, and the United States). The war came n.d.), Croatia, Montenegro, and Serbia crime and firearms proliferation in the to an end in 1995 with the signing of the top the list of gun-related suicides in region and firearms trafficking outside Dayton Peace Accord; see Hoare (2010). Europe. Country-by-country compari- of the region. While many international 4 The SFRY was a federation consisting of sons indicate that the rate of gun-related organizations and regional govern- six republics (Bosnia and Herzegovina, suicides in Montenegro is double the ments claim that organized crime is Croatia, Macedonia, Montenegro, Serbia, rate of Serbia and nearly three times the primary concern when it comes to and Slovenia), two autonomous provinces (Kosovo and Vojvodina), and multiple that of Croatia. Moreover, female sui- arms trafficking, very little data is ethnic and religious identities—includ- cides in Montenegro were also high, available to back up this assertion, ing Albanians, Croats, Bosnian Muslims when compared not only to Serbia and making this an important avenue for (or Bosniaks), Hungarians, Macedonian Croatia, but also to other countries future research. Slavs, Montenegrins, Serbs, Slovenes, in Europe (WHO, n.d.). Although a A third issue concerns the link Turks, and Romany (Bromley, 2007, p. 2). detailed analysis of suicide in the region between firearms legislation, the levels 5 Prior to the dissolution of the SFRY, the is not readily available, scholars have of firearms ownership, and armed vic- territorial defence force of the six repub- 510 000 examined longitudinal trends of suicide timization. Preliminary results suggest lics comprised around , soldiers, 195,000 of whom were on active duty in Croatia and found a close correla- that the tightening of firearms laws in (Bromley, 2007, p. 3). tion between the context of war—in the region since the late 1990s is cor- 6 By 1991, the Serbian government had terms of insecurity and the prolifera- related with a decrease in homicide inherited an estimated 3.5 million small tion of firearms—and an increase of rates. It is not clear, however, whether arms and light weapons from the SFRY firearm-related homicides in the war these stricter laws have an effect on the (Griffiths, 2010, p. 184). and post-war period (Bosnar et al., level of firearms ownership. 7 Arms smuggled into Croatia came primar- 2005; Cengija et al., 2012). Fourth, this Issue Brief calls for ily from the former Soviet and Warsaw Pact states, as well as Latin America— more research on the circumstances especially Argentina (Hajdinjak, 2002, under which armed victimization takes p. 9; Bromley, 2007, p. 13). Conclusion place. Although the findings presented 8 The primary sources of weapons meant In examining the dynamics of hand- here suggest that men and women for the Bosnian forces were the Third gun ownership and armed violence in experience armed violence in different World Relief Agency and large-scale ship- ways, the contexts in which individu- ments from Iran (Bromley, 2007, p. 10). the Western Balkans, this Issue Brief als of both sexes experience armed 9 See, for example, Stojarova (2007). has sought to achieve two goals. The violence remain to be explored. 10 Organized crime groups in Albania first is to provide historical analysis of might be an exception as they appear to the small arms problem in the region be based around kinship and clan ties in order to unpack the reasons behind that are prevalent in the society (Hysi, the high numbers of firearms in circu- Notes 2004, p. 544). Nevertheless, they collabo- lation, including cultural heritage, the 1 This Issue Brief uses the term ‘Western rate with the other organized crime groups Balkans’ to refer to Albania, Bosnia and in the region, as well as those in Italy (CSD, history of armed conflict and political Herzegovina, Croatia, Kosovo, Montenegro, 2004, p. 88; Massari, 2013, p. 80). instability, and the proliferation of Serbia, and the former Yugoslav Republic 11 This is not to say that forms of organized organized crime in the region. The sec- of Macedonia (referred to as ‘Macedonia’). crime did not exist in the region prior to ond goal is to provide a more detailed The designation of Kosovo is without this period. Rather, it can be argued that overview of handgun ownership and prejudice to positions on status, and is in the nature of organized crime in the West- armed victimization by analysing data line with UN Security Council Resolution ern Balkans today is a contemporary form obtained from a region-wide house- 1244 and the International Court of Justice of the phenomenon. See, for instance, Opinion on the Kosovo declaration of Bougarel (1999) and Xenakis (2001). hold survey. independence. Croatia became member 12 This is an average of the data provided Despite the insight obtained from of the European Union on 1 July 2013. for the reported countries and regions. the analysis presented in this paper, 2 Two exceptions are SEESAC (2006a) and See Gallup Balkan Monitor (2010, p. 36). a number of questions remain unan- Schwandner-Sievers (2005). 13 See Grillot (2010). swered, opening the door for further 3 The War of Yugoslav Secession has been 14 Criminologists take homicide to be both research. characterized as the most significant armed the most reliable measure of crime and First, the Issue Brief highlights the conflict to have taken place on European an accurate indicator of a country’s over- all level of criminal violence (Fox and need for greater attention to the issue soil since the end of World War II. It had a significant effect not only on the economy Zawitz, 1998; Howard, Newman, and of culture and firearms proliferation. and political environment in the region Pridemore, 2000; Trent and Pridemore, Although it has been argued that cul- (which gave rise to a new political and 2012). Nevertheless, homicide statistics tural legacies are not necessarily able territorial order, with the emergence of have their drawbacks, one of the most to shed light on why firearms are so Bosnia and Herzegovina, Croatia, the significant being that due to their narrow
http://www.smallarmssurvey.org 13 focus on intentional lethal violence, they Crime Drop: New Directions in Research. Bromley, Mark. 2007. ‘UN Arms Embargoes: underreport the various other violent Basingstoke: Palgrave Macmillan. Their Impact on Arms Flows and Target ways in which individuals lose their Anastasijevic, Dejan. 2006. ‘Organized Crime Behaviour—Case Study: Former Yugo- lives (Krause, 2009; Geneva Declaration in the Western Balkans.’ Paper presented slavia, 1991–96.’ Stockholm International Secretariat, 2011). at the First Annual Conference on Human Peace Research Institute. Stockholm: 15 Between 2004 and 2009, the average annual Security, Terrorism, and Organized Crime SIPRI. global violent death rate was 7.9 per 100,000 in the Western Balkan Region, Ljubljana, Carapic, Jovana. 2014. ‘Order and Authority inhabitants. This number decreased slightly 23–25 November. Within (and beyond) the State.’ Unpub- in recent years, with the most recent data —. 2008. ‘What’s Wrong with Serbia?’ Belgrade: lished doctoral thesis. Geneva: Graduate suggesting that between 2007 and 2012, European Stability Initiative. 31 March. Institute of International and Develop- the annual global violent death rate was 14 Small Arms Survey Issue Brief Number 4 September 2014 content/kriminalni-obracuni-vracaju- Judah, Tim. 2000. The Serbs: History, Myth, and building in Bosnia and Kosovo.’ Problems li-se-devedesete-na-beogradske-ulice/ the Destruction of Yugoslavia. New Haven of Post-Communism, Vol. 51, No. 3, pp. 53–60. 24625599.html> and London: Yale Nota Bene and Yale Radojevic, Nemanja and Dragana Cukic. 2012. — and Andy Heil. 2012. ‘Serbia Tries to Scrap University Press. ‘A Comparative Analysis of Homicides Its Gun Habit.’ Radio Free Europe/Radio KAS (Kosovo Agency of Statistics). n.d. ‘Total Committed with Knives and Firearms Liberty. 21 December. Violence Deaths, 2002–2011.’ Pristina: KAS. after Introducing Legal Restrictions.’ Gow, James. 2003. The Serbian Project and Its http://www.smallarmssurvey.org 15 SORS (Statistical Office of the Republic of —. 2003. Convention against Transnational WHO (World Health Organization). n.d. Serbia). 2012 ‘Statistical Yearbook 2012.’ Organized Crime. New York: UN. ‘European Health for All Database Belgrade: SORS. About the Small Arms Survey improvements in human security worldwide. To date 112 states have endorsed the Declaration. The Small Arms Survey serves as the principal interna- The Small Arms Survey hosts the Geneva Declaration tional source of public information on all aspects of small Secretariat, and provides policy-relevant information to arms and armed violence, and as a resource centre for support armed-violence prevention and reduction. governments, policy-makers, researchers, and activists. In For more information, visit www.genevadeclaration.org. addition to Issue Briefs, the Survey distributes its findings through Research Notes, Working Papers, Occasional Papers, Special Reports, Handbooks, a Book Series, and its annual Credits flagship publication, the Small Arms Survey. The project has an international staff with expertise in Author: Jovana Carapic security studies, political science, international public policy, Copy-editor: Tania Inowlocki law, economics, development studies, conflict resolution, Proofreader: Stephanie Huitson sociology, and criminology, and works closely with a world- Design and layout: Rick Jones ([email protected]) wide network of researchers and partners. The Small Arms Survey is a project of the Graduate Institute of International and Development Studies, Geneva. Contact details For more information, please visit: www.smallarmssurvey.org. Small Arms Survey 47 Avenue Blanc, 1202 Geneva, Switzerland About the Geneva Declaration on Armed t +41 22 908 5777 f +41 22 732 2738 e [email protected] Violence and Development The Geneva Declaration on Armed Violence and Develop- ment is a high-level diplomatic initiative designed to help states and civil society actors to achieve measurable reduc- tions in the global burden of armed violence and tangible 16 SmallSudan ArmsIssue BriefSurvey NumberIssue Brief 6 April Number 2007 4 September 2014