ARMED VIOLENCE ASSESSMENT Issue Brief Number 1 September 2011

India’s States of Armed Violence Assessing the Human Cost and Political Priorities

A traditional preoccupation with south. Divergence can be seen also terrorist activity in 2009, when there military threats to the state has long among major cities, with violence were 32,369 homicide victims and dominated Indian policy and activism much more of a problem in , for 2,231 deaths linked to terrorism. aiming to prevent and reduce armed example, than in Calcutta. Terrorist violence between 1994 and violence. This realist perspective has This Issue Brief focuses primarily 2009 resulted in 58,288 deaths, an the effect—whether intended or not— on the three problems of insurgency, average of more than 3,600 per year. of displacing consideration of other terrorism, and criminal violence in More than half of the dead, 52 per sources of danger related to armed vio- comparative perspective. In addition cent, were reportedly civilians and lence. In recent years, the government to examining patterns and trends, it members of the security forces. reviews government policy and spend- has increased its focus on secession- Between 1988 and 2009, the Kashmir ism and foreign-sponsored terrorism, ing devoted to tackling the problem of conflict caused at least42 ,657 deaths, influenced by external events such as armed violence. Among its key findings according to official data (more than the 11 September 2001 attacks on US are the following: 80,000 according to other sources). targets, but also by domestic terrorist Criminal violence caused more than The Maoist (Naxalite) insurgency incidents, notably the attacks of 14 times as many violent deaths as has intensified in recent years, 13 December 2001 on the Indian par- liament and the Mumbai attack of 26– 29 November 2008. Yet other forms of violence are far more devastating in terms of numbers of killed, injured, victimized, and affected. Some of these other forms of violence are especially serious in India, such as caste and dowry crimes, while others, such as criminal violence and suicide, are found in all countries. Although India’s violence-related problems are serious and widespread, its rate of violent death is not especially high when compared with that of many other countries. In terms of interna- tional homicide rates, India is among the lower–middle ranking countries (see Table 1). But India’s national sta- tistics conceal wide variation among its 28 states and seven union territories. Armed violence is much more serious in the north, the north-east, and the

Maoist-affected regions. It is believed The Taj Mahal hotel burns during the Mumbai terrorist attack of 26–29 November 2008. to be far lower in much of the country’s © Altaf Qadri/AP Photo

www.india-ava.org 1 spreading to 20 of the country’s 28 Effective policy on armed violence light-weapons acquisition and prolif- states and close to one-third of the requires highest-level political com- eration. All these issues are the sub- country’s 626 districts. It led to 896 mitment and whole-of-government ject of domestic debate and will also deaths in 2009, of which 392 were coordination, bringing together the interest the international community those of civilians. ministries of Defence and Home of governments, non-government or- Rates of violent crime reported in Affairs, with systematic civil-society ganizations, and diaspora groups. 2009 varied greatly across the engagement. Evidence-based eval- country, ranging from fewer than uation is needed to ensure that 10 reported crimes per 100,000 baselines are established and tracked A note on statistics people in the north-eastern state of over time. As in many countries, data on armed violence in India should be regarded Nagaland to 111 reported crimes The India Armed Violence as suggestive rather than exact. For per 100,000 in the southern union Assessment (IAVA) project examines example, different data sources show territory of Puducherry. the many dimensions of armed vio- significantly different levels of homicide. In 2009 police received reports of lence, with the aim of broadening de- Most analysts regard criminal-justice 8,383 deaths from domestic vio- bate on the causes and responses to data as the most accurate, but public- lence, including dowry deaths, and armed violence. The themes intro- health sources—collated by the World 89,546 non-fatal cases of torture or duced here will be explored at greater Health Organization—consistently cruelty by husbands and relatives. length in future IAVA Issue Briefs. In show homicide in India to be twice as Government spending prioritizes forthcoming editions, Indian experts high. As Table 1 shows, this disparity is international threats over domes- will consider the geographic distribu- not unique to India. In general, criminal tic threats. The 2008–09 budget tion and causes of violent mortality data shows lower levels of homicide, allocated five times more funds and morbidity, the impacts of Maoist due to reliance on narrower, juridical to national defence (INR 1,056 and other insurgencies, the scale and billion, or USD 23 billion) than to distribution of caste violence, trends determinations. Public-health data policing and paramilitaries (INR in law enforcement and policing, and usually relies on extrapolation from 206 billion, or USD 4.5 billion). wider patterns of small-arms and samples of death certificates, which reflect the finding of a medical exam- Table 1 International homicide rates for selected countries, according to public-health and iner. Weaknesses and idiosyncrasies criminal-justice sources in national reporting systems affect both categories.1 Country Year Homicides per Homicides per A lack of accurate national data on 100,000, criminal- 100,000, public- violence and victimization also makes justice sources health sources it difficult to compare levels of violent 1 South Africa 2008 37 68 death within India. Analysis of the prob- 2 2008 22 25 lem at the national level must rely on 3 Russian Federation 2008 14 20 two principal sources, both flawed by serious weaknesses. The most exhaus- 4 Sri Lanka 2008 7.4 6.8 tive source of data on violent deaths 5 2008 6.8 3.4 and other forms of homicide is the 6 USA 2008 5.2 6.0 annual reports of the National Crime Records Bureau (NCRB), a police 7 Iran 2004 2.9 2.5 agency operating under the Ministry of 8 India 2007 2.8 5.5 Home Affairs. The NCRB collects data 9 Nepal 2007 2.2 13.6 from state and city police agencies. The 10 China 2007 1.2 2.1 data reflects major differences in the way in which the police agencies report on 11 Germany 2008 0.8 0.6 crime. Differences in state reporting 12 Indonesia 2004 0.7 9.3 practices might explain why southern 13 2008 0.5 0.5 states have the highest crime rates in

Note: Latest available data are shown, rounded to two significant digits. the country according to the NCRB, Sources: UNODC Homicide statistics (2003–08). The source of all public-health data is the World Health Organization, except for Brazil (Brazilian Ministry of Health), even though they are widely reputed Germany, and Russia (WHO European Health for All Database), and United States (Pan American Health Organization). Sources of criminal-justice data are Ministry of to have low crime rates (NCRB, 2011a). Justice (Brazil), national police (India, South Africa, and Sri Lanka), national statistics office (China, Nepal, and Pakistan), UN Survey of Crime Trends and Operations of Criminal Justice Systems (Germany, Iran, Japan, Russian Federation, and United States), and Interpol (Indonesia). Consequently, the official data used in

2 Small Arms Survey Issue Brief Number 1 September 2011 NCRB reports is suggestive rather than conclusive. No Indian government agency is known to keep comprehensive records of casualties from terrorist violence in India. Rather, the South Asia Terrorism Portal (SATP), a private institute in , collates data on terrorism and armed conflicts from published reports, including the news media and annual reports of the Ministry of Home Affairs. SATP data is widely used, but is limited by weaknesses in its sources. The categorization of the dead as terrorists or civilians, for example, or killed in battle or died after arrest, is notoriously unreliable. Allegations of encounter deaths, or extra-judicial kill- ings attributed to Indian security forces, are almost impossible to identify and interpret through such statistics.2 Whether this significantly affects overall statistics on violent death is unknown. Villagers at a public rally organized by Maoist rebels in Gaya district, Bihar, 2009. © Manish Bhandari/AP Photo Armed violence or terrorism? clearly the government’s priority, as as more contemporary issues such as Terrorism is India’s most prominent the annual reports of the Ministry of the lack of effective governance that is armed-violence issue. But it is not the Home Affairs reveal (MHA, 2010a). propelling the Maoist insurgency. cause of most deaths. That claim Insurgent and terrorist dangers in Major terrorist attacks in recent years belongs, rather, to individual acts of India are complex and multifaceted. include the following (SATP, 2009, 2010): murder. As Figure 1 shows, in 2009 Almost all the conflicts underlying more than 14 times as many violent Jaipur, 13 May 2008, killing two, deaths were attributable to criminal such violence are decades old, with injuring about 20 people; murder as to terrorist activity: 32,369 strong local roots (Acharya, 2006, p.320). Bangalore, 25 July 2008, killing two, cases of criminal murder, compared India’s many forms of secessionist injuring about 20; with 2,231 deaths from terrorism warfare and terrorist conflicts involve , 26 July 2008, killing (NCRB, 2011a; SATP, 2010). While old-fashioned nationalist or ethno- 56, injuring more than 200; violent crime might attract more day- nationalist movements, as in Kashmir Mumbai, 26–29 November 2008, to-day media coverage, terrorism is and the north-eastern states, as well killing 164 people and injuring at least 308; Figure 1 Comparing fatalities from crime and terrorism, 2005–09 New Delhi, 13 December 2008, killing 30, injuring more than 100; Murder Terrorism Pune, 13 February 2010, killing nine, 35,000 injuring 45; and 30,000 New Delhi, 20 September 2010, 25,000 injuring 2. 20,000 Islamist groups were reportedly 15,000 responsible for most of these terrorist 10,000 attacks. Indeed, the dangers of seces- 5,000 sionist insurgency have different 0 2005 2006 2007 2008 2009 ideological roots. The Maoist insur- Sources: NCRB, 2011; SATP, 2010 gency is concentrated in the large

www.india-ava.org 3 ‘Red Belt’ but has spread to 20 of 28 Terrorism and armed conflict each of the three main terrorist or insur- Indian states and roughly 200 of the gency threats: in Kashmir, in north- ‘Terrorism’ typically refers to politically country’s 626 districts, most of them east India, and Maoist insurgency. motivated secessionist violence, or to relatively remote and impoverished. internationally sponsored political Although the insurgency had previ- Terrorism and insurgency in Kashmir violence. It usually does not include ously been confined to tribal and ru- communal or sectarian violence Since India and Pakistan became inde- ral areas, it is now beginning to be felt (Hoffman, 2006, chapter 1). Even when pendent in 1947, Jammu and Kashmir in Indian cities as well (Ramana, is still the major source of contention the definition is narrowed in this way, 2009). On 15 February 2008, for and conflict between them. Kashmir India ranks among the world’s most example, some 400 to 500 Maoists at- joined India under an Instrument of terrorism-afflicted countries (USDoS, tacked police facilities in the towns of Accession in 1947. Pakistan claims 2009, p.141). Although many academic Nayagarh and Daspalla in Orissa, that because most people of the state observers find the concept of terrorism killing 14 (Venkataramani, 2010). are Muslims and were not consulted problematic, among Indian academics In dealing with these threats, India in the accession process, the territory the concept is widely accepted, even has used strategies of both negotiated should join Pakistan. India has resisted by observers who are most concerned settlement and military force, although demands for a referendum, preferring with root causes of terrorist violence the latter has been more prominent. a scheme known as ‘Line of Control (Acharya, Singhdeo, and Rajaretnam The results have been mixed: in places (LoC) plus’, under which it would 2010). SATP statistics show that between such as Punjab the government has concede areas taken by Pakistan in the 1994 and 2009 terrorist violence resulted used military means to contain mili- 1947–48 war (Chari, 2006, pp. 130–31). in 55,643 deaths, an average of more tancy almost entirely, while in Mizoram India rejects external mediation, insist- than 3,400 per year. Of this total, 52 (see below) the insurgents were brought ing that the bilateral Shimla Agreement per cent of the dead were reported to into the mainstream through their of 1972 provides the framework for participation in the electoral process be civilians or members of the security negotiation of Kashmir issues. (Gill, 1997; Bhaumik, 2007, pp. 12–13). forces (see Figure 2). India and Pakistan fought major The government has begun to address Given the magnitude of the threat, wars over Kashmir in 1947–48 and the Maoist insurgency by using a mix- it is not surprising that the Indian 1965, and have engaged in a number of ture of development incentives and government ranks terrorism among smaller confrontations, most notably military force, but no one is sure how its foremost national security priorities. the Kargil war of 1999. Beginning in long this will take, or how many lives Recent attacks on urban centres and the 1989, foreign insurgents, including will be lost before the threat is contained. rapid spread of the Maoist rebellion veterans from wars in Afghanistan, Terrorism and separatism dominate reinforce this concern (Prakash, 2009). infiltrated Kashmir from Pakistan. Indian official priorities, as manifested Contrary to official claims, militancy These incursions, coupled with an esca- in the budgets of the Ministry of in Kashmir is not waning either, as lation of public uprisings, marked the Defence and the Ministry of Home shown by the sudden rebirth of violent bloodiest period of militancy in Kash- Affairs (see below). Realist political opposition and the harsh crackdown mir (Acharya, 2004, p. 55). Since then perspectives—which prioritize dan- since the summer of 2010 (The Economist, these militant groups have engaged gers to the Indian state—focus on the 2010). The following sections consider Indian security forces in a protracted military capabilities of neighbouring states and groups trying to under- Figure 2 Breakdown of fatalities from Figure 3 Fatalities from terrorist violence mine the authority of the central terrorist violence, 1994–2009 in Kashmir, 1988–2009 government (Chakma, 2009). The Terrorists (27,529) Terrorists (22,174) emphasis on internationally contested Civilians (22,286) Civilians (14,566) areas, especially Kashmir and the Security forces (5,828) Security forces (5,917) north-east, overshadows other threats to human security, such as those arising from communalism, sectarianism, and crime. This gap is attracting greater recognition. Many among India’s vast NGO community acknowl- edge the discrepancy (Acharya and Acharya, 2002).

Source: SATP, 2010 Source: SATP, 2010

4 Small Arms Survey Issue Brief Number 1 September 2011 Figure 4 Violence trends in Jammu and Kashmir present century. But in 2010 the con-

Incidents Casualties flict re-emerged suddenly as a locally led uprising, where civilian protests 3,500 are more important than insurgent 3,000 attacks, in a scenario comparable to the 2,500 Palestinian Intifada of 1987–91 (The 2,000 Economist, 2010). Outbreaks of violence 1,500 still occur. In lop-sided clashes, Indian 1,000 security forces killed 100 Kashmiri 500 protestors and by-standers in 2010. 0 2004 2005 2006 2007 2008 2009 Protestors have also been perpetrators

Sources: MHA, 2010a, p. 6 of some deadly violence, as in Novem- ber 2010 when Kashmiri men killed two police officers (The Economist, 2010). conflict which has cost at least 42,657 for past offences including enforced The continued violence suggests lives, according to official data, and disappearances of thousands of people that exclusive reliance on armed forces more than 80,000 according to other during the armed conflict in Kashmir to deal with terrorism or insurgency sources (SATP, 2010; Mishra, 2010). since 1989’ (AI, 2009). brings limited results. At the same This is an average of at least 1,900 The Ministry of Home Affairs con- time, democratic governance and a deaths per year between 1988 and tends that the incidence of violence massive infusion of development aid 2009—and possibly twice as many has declined progressively since 2004 to the state has not contained or re- (see Figure 3). (see Figure 4). It sees a ‘perceptible versed discontent among Kashmiris, Estimates of the number of Indian improvement’ in the overall security especially among surrendered militants. soldiers and paramilitary troops situation in Kashmir, due to ‘several The problems posed by reintegrated deployed to Kashmir range from holistic measures taken by the govern- surrendered militants are not unique to 170,000 to 500,000 (AFP, 2011; BBC, ment and the people’s yearning for India. Rather, they illustrate the impor- 2011). Even the lowest estimate would peace’ (MHA, 2010a, p. 6). Successes tance of issues that have proved hard make this the largest military deploy- include countering the challenge to manage everywhere (Muggah 2009).3 ment in the world today, larger than ‘posed by the terrorists and violence the total armies of Britain, France, or sponsored from across the border’ Terrorism and insurgency in Germany, bigger than the entire Inter- (MHA, 2010a, pp. 7–11). The lack of north-east India national Security Assistance Force in independent data makes it difficult to North-east India is the site of consider- Afghanistan (IISS, 2010, pp. 129, 134, evaluate such statements, however, able ethno-political unrest and armed 168; NATO, 2010). Accusations of and the resumption of violence in violence, in the form of ethnic seces- human-rights abuses by the security 2010 suggests that these statements sionist movements and even ethnic forces are rampant. Amnesty Interna- should be viewed with caution. cleansing. Important groups in the tional’s 2009 report on abuses in Jammu These measures, coupled with more multi-ethnic and multi-cultural states and Kashmir claims that in 2008 secu- effective security operations by Indian of Assam, Arunachal Pradesh, Manipur, rity forces killed at least 40 people for forces and more sensitive official leader- Meghalaya, Mizoram, Nagaland, defying curfew restrictions. The report ship, led to a steady decline in fatalities Sikkim, and Tripura have never fully also found that ‘impunity continued in Kashmir in the early years of the accepted integration into the country. They remain politically sensitive and Figure 5 Fatalities from terrorist violence in India's north-east, 2005–09 prone to revolt. As measured in fatali- ties, however, most of these conflicts 1,400 have been relatively stable in recent 1,200 years (see Figure 5). 1,000 Although the roots of north-eastern 800 conflict vary among states, there are 600 some common features. In the state of 400 Assam several insurgencies fight for 200 attention and domination. The xeno- 0 2005 2006 2007 2008 2009 phobic United Liberation Front of Source: SATP, 2009a Assam (ULFA) spearheads violent

www.india-ava.org 5 Figure 6 Fatalities from Maoist violence, 2005–09 declined in intensity in the late 1970s, when most senior leaders had died 1,400 or were in prison, the movement is 1,200 expanding once more. 1,000 Maoist violence seems to be increas- 800 ing in terms of both number and lethality 600 of attacks. These escalations are taking 400 place despite the ongoing Operation 200 Green Hunt: a ‘coordinated and joint 0 2005 2006 2007 2008 2009 action’ by central government para-

Source: SATP, 2009a military and state police forces which began in November 2009 in affected opposition to non-Assamese authority, toral process (Bhaumik, 2007, pp. 12–13). states (MHA, 2010a, p. 6). The number to legal migration from the rest of In Assam, by contrast, the situation of casualties caused by Maoist activi- India, and to illegal migration from remains precarious. Local residents ties has increased in recent years (see . In the same state, the Bodo and advocacy groups denounce exces- Figure 6) to 896 dead in 2009, of whom uprising arose from tribal perceptions sive reliance on the military to stem 392 were civilians (SATP, 2009b). of neglect and discrimination (Bhaumik, violence which has led to human-rights In a widely cited statement, Prime 2007, p. 3). The insurgency among Mizo violations (Nepram, 2009). Legislation Minister said that people originated from a perceived loss such as the Armed Forces (Special left-wing extremism poses ‘perhaps of identity to Assamese domination Powers) Act of 1958, which gives the the gravest internal security threat’ and discrimination (Bhaumik, 2004, military legal immunity for their facing the country, adding that ‘despite p. 225). In Nagaland, militancy stems actions, reinforces the tendency to efforts, the level of violence in the from the independence demands of rely on violent suppression rather affected states continues to rise’ (Indian indigenous people represented by the than negotiation (AFSPA, 1958). Express, 2009). The movement has suc- Naga National Council. The insurgency cessfully infiltrated tribal areas, where in Tripura involves parochial religious The toll of Maoist insurgency illiteracy levels are high and where and ethnic issues. In Manipur, the there is a lack of basic facilities and In the Indian heartland, various left- unrest is rooted in attempts to protect almost no official government pres- the multi-ethnic and multi-cultural wing extremist groups collectively ence. According to the Indian Ministry character of the native population operating under Naxalite or Communist of Home Affairs: (Saikia, 2001). Party of India (Maoist) leadership New Delhi’s efforts to deal with continue to perpetrate violence in the Left Wing Extremists operate in north-eastern militancy rely on military extensive ‘Red Belt’ of Andhra Pradesh, the vacuum created by functional force and political accommodation, Bihar, Chhattisgarh, Jharkhand, inadequacies of field level govern- typically bringing rebels into state Madhya Pradesh, Maharashtra, Orissa, ance structures, espouse local government. This combination has West Bengal, and other areas. The demands, and take advantage of yielded mixed results. In cases such as Maoist insurgency began as a peasant prevalent dissatisfaction and feel- Mizoram, the government has brought uprising in 1967 in the village of ings of perceived neglect and rebel groups into the mainstream elec- Naxalbari in West Bengal. Although it injustice among the under privileged and remote segments of population. (MHA, 2010a, p. 17) Figure 7 Reported murders, attempted murders, and related deaths, 2005–09 Murder Attempted murder Culpable homicides Dowry deaths Suicides Accidental shooting Maoist insurgents engage in system-

35,000 atic attacks on development works in an effort to undermine the government’s 30,000 authority and perceived effectiveness. 25,000 They have targeted school buildings, 20,000 railways, roads, and power and tele- 15,000 com infrastructure (MHA, 2010a, p. 17). 10,000 The threat posed by the rebellion has 5,000 a national reach. Maoist strongholds 0 2005 2006 2007 2008 2009 in the states of Andhra Pradesh, Sources: NCRB, 2010, 2011a, 2011b Chhattisgarh, Jharkhand, Orissa, and

6 Small Arms Survey Issue Brief Number 1 September 2011 West Bengal contain 85 per cent of 111 reported crimes per 100,000 people ing dowry deaths, in 2009. Even more India’s coal resources, and all are in the southern union territory of common are non-fatal cases of torture heavily affected by Maoist depreda- Puducherry (NCRB, 2011a, pp. 23, 26). or cruelty by husband and relatives, tion (Magioncalda, 2010). Since coal with 89,546 cases reported to police in constitutes more than 40 per cent of Violent death: murder and suicide 2009 (NCRB, 2011a, p. 81). Despite the India’s primary energy and 70 per stigma and a harsh punitive regime, According to the NCRB, the most fre- cent of its fuel for electricity genera- dowry gifts remain a local custom. quent motives cited for murder and tion, the implications extend across ‘With get-rich-quick becoming the culpable homicide (manslaughter) are the country (USEIA, 2010, pp. 1, 8). new mantra, dowry became the per- personal vendetta, disputes over prop- The Maoist threat is therefore both fect instrument for upward material erty, financial gain, intimate-partner a manifestation of the economic and mobility’, and consequently dowry conflicts and sexual affairs, dowry, social underdevelopment that has harassment has become a part of fam- politics, communalism, and lunacy, as plagued rural India for decades and ily life (Vinayak, 1997). According to a defined by the NCRB (NCRB, 2010, an obstacle to the future development study by the Institute of Development pp. 55–56). The victims tend to be of those same areas. It is an example and Communication, ‘the quantum of young adult men, with 45 per cent of of how ‘a sense of injustice, related dowry exchange may still be greater 18 particularly to gross inequality, can be murder victims aged between and among the upper classes, but 80 per a good ground for rebellion—even 30 years (NCRB, 2010, p. 59). Since cent of dowry deaths and 80 per cent of bloody rebellion’ (Sen, 2008, p. 8). young adult men are the most eco- dowry harassment occurs in the middle Improving the lives of the people in nomically productive segment of the and lower strata’ (Vinayak, 1997). rural areas requires concerted devel- population, their deaths represent a Suicide is not exceptionally common opmental efforts on the part of the significant loss of productive human in India, but suicide, often by pesticide government, but these efforts are being, capital for the country. consumption, ranks among the fore- and will continue to be, compromised Dowry deaths are an especially most causes of fatal deaths in rural by the Maoist threat. serious Indian pathology. Official areas (Eddleston and Konradsen, 2007). statistics show that there were 8,383 Rural areas are home to the majority deaths from domestic violence, includ- of the population (about 72 per cent Crime, domestic violence, suicide, and unintentional injury Terrorism is an issue of immense importance for India, but crime, domes- tic violence, and suicides are just as pressing. NCRB statistics show a con- tinuous increase in the reporting of crime, both in absolute numbers and proportionately. In 2009 police received more than 6.6 million complaints re- lating to criminal incidents, compared with 5 million in 2005, a figure which represents an increase of 30 per cent (NCRB, 2011a, p. 23; see Figure 7). In the absence of valid and reliable national polling, it is impossible to determine whether this rise reflects an increased willingness to report crimes or an increase in their actual incidence. In terms of incidences of violent crimes (crime against the body) reported in 2009, rates varied greatly across the country, ranging from fewer than 10 reported crimes per 100,000 people in A shelter for victims of dowry violence, New Delhi. the north-eastern state of Nagaland to © Elizabeth Dalziel/AP Photo

www.india-ava.org 7 in the 2001 census), most of whom a lack of personnel, uneven deployment concern is the crime–terrorism nexus. depend on agriculture for their liveli- of available forces, and a lack of equip- Crime syndicates help terrorist groups hoods (MHA, 2001). Farmers are often ment and training limit the effectiveness by providing resources to support, compelled to commit suicide to escape of the security forces. NCRB data shows conceal, or conduct their activism, ruinous debt following crop failures that in 2009 just 49.2 police personnel while criminal entrepreneurs benefit or household disasters (Patel, 2007; were deployed for every 100 square from terrorists’ military skills and Nagraj, 2008). According to an April kilometres, and just 1.3 constables per networks (Gunaratna and Acharya, 2009 report, some 1,500 farmers in the 1,000 people (NCRB, 2011a, p. 169). 2007, p.100). In one prominent exam- state of Chhattisgarh committed sui- Endemic corruption has reportedly ple, two perpetrators of the 1993 cide after being driven to debt by crop made the police more susceptible to Mumbai bomb blasts, Tiger Memon failure in the previous year (The Inde- the influence of criminals, unscrupu- and Moolchand Shah (also known as pendent, 2009). The provinces most lous politicians, and private entre- Choksi), are alleged to have used affected by Maoist insurgency, such as preneurs (Venugopalan, 2002, p. 97). criminal networks to channel illegal Andhra Pradesh and West Bengal, also Somaiah describes the widespread pop- earnings and fund their bombing have the highest number of suicides ular perception that police personnel— operations. Likewise, Dawood Ibrahim (NCRB, 2011b, pp. 171-72). whether from the central government, masterminded the 1993 Mumbai bomb As shown by the NCRB, consuming state, or community—are not apoliti- blasts and also has a huge national poison and hanging were the most cal or impartial in the application of and international criminal network common methods of suicide, used in the laws (Somaiah, 2002, p. 908). (Sarkar and Tiwary, 2001). 34 and 32 per cent of cases respectively, Venugopalan finds that it has ‘become Criminal gangs are heavily involved followed by self-burning and drowning, common practice for politicians and in the trade in illegal firearms (Lal,2007 ). at 9 and 6 per cent respectively (NCRB, bureaucrats to use the police in their Most crimes are committed with ille- 2011b, p. 184). The largest number of power struggle, thereby undermining gal weapons (NCRB, 2011a, p. 340). suicides was reported among people its independence and accountability’ During a ten-year period police who were self-employed: almost 40 (Venugopalan, 2002, p.97). Corruption seized 4,500 illicit arms. NCRB statis- per cent of the total of 127,151 officially leads police to under-report crime and tics suggest that licensed firearms were registered victims in 2009 (NCRB, potentially avoid thorough investiga- used in only 371 murder cases in 2011b, p.182). tions. Although this does not apply to 2009, while 2,722 cases involved unli- all Indian policing, issues of police cor- censed firearms (NCRB,2011 a, p. 340). The context of violent crime ruption, as well as inadequate facilities, Although some murders are commit- A number of socio-economic factors equipment, and training, are widely ted with licensed, legally owned guns, contribute to crime, including pov- accepted as part of the country’s grow- most crime guns are illegally made erty, inequality, unemployment, rapid ing crime problem (Verma, 1999). craft weapons (kattas), mostly made in urbanization, and uncontrolled urban Organized criminal networks the states of Uttar Pradesh and Bihar. migration (GD Secretariat, 2007, chap- exacerbate the problem of crime in A small proportion of illegal guns are ter 3). The inefficiency of the security cities. These networks extort money, manufactured abroad and smuggled apparatus and the pervasive sense of kidnap hostages for ransom, and into the country. impunity are further influencing fac- launder money, besides trafficking tors. The causes of inefficiency are not firearms, women and children, and Women and violence all inherently structural; in many cases drugs (Lal, 2007). A particular cause of Women are especially vulnerable to armed violence, although their vul- Figure 8 Reports of crime against women nerability and suffering is overlooked No. of incidents reported in official statistics. But reports of crimes against women have grown 210,000 rapidly in recent years (Figure 8). Not 180,000 only are women especially vulnerable 150,000 to direct violence, they are dispropor- 120,000 tionately victimized by less visible, 90,000 indirect effects. Both aspects require 60,000 better monitoring and more aggres- 30,000 sive action. 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 The Geneva Declaration notes that, Source: NCRB 2011a ‘Women and girls are affected by armed

8 Small Arms Survey Issue Brief Number 1 September 2011 Figure 9 Central-government defence and police/paramilitary budgets (INR billions) tral police forces, and the intelligence MoD MHA agencies—received 25 per cent increases in their budgets (Homeland Security 1,600 Research, 2009). Combined revenue 1,400 1,200 and capital expenditures (operations 1,000 and equipment) for the police and para- 800 military forces of the MHA increased 600 dramatically from IND 106 (USD 2.4 400 billion) to INR 206 billion (USD 4.5 200 billion) between 2004 and 2009, a trend 0 2004–05 2005–06 2006–07 2007–08 2008–09 2009–10 2010–11 that seems likely to continue (PTI, 2009).

Sources: MHA, 2009, p. 116; MHA, 2010b, p. 173; MoD annual reports Most of the money spent on defence supports conventional high-intensity combat forces and nuclear military violence in different ways, including of violence against women. Careful capabilities. MoD capital spending— direct and indirect conflict violence, tracking of vulnerable groups, starting procurement of equipment—is domi- and by lethal and non-lethal non- with women, is an essential step for nated by investment in major weapon conflict violence’ (GD, 2008, p. 106). Indian violence monitoring. systems for international security The World Bank’s authoritative World (Behera, 2010). But while most of the Development Report 2011 concludes Budgets and national increase is aimed at responses to that the direct impact of violence falls foreign-state threats, some will go primarily on young men, those most priorities towards domestic counterterrorism, likely to be directly involved in crime, Government spending continues to especially since the November 2008 gangs, and conflict. Women suffer dis- emphasize international threats more attacks. It is difficult to fully distin- proportionally from the indirect effects than threats to human security. In guish spending on the military from of violence, especially from the destruc- 2008–09, the most recent year for which spending on internal security, since tion of their families. While they are accurate data is available, the budget much of the Army is deployed domes- less likely to die from violent attack, for policing and paramilitaries was tically for internal security (India Today, women and their children are more INR 206 billion (USD 4.5 billion), about 2010). Cross-border terrorist threats, likely to suffer subsequent emotional one-fifth as much as the sum spent on and the possibility of a terrorist attack trauma, impoverishment, and home- national defence that year: INR 1,056 triggering a conventional war between lessness (World Bank, 2011, p. 6). billion (USD 23 billion) (see Figure 9). India and Pakistan, further blur the To its credit, the NCRB does stress There are limits to such compari- distinction between spending on inter- several kinds of violent crime specifi- sons, however. The budgets of both nal security and spending on external cally directed at women, especially the MoD and the Ministry of Home security. rape, dowry death, sati, and sexual Affairs (MHA) have grown in recent harassment (NCRB, 2011a, chapter 5). years, often rising at different rates; Most data collected by the NCRB, but defence spending averages five to Implications however, still does not disaggregate seven times the outlays for policing The three faces of armed violence crime and violence by gender, and and paramilitaries. The MHA policing reviewed here differ enormously in despite the grievous impact on women, and paramilitary budget is not com- scale and effects. Terrorism and insur- official statistics still do not emphasize prehensive, since other ministries and gency have effects which go far beyond gendered aspects of broader aspects state governments control much law- the direct deaths and injuries that they of crime and violence, or the indirect enforcement spending. And the MoD cause, since they undermine state effects that affect women most. contributes to counterterrorism opera- security and economic development. Civil-society organizations have tions. These limitations notwithstanding, But in terms of the sheer numbers of been especially instrumental in raising the differing budgetary emphases lives destroyed, criminal violence and awareness of the problem. But India are a clear sign of the difference in suicide deserve more attention. It is still lacks crucial instruments to guide national priorities. clear that better data on armed vio- policy, especially mechanisms to meas- Following the Mumbai attacks in lence is needed in order to clarify ure violence by husbands (intimate- November 2008, the agencies respon- national priorities and develop more partner violence) and relatives, widely sible for homeland security—primarily effective policy. Public survey research regarded as the most common forms the paramilitary forces, state and cen-

www.india-ava.org 9 Salwa Judam militia stand guard against a possible Maoist attack during re-polling in the village of Pandewar, November 2008. © Adrian Fisk

could help to clarify the severity and in India—such as caste violence and principles of whole-government re- nature of armed violence, and over- dowry crime—others resemble prob- sponsibility and cooperation between come persistent doubts about the lems faced by other countries. The the state and civil society—principles reliability of existing data. Indian government’s awkward or that are winning ever greater accept- High crime rates, suicides by limited engagement in international ance elsewhere. farmers, gender inequality leading to dialogues on armed violence and even A comprehensive and holistic domestic violence and dowry deaths, disarmament has prevented the country approach to the problem needs to communal violence, and sectarian vio- from benefitting fully from experiences address not only immediate dangers lence are all symptomatic of India’s and lessons learned elsewhere. posed by armed men, but also infra- failure to develop evenly in either A problem that appears repeatedly structure needs and economic devel- social or economic terms. Terrorism, in this review is the lack of official co­ opment, provision of basic services, insurgency, and secessionist move- ordination. Official responses to armed more effective democratic representa- ments are rooted in extreme poverty, violence tend to be fragmented among tion, and the quality of governance social inequality, and ethnic tensions bureaucracies and the central and state (Christensen and Lægreid, 2007). The that are neither adequately recognized governments, with limited collabora- same ideas should underpin counter- nor managed effectively. Since terror- tion among institutions, or between insurgency responses to sub-state ism, insurgency, and violent crimes government and non-government violence, violence by youth gangs, and share underlying causes, it makes actors. The government appears will- neighbourhood violence (Kilcullen, sense to tackle the problems together. ing to increase spending on armed 2010). Tackling armed violence also How India faces its challenges violence, but much slower to develop requires systematic intervention by from armed violence remains unclear. a coherent policy to deal with the NGOs, an area in which India excels. Although some problems discussed problem, addressing the full spectrum That India needs to spend more on here are unique or exceptionally severe of armed violence and embracing the addressing armed violence is obvious,

10 Small Arms Survey Issue Brief Number 1 September 2011 but it also needs to consider how it AFP (Agence France-Presse). 2011. ‘India Geneva: Geneva Declaration Secretariat. spends in response to these problems. to Pull 10,000 Troops from Kashmir.’ Gunaratna, Rohan and Arabinda Acharya. 13 February. 2007 Interventions must be based on sound . ‘Terrorist Finance and the Criminal AI (Amnesty International). 2009. Amnesty Underground.’ In Michael Innes, ed. evidence and judicious use of cost– International Report 2009: State of the Denial of Sanctuary, Understanding Terror- benefit analysis to measure and assess World’s Human Rights. London: Amnesty ist Safe Havens. Westport, Connecticut: the benefits associated with the preven- International. Praeger Publishers. tion and reduction of armed violence. The Armed Forces (Special Powers) Act Hoffman, Bruce. 2006. Inside Terrorism, revised. The central government is the most (AFSPA). 1958. New Delhi: Ministry of New York: Columbia University Press. natural leader of such integrated re- Home Affairs. Homeland Security Research. 2009. ‘India’s BBC (British Broadcasting Corporation). Homeland Security Budget and Require- sponses. But coordinated institutional 2011. ‘India “to Cut Kashmir Troops by ments for 2009/10.’ Washington, DC, cooperation—among ministries, between a Quarter”.’ 14 January. 22 July. the central and state governments, Behera, Laxman K. 2010. ‘Budgeting for IISS (International Institute for Strategic and between government and private India’s Defence: An Analysis of Defence Studies). 2010. Military Balance 2010: The organizations—is only now beginning Budget 2010-11 and the Likely Impact of Annual Assessment of Global Military Capa- to occur. the 13th Finance Commission on Future bilities and Defence Economics. London: IISS. Defence Spending.’ Institute for Defence The Independent (London). 2009. ‘1,500 Farmers Studies and Analyses (IDSA) Comment, Commit Mass Suicide in India.’ 15 April. 3 March. India Today. 2010. ‘Full Text of the Budget Notes Bhaumik, Subir. 2004. ‘Ethnicity, Identity and Speech.’ 16 February. This Issue Brief was written by Arabinda Religion: Separatist Movements in India’s Indian Express. 2009. ‘Naxalism Gravest Inter- Northeast.’ In Satu P. Limaye, Mohan nal Security Threat: PM.’ 15 September. Acharya and Aaron Karp. Sonal Marwah Malik, and Robert G. Wirsing, eds. Kilcullen, David. 2010. Counterinsurgency. updated data and prepared the tables on Religious Radicalism and Security in South Oxford: Oxford University Press. international homicide rankings. Anjali Asia. Honolulu: Asia Pacific Center for Lal, Rollie. 2007. ‘South Asian Organized Crime Krishnan, National University of Singa- Security Studies. and Linkages to Terrorist Networks.’ In pore, assisted with research. —. 2007. Insurgencies in India’s Northeast: Kimberley L. Thachuk, ed. Transnational Conflict, Co-option and Change. Working Threats: Smuggling and Trafficking in 1 For a detailed discussion of data sources Paper. Washington, DC: East–West Center. Arms, Drugs, and Human Life. Westport, and armed violence monitoring systems Chakma, Bhumitra. 2009. ‘South Asia’s Real- Connecticut: Praeger Publishers. see Gilgen and Tracey, 2011. ist Fascination and its Alternatives.’ Magioncalda, William. 2010. ‘A Modern 2 ‘Encounter deaths’ are killings by the Contemporary Security Policy, Vol. 30, No. 3. Insurgency: India’s Evolving Naxalite police forces, whereas ‘extrajudicial kill- December, pp. 395–420. Problem.’ South Asia Monitor, No. 140. ings’ tend to mean paramilitary actions. Chari, P. R. 2006. ‘Sources of New Delhi’s Washington, DC: Center for Strategic 3 Based on discussions with Akanksha Kashmir Policy.’ In Wahenguru Pal Singh and International Studies. Mehta, Research Analyst, International Sidhu, Bushra Asif, and Cyrus Samii, eds. MHA (Ministry of Home Affairs). 2001. India Centre for Political Violence and Terrorism Kashmir New Voices, New Approaches. at a Glance—Rural–Urban Distribution. Research, Singapore, on the basis of her London: Lynne Rienner. New Delhi: Office of the Registrar General interview with surrendered militants in Christensen, Tom and Per Lægreid. 2007. ‘The and Census Commissioner, MHA. Kashmir, 26–29 April 2010. Whole-of-Government Approach to Public —. 2009. Government of India Ministry of Home Sector Reform.’ Public Administration Review, Affairs Outcome Budget 2008–09. New Delhi: Vol. 67, No. 6. November–December, MHA, p. 116. pp. 1059–66. —. 2010a. Annual Report 2009–2010. New Delhi: Bibliography The Economist. 2010. ‘Kashmir’s Troubles: MHA. Acharya, Amitav. 2004. Age of Fear: Power Shaking the Mountains.’ 29 December, —. 2010b. Government of India Ministry of Home versus Principle in the War on Terror. New pp. 38–39. Affairs Outcome Budget 2010–2011. New Delhi: Rupa. Eddleston, Michael and Flemming Konradsen. Delhi: MHA, p. 173. Acharya, Amitav, Subrat K. Singhdeo, and 2007. ‘Commentary: Time for a Re-assess- Mishra, Pankaj. 2010. ‘Kashmir: The World’s M. Rajaretnam. 2010. Human Security: From ment of the Incidence of Intentional and Most Dangerous Place.’ New York Review Concept to Practice: Case Studies from Northeast Unintentional Injury in India and South of Books, 4 March. India and Orissa. Singapore: World Scientific. East Asia.’ International Journal of Epidemi- MoD (Ministry of Defence). 2011. Ministry of Acharya, Arabinda. 2006. ‘Terrorism and ology, Vol. 36, pp. 208–11. Defence Annual Report Year 2009–10; Year Regional Security: Deconstructing the Geneva Declaration Secretariat (GD Secre- 2008–09; Year 2007–08; Year 2006–07; Year Global War on Terror in Asia.’ In Global tariat). 2007. The Global Burden of Armed 2005–06; Year 2004–05. New Delhi: MoD. Terrorism: Genesis, Implications, Remedial Violence. Geneva: GD Secretariat. Muggah, Rob. 2009. Security and Post-Conflict and Counter Measures. Islamabad: Institute Gill, Kanwar Pal Singh. 1997. The Knights of Reconstruction: Dealing with Fighters in the of Regional Studies. Falsehood. New Delhi: South Asia Terror- Aftermath of War. New York: Routledge. — and Amitav Acharya. 2002. ‘Human Secu- ism Portal. Nagraj, Karkada. 2008. ‘Farmers Suicide in rity in Asia: Conceptual Ambiguities and Gilgen, Elisabeth and Lauren Tracey. 2011. India: Magnitudes, Trends and Spatial Common Understandings.’ Man and Contributing Evidence to Programming: Patterns.’ Madras: Madras Institute of Development, Vol. XXIV, No. 4, December. Armed Violence Monitoring Systems. Development Studies, March.

www.india-ava.org 11 NATO (North Atlantic Treaty Organization). RUPE (Research Unit for Political Economy). USDoE (US Department of Energy, Energy 2010. ‘International Security Assistance 2005. Aspects of India’s Economy, Nos. 39 Information Administration). 2010. Force and Afghan National Army Strength and 40. Mumbai: RUPE, June. Country Analysis Briefs: India. Washing- and Lay Down.’ 1 February. Saikia, Jaideep. 2001. Contours: Essays on Security ton, DC: US Department of Energy, NCRB (National Crime Records Bureau). 2010. and Strategy. Guwahati: Sagittarius Print. Energy Information Administration. Crime in India—2008. New Delhi: NCRB Sarkar, Sumita and Arvind Tiwary. 2001. USDoS (US Department of State). 2009. Coun- (MHA). ‘Combating Organised Crime: A Case try Reports on Terrorism 2008. Washington, —. 2011a. Crime in India—2009. New Delhi: Study of Mumbai City.’ Faultlines, Vol.12. DC: USDoS, Office of the Coordinator for NCRB (MHA). SATP (South Asia Terrorism Portal). 2009a. Counterterrorism. —.2011b. Accidental Deaths and Suicides in ‘Fatalities in Terrorist Violence in India’s Venkataramani, R. 2010. Naxalites’ Urban Push: India—2009. New Delhi: NCRB (MHA). Northeast.’ New Delhi: SATP. Will They Succeed? Institute for Peace and Nepram, Binalakshmi. 2009. The Emergence —. 2009b. ‘Fatalities—Left-Wing Extremism.’ Conflict Research Issue Brief, No. 138, of the Manipuri Women Gun Survivor’s New Delhi: SATP. 10 February. Network. New Delhi: Control Arms —. 2010. ‘India Fatalities 2010.’ New Delhi: Venugopalan, M. G. 2002. ‘Police and Political Foundation of India (CAFI). SATP. Parties: The Politician–Police–Criminal Patel, Raj. 2007. Stuffed and Starved. London: Sen, Amartya. 2008. ‘Violence, Identity and Nexus.’ In P. J. Alexander, ed. Policing in Portobello Books. Poverty.’ Journal of Peace Research, Vol. 45, Indian in the New Millennium. Mumbai: Prakash, Ved. 2009. . New No. 5, pp. 5–15. Allied Publishers. Delhi: Kalpaz Publications. Somaiah, P. E. 2002. ‘Police Image.’ In P. J. Verma, Arvind. 1999. ‘Cultural Roots of Police PTI (Press Trust of India). 2009. ‘Ministry of Alexander, ed. Policing in India in the New Corruption in India.’ In Policing: An Home Affairs Gets Whopping 33% Hike.’ Millennium. Mumbai: Allied Publishers. International Journal of Police Strategies & 6 July. UNODC ( Office on Drugs Management, Vol. 22, No. 3, pp. 264–79. Ramana, P. V. 2009. ‘Maoists’ Tactical United and Crime). 2011. Homicide Statistics. Vinayak, Ramesh. 1997. ‘Victims of Sudden Front (TUF) and Urban Movement.’ Defence and Strategic Analysis, 10 July.

About the India Armed Violence Credits Assessment Content editor: Diana Rodriguez The India Armed Violence Assessment (IAVA) promotes Copy-editor: Catherine Robinson research and supports Indian social-science research Proof reader: Donald Strachan communities dedicated to studying the causes and conse- Design and layout: Richard Jones ([email protected]) quences of armed violence. Developed in coordination with Indian partners, it explores wide-ranging issues related to the instruments, actors, and enabling institutions that shape Contact details security. The IAVA intends to catalyse evidence-based Sonal Marwah, IAVA Project Coordinator: debate in India and facilitate Indian contributions to global [email protected] policy and programming on related issues. The project is Aaron Karp, Senior Consultant, Small Arms Survey: supported by the Small Arms Survey. [email protected] IAVA Issue Briefs review the state of knowledge on key India Armed Violence Assessment themes associated with armed violence. Commissioned EP-16/17 Chandragupta Marg, , by the Small Arms Survey, Issue Briefs summarize major New Delhi-110021, India findings and insights on issues related to conflict and Small Arms Survey crime-related violence, perpetrators and victims, preven- 47 Avenue Blanc, 1202 Geneva, Switzerland tion and reduction, and strategies to contain violence. They stress data-based research findings on the scale, forms, t +41 22 908 5777 f +41 22 732 2738 and severity of armed conflict, contributing elements, and the impact of policy responses.

All IAVA Issue Briefs are available in English and Hindi. They can be downloaded at www.india-ava.org. Print copies are available from the Small Arms Survey. A project of the Small Arms Survey

12 SmallSudan ArmsIssue BriefSurvey Number Issue Brief 6 April Number 2007 1 September 2011