1. Introduction
When India gained independence in 1947, its new rulers chose to build a constitutional, democratic, secular state. At the time, this was a startling choice.Footnote 1 India was a large, poor, ethnically varied collection of linguistically distinct provinces, many of which had been uneasily corralled into the new country. Independence had been accompanied by the catastrophically violent carving of the British colony into Hindu-majority India on the one hand and Muslim-majority Pakistan on the other.Footnote 2
Despite this traumatic start, India’s secular, democratic Constitution has endured and, under its umbrella, so have civilian government, judicial independence, and regular national and state elections. However, the country’s legacy of sectarian violence has also endured. Riots and pogroms have recurred in India over the past six and a half decades.Footnote 3 In addition to low-grade sectarian violence, India has seen several large-scale, systematic attacks against religious minorities which have been tolerated, even aided, by the state. Hundreds of people have been killed, hundreds injured, and thousands displaced in such violence. This article examines government reparation after four of the worst episodes of mass sectarian violence in independent India.
Many scholars have studied reparations in countries transitioning from authoritarianism to democracy. Be it in Latin America, Africa, Asia, or Eastern Europe, new regimes in transitional democracies have had to decide how to reckon with the abuses of the old order, including whether and how to make amends to victims of state violence. But discriminatory, planned, systematic persecution takes place—and even prospers—in democratic regimes. India is far from being the only country whose recent history includes targeted violence on a large scale against people belonging to religious or ethnic minority groups. Post-colonial countries including, inter alia, Bangladesh, Pakistan, the Philippines, and Kenya have all experienced such violence. This article analyses reparation after mass violence in a stable, multi-ethnic democracy, and reflects how and why such reparation falls short.
I discuss my methodology in Section 1 below. In Section 2 of the article, I consider international standards on reparations, focusing in particular on the United Nations Basic Principles and Guidelines on the Right to a Remedy and Reparation 2005. In Section 3, I discuss how the Indian state dealt with reparations after mass sectarian violence in Nellie, Assam, in eastern India in 1983; in Delhi, the national capital, in 1984; in Bhagalpur, Bihar in eastern India in 1989; and in the western state of Gujarat in 2002. I then evaluate these efforts in Section 4. I argue that Indian governments have never actually attempted to make reparation to victims of mass violence. While the state has established administrative programmes to help victims after each episode of mass violence, it has failed to acknowledge any legal obligation to do so. The help extended is officially categorized as “ex gratia,” namely as being voluntary rather than required. I argue further that, although reparation is not recognized as the state’s legal duty, it is nevertheless a political expectation so settled as to be considered normative. Accordingly, there is a well-established repertoire of assistive measures that governments have delivered after mass violence against religious minorities. I then highlight the ways in which these administrative aid programmes are unimaginative and meagre. In Section 5, I examine why reparative measures fall short, despite the capacity, resources, and room for manoeuvre available within a politically stable state. I argue that the political opportunity structure for reparations is limited by the fact that sectarian mobilization is embedded within electoral politics in India. Just as sectarian electoral politics led governments to tolerate—and even actively support—mass violence against religious minorities on the four occasions that I examine, it also constricts the constitutional checks and balances that might lead to adequate reparation after violence subsides. I argue further that the opportunities to negotiate better support for victims of particular episodes of mass violence expand over time. Finally, I suggest that a law on reparations would constrain state bias and strengthen opportunities for victims of mass violence to demand the support that they need.
2. Methodology
Aside from sporadic journalistic coverage, publicly available information about how Indian governments have helped survivors of mass violence is relatively scant. This article relies primarily on official records which were obtained using India’s law on the right to information.
In 2005, the Right to Information (RTI) Act granted Indians the right to access a wide swathe of official records without having to justify why they want this information or demonstrate that they are directly affected by it.Footnote 4 I used this recently minted law heavily between 2009 and 2011, while co-ordinating a project to access and analyze official records on mass violence. Over the course of the project, the research team filed 824 applications for information, including, inter alia, records about reparation, with various public authorities.Footnote 5 I drafted the right-to-information applications related to reparation, after researching the relevant laws, policies, and procedures as well as interviewing retired civil servants about how these norms were applied in practice and consulting right-to-information experts about formulating effective applications. My fellow researchers shared the practicalities of sending and tracking applications, following up on the ones that succeeded and appealing those that were refused.Footnote 6
I applied for information about government assistance in the immediate aftermath of violence, as well as longer-term support for victims. Since it was unclear which public authorities handled reparations, right-to-information applications were sent to the Ministry of Home Affairs and the Office of the prime minister within the national government, as well as departments for home affairs within the state governments of Assam, Delhi, Bihar, and Gujarat. The local administration in districts where violence had taken place was also asked for information.Footnote 7
These applications elicited government memos, official announcements, and bureaucrats’ reports related to assisting victims. Governments also disclosed reports by official Commissions of inquiry which had been appointed after each instance of mass violence to examine its causes and consequences. Material seemingly unrelated to reparations, such as criminal complaints and trial records, sometimes contained useful information.
In addition to records that were obtained through right-to-information applications, I also analyze official material already in the public domain, including debates in national and state legislatures and judicial decisions. Alongside material generated by state processes, I draw upon reports by journalists and non-governmental organizations (NGOs) about the four episodes of mass violence discussed below.
My strong focus on state records has inherent limits. I have not interviewed state functionaries or survivors of mass violence. Therefore, I engage primarily with what governments formally declared that they would do, and only very partially with what they delivered on the ground. That said, those formal commitments—what governments have decided is warranted for the people that they failed to protect from systematic, violent persecution—are diagnostic and, therefore, worth analyzing in their own right.
To my knowledge, this article is the most comprehensive analysis of state assistance to victims of mass sectarian violence in India. While other commentators have analyzed government aid after specific instances of sectarian violence, my analysis is unusual in encompassing reparative measures after four of the most serious such occasions, and in having access to official records related to each one.
Before turning to this material in Section 4, I consider international standards on reparation in Section 3 below.
3. Standards, Guidance, and Good Practice
The 1980s and 1990s saw democratic transitions in many countries. States transitioning from authoritarianism to democracy had to decide how to confront the deposed regime’s misdeeds. This difficult predicament prompted responses ranging from impassive silence in Indonesia, to official fact-finding and formal amnesties by governments in Chile and South Africa, to selective prosecution of junta leaders in Argentina, to the very controversial “lustration” of communist-era officials in Poland.
Over time, the challenges of confronting past atrocities came to be articulated as a distinct set of “transitional justice” concerns: uncovering and acknowledging past abuse, punishing abusers, and making reparation to victims. As the tripartite goals of truth, justice, and reparation became established in the international lexicon, scholars debated which of these goals was the most important. Human rights advocates such as Neier argued that criminal prosecution was paramount, while acknowledging that governments may be forced to forgo it in order to maintain the new political order.Footnote 8 In the South African context, Bishop Tutu emphasized that public truth-telling fosters forgiveness. Scholars such as Minow and HaynerFootnote 9 underscore the importance of reparations, arguing that reparative measures promote a new social contract in a way that prosecution and punishment do not.Footnote 10 Most transitional justice experts now agree that all these processes are crucial and mutually reinforcing,Footnote 11 though the scope, sequencing, and priority of each are inevitably shaped by political context and constraints.
These interconnected transitional justice processes have gathered normative strength over time. International law recognizes truth, justice, and reparation as duties owed by states to people who have suffered grave violations of human rights. State obligations in this regard are codified in the Basic Principles and Guidelines on the Right to a Remedy and Reparation for Victims of Gross Violations of International Human Rights Law and Serious Violations of International Humanitarian LawFootnote 12 (“Basic Principles”). I focus below on the provisions specific to reparations in the Basic Principles.
The Basic Principles, which were adopted by the United Nations General Assembly in 2005, claim to collate existing obligations under international law rather than imposing fresh duties on states.Footnote 13 Foundational human rights texts, including the Universal Declaration of Human RightsFootnote 14 and the International Covenant on Civil and Political Rights (“ICCPR”),Footnote 15 do indeed recognize the right to an effective remedy for violations of human rights. The UN Human Rights Committee has clarified that this general right to an effective domestic remedy encompasses the right to reparation. Absent reparations, the Committee argues, the state has not discharged its duty towards individuals whose ICCPR rights have been violated.Footnote 16 The Convention against Torture obligates states to compensate victims of torture, and provide for “as full rehabilitation as possible.”Footnote 17 A right to reparation is also recognized in the Convention on the Rights of the Child.Footnote 18
This overarching right, as well as the corresponding duties of the state, is disaggregated in some detail in the Basic Principles. They prescribe “adequate, effective and prompt reparation” by the state to individuals and communities who suffer a gross violation of human rights law or serious violation of international law.Footnote 19 The Principles note that reparation is owed for violations flowing from the state’s acts or omissions, and must be made by the state itself.Footnote 20 The state’s efforts must be proportionate to the gravity of the violation and the harm suffered by victims.Footnote 21 Further, reparative measures should encompass restitution, compensation, rehabilitation, satisfaction, and guarantees of non-repetition, namely long-term, systemic reforms to ensure violence will not recur.Footnote 22
This capacious understanding of reparation extends considerably further than what a reparations programme is likely or able to deliver in practice. As the Special Rapporteur on the promotion of truth, justice, reparation, and guarantees of non-recurrence has pointed out, operationally, systemic reform, truth-telling, and criminal justice would fall outside the ambit of most official attempts at reparation.Footnote 23 He distinguishes between measures that directly benefit victims and measures that have reparative effects but do not focus on the victim, such as criminal prosecution of perpetrators or institutional reform to address bias amongst the police.Footnote 24 I adopt the same distinction. I am examining the direct support offered by Indian governments to victims of mass violence. Accordingly, I do not delve into the Basic Principles’ provisions on systemic reform (“guarantees of non-repetition”) or those aspects of satisfaction that would, in a practical sense, fall under the rubric of criminal justice or fact-finding.
Amongst the reparative measures that would directly benefit victims, the Principles include, first, restitution—trying to restore the victim to her position before she suffered harm. Restitution includes “restoration of liberty, enjoyment of human rights, identity, family life, and citizenship, return to one’s place of residence, restoration of employment and return of property.”Footnote 25 If restitution is unviable, the Principles prescribe compensation for (1) physical or mental harm; (2) lost opportunities, including employment, education, and social benefits; (3) material damages and lost earnings; (4) moral damage; and (5) costs required for legal, medical, and psycho-social services.Footnote 26 Rehabilitation is defined to mean medical and psychological care as well as legal and social services.Footnote 27 Under the rubric of satisfaction, the Principles include, inter alia, “effective measures” to stop ongoing violations; searching for the disappeared and helping people recover the remains of loved ones; public acknowledgement and apology; commemorations and tributes to victims; and incorporating a fair account of the violations in public education.Footnote 28
These tenets on reparation—the international state of the art—have been widely neglected at the national level.Footnote 29 Instead of reparation, many governments have offered development initiatives or welfare benefits to victims of grave human rights violations. Some governments have cited resource constraints as a reason for withholding reparations in whole or part. Numerous governments have implemented reparation programmes, but attended only tenuously to the standards in the Basic Principles. Nevertheless, these experiences across various countries have yielded practical guidance on how to design reparations that would reflect the Basic Principles more fully, and avoid harmful trade-offs.
Reparation measures can be individual or collective, monetary or non-monetary, juridical or administrative.Footnote 30 The particular mix of measures should be tailored to the types of harm reparations are attempting to allay and the people they aim to reach.Footnote 31 For example, where the number of recipients is relatively large, administrative reparations programmes that distribute standardized aid for different categories of harm are advocated by experts such as De Greiff, and Issacharoff and Mansfield. Compensating a large number of victims to the juridical restitutio in integrum standard laid down by the Basic Principles might not be possible. These authors also warn that highly individuated reparations have the potential to divide victims and foster grievances.Footnote 32 In addition, administrative programmes spare beneficiaries the burden of litigating to claim their due, and allow for a mix of measures that mitigate different facets of the harm suffered.Footnote 33 By contrast, where reparations are owed to small numbers of victims, individual determination of each victim’s losses would be more viable practically and politically.
Experts emphasize certain procedural and substantive safeguards for reparation programmes. The Special Rapporteur has stressed that a state must expressly acknowledge its responsibility to victims if measures intended to help them are properly to be considered reparation. If victims are not recognized as rights-bearing, and the state as correspondingly duty-bound, the Special Rapporteur argues that supposed reparation programmes are more accurately “mechanisms to distribute indemnification benefits.”Footnote 34 Some scholars argue that such indemnification is unlikely to comfort victims and educate society at large.Footnote 35
Experience indicates that reparation programmes benefit greatly from being participatory.Footnote 36 Reparation efforts have often excluded victims from disadvantaged groups or disregarded harm that is disproportionately borne by the less powerful. For example, women and girls who have suffered sexual violence might be ignored entirely when designing reparative measures. Measures in response to mass atrocity are likely to be more tailored to grassroots needs and less exclusionary if victims are consulted when designing a reparation programme, distributing its benefits and monitoring its efficacy.
Just as large reparation programmes should strive to be inclusive, they must treat beneficiaries equally.Footnote 37 Equal treatment in this context should mean ensuring that groups with particular vulnerabilities, such as disability or a heightened risk of violence, can access programme benefits. Experts note that effective reparation programmes should also try to circumvent discrimination within affected communities that might deprive many victims of their due.Footnote 38 For example, the Special Rapporteur notes how some programmes depart from traditional family and inheritance law to ensure that women can access benefits, while others have implemented measures to give women greater financial autonomy, such as shares in micro-credit programmes.Footnote 39
The Basic Principles require that state reparation be comprehensive and fully compensatory. In practice, large reparation programmes following mass atrocities rarely reach this high threshold.Footnote 40 But, while the restitutio in integrum standard is a difficult one to meet, reparations must be commensurate to the harm suffered by victims.Footnote 41 Further, whatever the balance between individual and collective, monetary and non-monetary measures in a reparations programme might be, experts emphasize that these different measures should cohere internally. Reparations should also dovetail effectively with parallel state processes such as truth Commissions, fact-finding, or criminal proceedings.Footnote 42 Moon argues that governments tend to use reparation efforts as an alibi for neglecting the state’s past abuses and shielding the culprits.Footnote 43 Cynical expediency of this sort is likely to undermine reparation efforts in the same way as reluctance to acknowledge responsibility by the government does.
Bearing in mind this cautionary guidance and the international legal standards on reparation, I turn below to reparation efforts after four episodes of mass sectarian violence over a 20-year period in India.
4. Mass Violence and Reparation: Four Episodes
In this section, I outline briefly the context and nature of attacks against religious minorities in Nellie, Assam, in eastern India in 1983; in Delhi, the national capital, in 1984; in Bhagalpur, Bihar in eastern India in 1989; and in the western state of Gujarat in 2002. I also discuss how national and state governments assisted victims after each of these events.
4.1 Nellie, 1983
At least 1,800 Bengali Muslims were murdered over the course of a few hours on 18 February 1983 in the small town of Nellie in the state of Assam—unofficial estimates put the number of people killed at 3,000. While Muslims in India had often been targets of violence, the concentrated slaughter in Nellie was the single worst episode of mass violence in India since independence.Footnote 44
In the year preceding the Nellie massacre, Assam had been riven by state-wide agitation against illegal immigration from neighbouring Bangladesh.Footnote 45 Led by the All Assam Students Union (AASU), the anti-foreigner movement demanded that Bangladeshi immigrants be removed from electoral rolls and evicted. AASU leaders argued, fairly credibly, that the ruling Congress party had registered illegal immigrants as voters, gaining a vote bank in exchange for this strategic benefaction. After declaring an emergency in Assam in response to violent clashes and civil disobedience, the national government, also led at the time by the Congress party, ill-advisedly announced state elections in 1983, in the expectation that it might win. Impending elections heightened hostility towards Bengali Muslims, many of whom had lived in Assam for generations, some of whom had indeed crossed the Bangladesh border more recently to establish a precarious existence as religious, ethnic, and linguistic minorities in Assam.
On 18 February 1983, Bengali Muslims in Nellie were attacked by a 1,000-strong, armed throng of people from nearby villages.Footnote 46 Seven and a half hours of uninterrupted butchery left over 2,000 people dead and many more injured. The attackers had been amassing around Nellie for three days, wielding weapons, beating drums, shouting slogans, and closing off escape routes.Footnote 47 Residents had asked the police for help as the belligerent crowd gathered, but their requests—including the specific fact that over a thousand armed men were surrounding Bengali Muslim settlements—were ignored.Footnote 48
In 1985, the anti-foreigner movement negotiated a peace deal with the central government.Footnote 49 Fresh elections in Assam brought movement leaders to power, some of whom had allegedly participated in the Nellie massacre and many of whom had led other attacks before the 1983 elections. Under this leadership, the state government dropped all criminal charges—however serious—related to election violence across Assam. No one has been prosecuted for the Nellie massacre since 1983.Footnote 50
Soon after the Nellie massacre, local officials arranged emergency relief as prescribed in the Assam Relief Manual.Footnote 51 The colonial-era (though periodically updated) Relief ManualFootnote 52 focuses primarily on relief following floods, directing district officials to adapt flood-relief guidelines to other upheavalsFootnote 53 and to “give relief to … deserving cases.”Footnote 54
The district Commissioner is charged by the Manual with supervising “relief centres,” organizing emergency medical relief, setting up temporary medical facilities in relief camps if necessary,Footnote 55 and making special provisions for the “infirm, destitute, orphans, children, expectant/nursing mothers.”Footnote 56 Beyond this, the Commissioner has the discretion to establish relief works and distribute financial aid in the form of agricultural inputs, soft loans, and house repair grants.Footnote 57 The Manual instructs the district Superintendent of Police to prevent and investigate crime after a calamity, rescue people and transfer them to relief camps, and help people to re-establish contact with their families.Footnote 58
By 2009, when I applied for information to be released, many records about how far the local administration implemented relief guidelines had been destroyed. The Superintendent of Police at the time recorded that Bengali Muslims had sheltered in a local school on the night of 18 February before being moved to a relief camp guarded by paramilitary forces.Footnote 59 Nellie had limited security cover for some days after the massacre, but the armed forces were in short supply. The district Commissioner noted that “minority pockets in remote areas were leaving their villages out of fear,” but the area could not be offered more protection because “most of the bridges were burnt and roads damaged by agitationists.”Footnote 60
In the months following the massacre, local police registered criminal complaints, termed first information reports (FIRs) under Indian law, from survivors in Nellie. These FIRs would have been little use in facilitating criminal investigation. Rather than details of the particular crime at issue, most contained vague, pro forma text, stating that “a number of miscreants” had attacked Nellie. But these FIRs recorded property lost by the complainant, suggesting that the district administration was not intending to investigate the assaults and murders committed in Nellie, but was planning to disburse aid and deputing the police to gather information in this regard.Footnote 61 The property logged in these FIRs suggests a community living in considerable destitution, with complainants reporting losses—a couple of chickens, a goat, a single brass container—that would loom large only for the poor.Footnote 62
After relief camps were closed, the Assam government allowed people to avail of emergency services after they returned to their homes, “till the harvesting of the next crop.”Footnote 63 Emergency relief was followed by financial aid for people harmed during the three very violent months preceding elections in Assam. The Assam government announced payments of Rs. 5,000 to the families of people who had been killed and Rs. 5,000 for homes destroyed by violence; free seeds for paddy farmers; small subsidies and loans for replacing livestock—“lost bullocks and milch cattle”; restarting businesses and rebuilding shops.Footnote 64
4.2 Delhi, 1984
Eighteen months after the Nellie massacre, at least 2,733 Sikhs were killed in Delhi by mobs led by the ruling Congress party—unofficial estimates put the death toll at 4,000.Footnote 65
In the early 1980s, separatist groups demanding independence for the Sikh-majority state of Punjab became increasingly violent, murdering senior officials and politicians as well as members of the public. The government responded with severe counter-terrorism legislation as well as no-holds-barred policing, resorting routinely to torture and extrajudicial killing.Footnote 66 On 4 June 1984, the prime minister, Indira Gandhi, ordered the army to attack separatist leaders hiding in the Golden Temple in Amritsar, one of Sikhism’s holiest shrines. Several innocent worshippers were killed in the gun battle that ensued. Forty-one other gurudwaras were also raided by the army.
On 30 October 1984, Mrs Gandhi was shot at close range by her Sikh bodyguards. The day after the prime minister’s assassination, Sikh neighbourhoods across Delhi were systematically attacked by armed gangs whose ranks were swelled by Congress party functionaries, including some senior politicians. Sikh families were killed and injured, Sikh women sexually assaulted, and Sikh homes and businesses set on fire. The Delhi police pointedly did not intervene. Indira Gandhi’s son, Rajiv Gandhi, had stepped into the prime ministerial role shortly after her death. Despite being urged to, he delayed deploying the armed forces to control violence, tolerating—and perhaps tacitly endorsing—the bloodshed by his colleagues in the Congress party.
In the years since, the Delhi government-appointed multiple ad hoc Commissions and committees to apportion responsibility for this episode of mass violence.Footnote 67 Some of these bodies were highly partisan, scapegoating selected civil servants while protecting Congress leaders.Footnote 68 Others made fairly modest recommendations, but were neglected by the Delhi government. The crimes committed during October 1984 were poorly investigated and incompetently tried, if at all; some criminal cases were re-opened due to concerted pressure from civil society groups, and trials sputter on even today.Footnote 69 Some of the key culprits have held ministerial positions under Congress-led governments.Footnote 70
On 3 November 1984, even as the police were ignoring violence against Sikhs, the Delhi government set up 12 relief camps guarded by the army. Three days later, it announced monetary aid for victims, but simultaneously began evacuating people from relief camps.Footnote 71 Victim groups challenged camp closure in the Delhi High Court. Their homes had been burnt and looted, and—as Congress party workers menaced them even in camps with police cover—peace was clearly still fragile. High Court intervention kept the camps open for another ten days, but they remained deprived of basic supplies, including food, water, and toilets.Footnote 72
The government’s aid package offered Rs. 10,000 to families of the deadFootnote 73 —double the amount announced by the Assam government for similar loss the previous year but still shockingly low. Aid to the disabled and injured, as well as to people whose homes were damaged was also extremely low.Footnote 74 A year later, the government doubled aid for families of the dead, and marginally increased aid to the injured.Footnote 75
The government initially dispensed with elaborate paperwork when disbursing money. A survivor could secure aid if an eyewitness corroborated her account of loss suffered.Footnote 76 When the government increased aid in later years, however, this initial flexibility was replaced with requirements that unfairly excluded many people. Applicants for aid had to demonstrate that they had reported their losses to the police but, since the police had often refused to register complaints in 1984, this was a difficult requirement to fulfil and an unreasonable one to impose.Footnote 77 Just as carelessly, the government restricted assistance for commercial property losses to people who lacked property insurance, but did not register that insurance companies—most of them state-owned—were rejecting claims by victims on the grounds that insurance policies did not cover sectarian violence.
Sikh organizations lobbied the government as well as Commissions of inquiry to resolve these roadblocks and improve support for victims. Their efforts yielded three further increases in financial aid. The Delhi government also sold apartments to women widowed in 1984 at subsidized rates and offered them skills training.Footnote 78
The most significant increase in cash assistance came after Bhajan Kaur, whose husband was killed in 1984, petitioned the Delhi High Court and argued that government aid was too low. The High Court agreed with Mrs Kaur. After surveying court-ordered compensation in a range of cases in which the state had committed human rights abuses, the High Court ordered the Delhi government to give Bhajan Kaur, and other people in her situation, Rs. 300,000 minus the aid already given.Footnote 79 Significantly, the court stated that the government had a duty to prevent “riots” and, since it had failed to do so, was responsible for the loss suffered by victims of violence in 1984.Footnote 80 In reaching this conclusion, the court drew upon findings against the Delhi police by the government-appointed R.N. Misra Commission of Inquiry, which detailed “the passivity, callousness and indifference” of the police.Footnote 81 The court also noted that the government could not plead ignorance of violence so public, nor could it blame poor resources, since the events of 1984 took place in the national capital rather than somewhere remote and difficult to access.Footnote 82
In 2006, the national government significantly increased financial assistance for survivors of mass violence in 1984.Footnote 83 It raised cash aid to family members of the dead to reflect the Delhi High Court’s Bhajan Kaur decision. The disabled, the injured, and property owners who suffered losses also received more generous sums of money. In addition, the government instituted small pensions and modest affirmative action in government employment for widows and children of men killed in 1984.
This jump in aid, markedly more than previous increments, owed much to political shifts nationally and within the Congress party. After a decade out of power, the Congress-led United Progressive Alliance had unexpectedly defeated the Hindu-majoritarian Bharatiya Janata Party (BJP) in national elections in 2004. At the time, mass violence targeting Muslims in BJP-controlled Gujarat in 2002 and the BJP’s lack of remorse were recent and vivid. The Congress tended to contrast its commitment to secularism with the BJP’s disdain for religious minorities, but the events of 1984 provided political opponents with a quick—and apposite—rebuttal to these assertions. In 2005, the last of several official inquiries into mass violence in 1984—the Nanavati Commission—finally concluded and placed its report before the Indian Parliament. While careful to exonerate Congress leaders and deny government complicity in violence, the report did recommend enhanced economic assistance for victims.Footnote 84
Dramatically increasing aid was a way for the Congress to reduce its vulnerability to charges of persecuting minorities. In addition, a few senior members of the Congress might genuinely have wanted to express regret to the Sikh community. The party was headed by Rajiv Gandhi’s widow, Sonia Gandhi, who later advocated a law to prevent sectarian violence. The prime minister, Manmohan Singh, also a long-standing member of the Congress, was Sikh himself. In 2006, when the national government placed its response to the Nanavati Commission of inquiry before the national Parliament, the prime minister apologized for the state’s failures in 1984.Footnote 85
4.3 Bhagalpur, 1989
In October and November 1989, approximately 1,000 people—most of them Muslim—were killed, several hundred injured, and about 50,000 displaced in long, drawn-out mass violence in Bhagalpur, Bihar.Footnote 86
The months preceding this conflagration had seen some friction between Hindu and Muslim groups in Bhagalpur, as the BJP’s national campaign to replace the sixteenth-century Babri mosque in the northern town of Ayodhya with a Hindu temple gained momentum. The mosque stood on the site where many believed that the Hindu god Ram had been born, and became a potent symbol of the Hindu-majoritarian narrative of a halcyon Hindu India humiliated by centuries of Muslim invasion. On 24 October, a scuffle broke out when a procession carrying bricks for the Ayodhya temple marched through a Muslim neighbourhood in Bhagalpur. The marchers alleged that a bomb had been lobbed at them, and the police fired upon the gathered Muslims. A government-appointed inquiry later found that the explosion had likely been firecrackers.Footnote 87
A few days later, armed gangs of hundreds and even thousands attacked Muslim neighbourhoods and spread outwards into Muslim villages surrounding the town.Footnote 88 Bhagalpur’s police force was strikingly partisan. Some policemen joined Hindu mobs, and most refused to aid Muslims under attack. As many Muslims as Hindus were arrested by the police, even though the overwhelming majority of attacks were against Muslims.Footnote 89 The police shot readily at Muslim mobs but hardly ever at Hindu ones.Footnote 90 In one of several massacres during the two weeks of intense violence, the Muslims of Chanderi village were assured by the army and the police that they would be protected, but were left instead to be attacked by Hindus from their own village; 60 Muslims were butchered and drowned in the local river.Footnote 91
By the mid-1980s, the Bihar government no longer treated its Famine Relief Code as a one-size-fit-all policy, and had developed more specialized guidelines for responding to sectarian and terrorist violence.Footnote 92 These guidelines prescribed monetary aid for victims and also outlined relief measures as well as post-crisis support over the short and medium term. While these guidelines were more nuanced than the Relief Code, the amount of money they allocated was minimal.
In December 1989, the Bihar government announced an aid package for victims of mass violence that offered considerably more cash assistance to families of the dead than the sums prescribed in its pre-existing guidelines.Footnote 93 The next of kin of people who were killed were given Rs. 100,000; by comparison, the Delhi government had given Rs. 10,000 for the equivalent loss five years previously. In comparison to payments for death, rates of aid for disability, injury, and property damage remained very low; the designated payments of Rs. 5,000 to the permanently disabled and Rs. 1,000 to the grievously injured would barely have made a dent in medical expenses, even in 1989.
Meagre assistance for damaged property was also consequential. Some Muslims in Bhagalpur were relatively well off as compared with Muslims in other parts of Bihar. Many were concentrated in Bhagalpur’s local textile industry.Footnote 94 An assessment by the Bhagalpur district Commissioner after mass violence noted “systematic attempts ... to weaken the other community [Muslims] economically.”Footnote 95 Muslim-owned shops were looted and burnt, as were looms and yarn belonging to Muslim weavers.Footnote 96 In rural areas, the tractors, cattle, and tools of Muslim farmers were stolen, their wells filled with stones, and, in some instances, their fields harvested and crops taken.Footnote 97 Unable to rebuild livelihoods, many Muslims migrated from Bhagalpur in the years that followed, joining the insecure ranks of casual construction labourers in north India, and many sank into poverty.
Government officials seemed to have reflected on the details of the recent violence when designing the aid package for victims. So extensive were the murders that many of the dead were buried in mass graves or thrown into wells, rivers, and ponds.Footnote 98 Recognizing that local water sources had been rendered unusable, and perhaps also recognizing that these wells and ponds were now distressing sites for survivors, the district administration installed water pumps in many villages.Footnote 99 Given the drowning and mass burials of Muslims, the administration formally recognized as dead those individuals who had been missing for two months, rather than seven years as would ordinarily be required, thereby allowing bereaved families to access cash assistance.Footnote 100 But official sensitivity to the particularities of recent violence was distinctly partial. Families of people killed by the police could not claim any assistance.Footnote 101 However, the assumption that individuals shot by the police were culpably belligerent was problematic, given that policing during mass violence had been extremely biased.
Twenty years after mass violence in Bhagalpur, the national government announced fresh aid to victims, comparable to the aid package offered to survivors of mass violence in 1984.Footnote 102 New Delhi had been lobbied by the Bihar government, as the ruling party in Bihar, the Janata Dal (United), courted Muslim votes and sought to assuage the sense of injustice felt by survivors in Bhagalpur. Political calculation also shaped the national government’s largesse. The new aid package was announced strategically close to impending elections for the Bihar state legislature, where the UPA—which held power at the national level at the time—was hoping to make electoral gains.Footnote 103
4.4 Gujarat, 2002
As in Bhagalpur in 1989, the Hindu-majoritarian movement was implicated in violence that led to over 2,000 deaths in the western state of Gujarat in February 2002. Formal politics in Gujarat was dominated by the BJP at the time, and its Hindu-Right affiliates such as the Rashtriya Swamamsevak Sangh (RSS) and the Bajrang Dal had strong grassroots support in the state.Footnote 104 On 27 February 2002, as a train ferrying members of the RSS and Bajrang Dal from Ayodhya after a ceremony at the site of the Babri mosque stood at Godhra train station in Gujarat, two carriages were set on fire. Fifty-eight Hindu pilgrims were burnt to death.Footnote 105
Hindu-Right politicians responded to these horrific killings by speculating that Muslims in Godhra were collaborating with Pakistan to inflict terror on Gujarat.Footnote 106 In the days that followed, large groups wearing the Hindu-Right colours of saffron and khakhi, armed with swords, explosives, and cooking-gas cylinders attacked Muslims across Gujarat.Footnote 107 Ministers in the government as well as members of the Gujarat legislature participated in these attacks.Footnote 108 The police did not intervene, allowing not just looting sprees, but multiple, hours-long massacres to take place.Footnote 109 In just one example, the home of a Muslim politician, a former Member of Parliament, was surrounded by about 15,000 people and, over seven hours, the 70 people sheltering within were murdered.Footnote 110 Muslim women were sexually assaulted, stripped, and gang-raped in large numbers. After violence subsided, two high-ranking police officials alleged that the Chief Minister Narendra Modi had instructed senior administrators on the day of the Godhra killings to allow violence against Muslims to unfold, and reports suggested that a minister had taken over the police control room in the state capital of Ahmedabad, ensuring attackers had a free run of the city.Footnote 111 The criminal proceedings that followed were concertedly partisan.Footnote 112 In an unprecedented measure, the Indian Supreme Court transferred criminal prosecution related to two particularly serious massacres out of Gujarat; the evident hostility towards Muslim victims displayed by the Gujarat authorities persuaded the court that a credible trial was not possible within the state.Footnote 113 Some years later, after prosecutorial bias undermined other trials related to mass violence, the Supreme Court appointed a special team to re-investigate nine massacres, bypassing Gujarat’s ordinary investigative machinery.Footnote 114
Unofficial estimates suggest that about 200,000 people were displaced from their homes during mass violence in 2002, while official figures put the number at approximately 100,000.Footnote 115 Compounded by its desertion of people under attack, the Gujarat government failed to establish relief camps where the displaced could shelter. NGOs stepped into the breach, only to find the government withholding water, sanitation, security, and cash. A month after camps opened, many still lacked security cover despite continuing violence and had yet to receive financial support.Footnote 116 Organizers challenged government neglect in the Gujarat High Court, but their petition was dismissed as frivolous.Footnote 117
When the government belatedly set up relief camps, it shut them down quickly. Across Gujarat, many Muslims faced threats and economic boycotts from their erstwhile neighbours when they sought to return home. Some were let back into their villages on condition that they recant complaints to the police. Faced with such hostility, many Muslim families migrated instead to “relief colonies” on land leased by religious charities on the outskirts of Gujarati towns.Footnote 118 These make-shift settlements—usually located far from public amenities in areas that were affordable at short notice—have solidified over time into more permanent neighbourhoods that remain neglected by city governments and bereft of welfare services.Footnote 119
Shortly after violence broke out, the Gujarat government announced financial assistance for victims.Footnote 120 The National Human Rights Commission (NHRC) discovered that cash support to Hindu families bereaved during the Godhra attack was double the support to Muslim families bereaved in subsequent violence.Footnote 121 Once exposed and criticized, this discriminatory differentiation was removed.Footnote 122 However, the Gujarat government refused to concede the point of principle involved, and couched the equalization as karsevaks—members of Hindu-Right organizations—voluntarily sacrificing financial aid that was their due.Footnote 123
The amount of cash aid the Gujarat government offered for various types of loss was surprisingly low—closer to what victims of mass violence had received in the much poorer state of Bihar 13 years previously than what the Delhi High Court had ordered more recently in its Bhajan Kaur decision. Low rates of financial aid created an opening for the BJP’s competitors to intervene. In 2006, the Congress and its allies in New Delhi announced financial aid from the national government for victims of mass violence in Gujarat, on par with what the Centre had announced the same year for Sikh survivors of mass violence in 1984.Footnote 124
The Gujarat government gave little, and grudgingly, to victims. It did not ease access to financial aid either. Human rights groups say that the government failed to issue identity documents to people in relief camps. Nor did the government ameliorate the requirement that a missing person would be treated as dead only after seven years, which left many families whose loved ones had been burnt or dumped in mass graves unable to claim financial assistance. The Gujarat police made little effort to recover the remains of people who had been burnt or anonymously buried. When some of the bereaved tried to uncover mass graves on their own, they were criminally charged and prosecuted by the state.Footnote 125
The NHRC recommended in 2002 that the Gujarat government rebuild places of worship that had been damaged and destroyed.Footnote 126 A year later, the Islamic Relief Committee of Gujarat (IRCG), a religious charity, challenged the government’s failure to do so in the High Court.Footnote 127 In the long-running judicial review that followed, the government maintained that it need not aid in rebuilding mosques because it had not rebuilt places of worship damaged in an earthquake and a terrorist attack in previous years. Hesitant, perhaps, to make an unpopular decision, the High Court urged the government to resolve the dispute through dialogue with the IRCG. When the government remained implacable for the better part of a decade, the court finally held in 2012 that failure to rebuild places of worship violated the constitutional right to equality, as state negligence had allowed the destruction in the first place.Footnote 128 The High Court ordered the government to assist in rebuilding damaged mosques, and constricted executive discretion in the matter by appointing judges to adjudicate funding applications and monitor government compliance.Footnote 129
5. How Reparation Programmes Fall Short
These efforts to aid victims of mass sectarian violence, spanning 20 years and four different states, share a common core of measures. Below, I appraise these measures against the Basic Principles and expert policy guidance, and highlight the ways in which they fall short.
5.1 Acknowledging Responsibility
While Indian governments resort to a well-established repertoire of reparative measures, they have never actually made reparation to victims of mass violence. After each episode of violence discussed in Section 3, the aid proffered to victims was categorized as “ex gratia.” Thus, Indian governments have tacitly rejected that their egregious failure to control violence against particular minorities obligates them to recompense the victims. As such, government support for victims is, in a legal sense, not reparation, but welfare.
Victims’ right to reparation, and the concomitant obligation on the state, has been unequivocally articulated by the Indian judiciary in decisions such as Bhajan Kaur. However, judicial recognition has not, thus far, shifted the government position. In practice, as I discuss in Section 5 below, governments face a strong expectation that they will assist people who have been attacked en masse, and they have usually done so in some measure. But conventional practice and political constraints do not amount to legal recognition.
5.2 Harm Covered
Relatively soon after each episode of mass violence, state governments announced a cluster of measures for victims. Across all four episodes, these measures responded to the same types of harm and loss—the killing of a family member, disability and injuries of varying intensity, and damage to personal and (everywhere but in Assam) commercial property. Governments entirely ignored emotional harm and the frequently debilitating effects of trauma and mental illness. Sexual violence and its impacts were also ignored. Economic harm was recognized only very partially. Longer-term harm, including inter-generational harm, was acknowledged, albeit belatedly, by the governments of Delhi and Bihar but disregarded by the governments of Assam and Gujarat. When aid was enhanced by the Delhi government, increases were initially restricted to families who had been bereaved rather than extending to victims who had suffered other types of harm. Victims who were no longer living in Delhi—many Sikh families migrated out of the city after 1984—were also deprived of increased aid. For example, Paramjeet Singh, who moved from Delhi to Punjab after being attacked in 1984, did not receive increased cash assistance in 2006 until he successfully challenged the Delhi government in court.Footnote 130
Under-inclusiveness might be due, at least in part, to the exclusion of victims from programme design and implementation. Neither the national government nor any of the relevant state governments formally consulted victims when designing reparative measures. Victims and civil society groups lobbied governments to improve support and implement the measures already announced, in the years following an episode of mass violence. While such advocacy might have influenced assistive measures over the years, governments have not pro-actively sought input from the communities targeted during mass violence. Consulting affected communities would potentially make government aid more nuanced and effective (though governments would need to guard against reproducing local discrimination and hierarchies in official measures). Sustained engagement with victims might also help to restore a sense of citizenship, and build trust in state institutions.
5.3 Financial Assistance
None of the governments involved attempted any restitution after mass violence. Government assistance was primarily financial, with victims receiving lump-sum payments depending upon the type of harm they had suffered. The governments of Delhi and Bihar introduced longer-term financial support for widows in the form of monthly pensions, but did so over 20 years after violence took place.
The money given to victims was not intended to be compensatory, in the restitution in integrum sense of covering losses attributable to harm suffered during mass violence. Cash assistance offered in the first instance after all four episodes of violence would, for most families, hardly have made a dent in the costs of dealing with serious injury or the death of an earning member of the family. Rates of cash assistance for damaged property were even lower relative to the market value of property.
Official records suggest that governments record the amount of aid they have disbursed, but they do not assess—even internally—the economic harm actually resulting from mass violence. A truer picture of economic harm would include the market value of restoring private property that was damaged and destroyed. It would also mean estimating the costs of violence against individuals. When people are murdered, sexually assaulted, beaten, attacked with weapons, and driven from their homes, serious and foreseeable consequences follow. The costs of absorbing many of these consequences, such as medical treatment, relocation, lost earnings, or rebuilding a business, can be estimated reasonably readily.
Other consequences are admittedly harder to anticipate and assess. The fees lost when someone drops out of college because she is fleeing to a relief camp are easily calculated, but how this educational disruption hampers her career trajectory is more difficult to estimate. Trauma might manifest as mental or physical illness months, even years, after someone was attacked. The lack of psycho-social public health services might lead grief and anxiety to congeal into debilitating mental illness. Fearful employment choices by survivors of mass violence, driven by the understandable desire to remain unobtrusive—choosing poorly paid work within one’s own community rather than better paid work outside it, for example—are unlikely to be economically optimal.
While these slower-burning impacts of mass violence are difficult to estimate for hundreds of people, the more obvious costs are not. Estimating them would compel governments to confront more fully the price of complicity in mass violence. It would underscore that a wider group of people have sustained loss than state reparations programmes have typically included. It would also highlight the intersecting, mutually reinforcing nature of different types of harm, as well as the inter-generational effects of violence. It should lead governments to offer far more generous financial support than it has done, and offer it promptly rather than protractedly, after several years of being lobbied. Recognizing the insidious, long-running, often permanent consequences of mass violence should lead to the further recognition that the state’s standard response of cash aid is a minimal measure, and victims need more comprehensive support.
5.4 Assistance in Kind
Support in kind, including seeds, low-interest loans, and insurance, was extended to victims in Nellie, Delhi, and Bhagalpur relatively soon after violence subsided. Many years elapsed before the governments of Delhi and Bihar introduced more substantial non-monetary measures for victims, including housing assistance, preference in public-sector job applications, and small pensions for widows. These measures implicitly acknowledge that the partners and children of people murdered in 1984 and 1989 suffered long-term costs that are very likely to have constricted their options and capacities. However, such measures excluded the disabled and seriously injured, who also suffered irreversible economic harm. Further, for many families of the dead, this help may have come too late to counteract the impoverishing effects of mass violence. In contrast to its counterparts in Assam, Delhi, and Bihar, the Gujarat government offered neither punctual nor belated support in kind to victims, and rejected the central government’s offer of such aid.
5.5 Relief and Rehabilitation
Emergency relief measures are not expressly included in the Basic Principles, or in international jurisprudence on the right to reparation, which focus on measures to redress abuse that is past.Footnote 131 Somewhat in contrast, relief is needed while violence unspools, when peace is still precarious and is intertwined with the state’s duty to curb violence and control crime. However, robust relief measures are crucial in their own right, and they also prefigure and shade into measures to rehabilitate victims. Crisis measures, in responding to injuries, trauma, and displacement, influence how people rebuild their lives, including resuming jobs and educations, testifying confidently in court, and availing public services for longer-term care, treatment, and education.
Unlike longer-term reparations, emergency relief in India is grounded in well-established norms. As discussed in Section 3, the Famine Code, a long-standing disaster response protocol, guides governments responding to mass violence, even though it was designed for environmental disasters rather than violent displacement. But, while relief protocols are well established, they are also rather narrow; these limits are evident in the relief provided after these episodes of mass violence.
Relief camps were established in response to each episode of mass violence, though reluctantly and tardily in Gujarat. These camps provided safety—they were guarded temporarily by soldiers or the police—and barebones supplies of food and water. However, camps did not have any psycho-social services. It is unclear whether any arrangements existed for accessing legal aid. Officials did not issue identity documents to displaced individuals. Local governments were under no obligation to reunite separated families, attend to abandoned minors, or search for missing people. Further, governments sought to close camps hastily—going as far as forcibly removing people in Delhi in 1984—which sparked legal challenge by camp residents in Delhi and Gujarat.
Families exiting camps in Delhi and Bhagalpur received small sums of money. But, beyond this, no governments offered further protection. In response to right-to-information applications, none of the relevant public authorities disclosed any records on how many people had been displaced by violence. Governments do not seem to have monitored whether people were allowed to return to their homes without fear and coercion, nor did public clinics and schools in violence-affected areas arrange special support for victims.
Thus, emergency relief following mass violence is skeletal, and longer-term rehabilitative measures non-existent. Without such support, communities devastated by violence, such as Bhagalpur’s Muslim weavers, never regained the ground that they lost. Many families displaced by mass violence are permanently displaced. Fearful of returning to their homes, or not allowed to return, they migrate to safer locations. However, without identity documents, they have struggled to access public services, and in some instances even the financial aid announced for victims. In Gujarat, displaced Muslims clustered in neglected shanty towns for years after 2002, while the government staunchly denied their existence until pressed by the Supreme Court to deliver welfare services to these isolated ghettoes.Footnote 132
5.6 Satisfaction
5.6.1 Recovering the Dead and Disappeared
After receiving emergency relief and a sum of money, most victims of mass violence have been expected to fend for themselves. While some aid in kind was extended by the governments of Delhi and Bihar, even these relatively progressive governments barely offered what the Basic Principles describe as satisfaction. Satisfaction includes measures such as, inter alia, public apologies, memorials, reuniting families, recovering the remains of the dead, and educating people about past atrocities.Footnote 133
Governments made no systematic effort to search for the disappeared or the dead after any of these bouts of orchestrated bloodshed. In Bhagalpur, the police not only failed to exhume mass graves after armed attacks abated, but they also colluded in disguising the graves by planting cabbages over buried bodies.Footnote 134 These vegetable beds were uprooted when vultures roosting on nearby trees aroused the suspicions of the senior district official investigating violence-affected areas.Footnote 135 Years later in Gujarat, Muslims in Pandharvada village were threatened with criminal prosecution when they tried to extract the bodies of their loved ones out of a mass grave on their own in 2006.Footnote 136 Rather than retrieving the remains of the dead, the police began guarding the grave from the bereaved.Footnote 137
5.6.2 Apologies
This neglectful, callous attitude also extends to official acknowledgement and apologies. The solitary official apology for mass violence came when the Indian prime minister apologized before Parliament for the government’s failure to control attacks against Sikhs in 1984. Neither the central government nor any of the relevant state governments have publicly accepted responsibility for equally egregious failures in 1983, 1989, or 2002. The lack of apology is sharpened by the fact that high officials have trivialized these episodes of violence. In 1984, then prime minister, Rajiv Gandhi—hurriedly appointed when his mother was assassinated—reacted to attacks on Sikhs by cadres of his own party by declaring at a public rally “when a mighty tree falls, it is only natural that the earth around it shakes a little.”Footnote 138 In 2002, soon after attacks on Muslims in Gujarat, the prime minister, BJP leader Atal Bihari Vajpayee, asserted that “wherever Muslims are living they don’t want to live in harmony.”Footnote 139 Gujarat’s chief minister, Narendra Modi, lauded his government’s refusal to set up relief camps when he rhetorically asked supporters at a public rally why Gujarat should run “breeding centres” for Muslims.Footnote 140
Such belittlement, when it comes from the head of government rather than a junior politician rallying the party faithful, validates violence and the subsequent neglect of its victims. An apology by someone in the same high office signals the equal citizenship of targeted communities and is important not just for people directly affected by violence, but for members of those communities as a whole. For many survivors, it would have been important to witness India’s first Sikh prime minister, Manmohan Singh, apologize for violence that a previous Congress-led government had tolerated, and then dismissed as a “natural” inevitability. Of course, the 21-year delay diluted the reparative potential of the prime minister’s remorse, as did the continued impunity of Congress politicians who had led mob attacks against Sikhs.
5.6.3 Rebuilding Community Resources
In Delhi, Bhagalpur, and Gujarat, armed mobs targeted not just private property belonging to religious minorities, but also community resources, whether water sources in Bhagalpur or places of worship in each instance. Official records from the Assam government do not disclose whether there was similar destruction in Nellie and, given the nature of violence—a focused, ferocious massacre lasting a few hours—there may not have been. In comparison, Delhi and Bihar recorded considerable damage to religious buildings, with the Bhagalpur Commission of inquiry specifically noting hours-long onslaughts on particular mosques.Footnote 141
Flattened shrines and vandalized wells are not merely collateral damage, but a core part of persecuting religious minorities. Like dismissive speeches by public officials, this destruction disparages the citizenship of non-Hindus. Thus, like official apologies, even token support by the state for restoring community resources is an important counter-measure. It provides visible recognition of the equal right to worship and inhabit public spaces as an Indian citizen, regardless of one’s faith.
The governments of Delhi and Bihar contributed towards repairing places of worship damaged during violence. The local administration in Bhagalpur installed water pumps in communities whose wells and ponds had been poisoned by decomposing bodies. In Gujarat, by contrast, the state government refused to aid in reconstructing broken mosques until it was ordered to by the High Court in 2012. Any government drawing up a reparations programme has to choose between competing goals, and might justifiably decide that measures other than rebuilding religious structures took priority. However, in Gujarat, this choice signalled not alternative priorities but general antagonism towards reparative measures for Muslims.
5.7 Summing Up: Shortfalls and Variance
Thus, Indian governments have failed to acknowledge that victims of mass sectarian violence have a right to reparation. Government aid has consistently been categorized as ex gratia and therefore compassionate rather than obligatory. While Gujarat’s efforts were particularly grudging, all the programmes developed within this welfare framework fall well short of international standards. They have excluded entire categories of harm. They have been extremely slow. Victims have not been consulted when designing aid or disbursing it. When viewed in the round, aid extended to victims after each episode of mass violence has not been commensurate with the harm that they suffered.
Contrary to the Basic Principles, these programmes do not attempt to restitute or compensate victims. The cash offered to victims for different types of harm was very low initially and, after some episodes of violence but not others, has increased after sustained negotiation. Many experts emphasize that reparation following large-scale atrocity should focus on rehabilitation and what the Basic Principles term “satisfaction” for victims. They urge governments to focus on rehabilitating communities, building peace, and honouring losses. Indian governments have failed to do this. While they have usually delivered emergency relief according to domestic standards, they have offered minimal rehabilitative services once the crisis has past and equally sparse gestures of respect or apology.
Moreover, official acknowledgement or assistance for long-term inter-generational loss, even from more progressive dispensations in Delhi and Bhagalpur, came late, in fits and starts, after survivors lobbied and litigated for it. It is clear that governments gave little thought to whether relief and reparation cohered internally, or dovetailed with the criminal justice process.
While disjointed and inadequate, aid measures have been impartial within the remit of a particular programme, with the dishonourable exception of discriminatory financial aid to victims in Gujarat in 2002. However, when designing aid packages, governments did not incorporate safeguards against discrimination within families such as, for example, designating women as equal beneficiaries of financial aid within a bereaved family.
Further, aid packages have varied across different episodes of mass violence. For example, cash assistance extended by the state to the bereaved in Nellie, Delhi, Bihar, and Gujarat has diverged considerably, even taking into account that nominal amounts would increase over time for the real amount of cash assistance to remain consistent. The Assam government gave Rs. 5,000 to families of people killed in the Nellie massacre as well as in other election-related violence in Assam in 1983. A year later, the Delhi government announced double that amount of money as aid for the bereaved. In 1989, families of the people killed in Bhagalpur were given Rs. 100,000 in aid by the Bihar government—twenty-fold higher than the aid given to similarly placed families in Nellie six years previously. The same amount of cash aid—Rs. 100,000—was given 13 years later to bereaved families in the far wealthier state of Gujarat. Cash assistance to the disabled and injured fluctuates similarly across different episodes of mass violence. These variations are deeply problematic. Were the government compensating victims to the restitutio in integrum standard, the money given to different individuals would clearly vary depending upon the particular losses each had sustained. In fact, victims have received standardized sums pegged to particular categories of harm. As such, variation in financial aid means that, in effect, the loss suffered by one person is accorded more weight than the equivalent loss suffered by another person in comparable circumstances.
Not only are initial aid packages unequal, but differentiation between victims has deepened over time. As discussed in Section 3 above, aid after mass violence in Delhi and Bhagalpur has been iterative, increasing significantly and shifting shape over time. By contrast, the Assam government has extended no additional help to victims of the Nellie massacre, and the Gujarat government has actively resisted victims’ demands for further support.
6. Negotiating the Shortfall
The discussion above indicates that the substantive weaknesses in government aid to victims of mass violence are systemic; they persist, albeit with some variations, across decades, regions, and political dispensations. As such, it is worth reflecting on aspects of the political environment in India that shape government decisions and civil society responses, and thereby shape these ungenerous, unimaginative efforts to make reparation.
Three of the four episodes of mass violence considered in this article were preceded by tumult: Assam was in turmoil in 1983; Delhi—and India as a whole—deeply shaken in 1984 by the prime minister’s assassination; and tensions ran high in north India in 1989 as Hindu-majoritarian groups campaigned to tear down an ancient mosque. Gujarat was calm in 2002, until the brutal burning of train carriages carrying pilgrims in Godhra train station. However, even taking into account the turbulence surrounding each episode of mass violence, on none of these occasions was India’s political system being fundamentally restructured. Unlike in transitional societies, constitutional principles, legal rules, and public institutions remained intact.
Social movement scholars define political and legal opportunity structures as the consistent features of the political system that aid or hamper popular mobilization.Footnote 142 These include how primary political institutions are structured, how stable they are, how open or closed, how accessible or remote, and whether they tend to accommodate challenge or to repress it.Footnote 143 Since India has been undisturbed by fundamental political transition, the “opportunity structure” for improving and challenging the state’s choices has fluctuated little across these episodes of mass violence. As a fundamentally stable, non-transitional state, India did not face the strained capacity and stark trade-offs of a state in the thick of creating a new political order. But this greater room for manoeuvre did not, in fact, lead to effective reparations.
I argue that particular features of the political opportunity structure dilute any principled commitment by Indian governments to repairing the harm wrought by mass violence. I highlight below the features of the political opportunity structure that stymie commensurate reparation, as well as those that encourage it. I then suggest legal reform that would improve the opportunity structure for reparation after mass violence and pressure the state to respond more effectively to victims.
6.1 Sectarian Electioneering and State Institutions
India’s violence-laden identity-based electoral politics facilitates mass violence against religious minorities. Sectarian violence, one-sided attacks as well as rioting between groups, is stitched into electoral politics in India. Most political parties within India’s energetic democracy are organized around caste interests and, to a lesser extent, religious affiliation. Political mobilization involves demands for better public services and aspirations for greater prosperity, but also fixates on historical and contemporary injuries—genuine or imagined—faced by particular communities. Religious minorities are particular targets of this politics of grievance. They are routinely pilloried by right-wing Hindu groups, and opportunistically maligned by more secular politicians as well.
Hindu-majoritarian political organizations, such as the Bajrang Dal and Shiv Sena, overtly rally men to defend the faith, organizing exercise drills and distributing weapons to their membership. Most mainstream political parties are not as transparently martial, but nevertheless partake in what Brass describes as “institutionalized riot systems,” maintaining a critical mass of members who can skirmish with rival parties and attack the people who support their competitors.Footnote 144 In regions where violence against religious minorities is endemic, popular discourse might characterize friction between different communities as primordial and justifiable. But underlying this discourse are political parties who leverage local disputes to stoke ethnic tensions and political rivalries, and maintain cadres who can be pressed into violence at short notice.Footnote 145 In some parts of the country, aggression has been such an established part of the political arsenal that it has identifiable rituals and routines. Festivals and religious celebrations often provide the occasion for it; rumours and disrespectful gestures are used to fuel it.Footnote 146
Three of the episodes of mass violence that I look at came in the wake of political movements disparaging the community which was ultimately attacked. The anti-immigrant agitation in Assam wanted Bangladeshi Muslims expelled from the state. Indian Muslims as a whole were vilified by the flamboyant campaign to replace the Babri mosque with a temple. Hindu-majoritarian organizations such as the RSS and Bajrang Dal had long been active in Gujarat; their grievance-laden vendetta against Muslims not only had cultural cache, but also—with the BJP in power and a chief minister who had helped to organize the Ram-temple movement in the 1990s—considerable influence on the government.Footnote 147 Experienced in the logistics of group attacks, political parties were able to seize quickly upon the horrific violence of Mrs Gandhi’s assassination or the Godhra train-burning and harness the momentum of long-running political campaigns in Bhagalpur and Assam to attack particular communities. Granted licence by the government, one-sided violence could be executed on an extraordinary scale.
Despite the shocking cruelty that followed, voters did not punish political leaders for condoning violence. In Assam, the leaders of the anti-immigrant movement were voted into power, though the direct perpetrators of the Nellie massacre played little role in formal politics. The ruling Congress party called early elections in 1984 and, after a campaign with advertisements that described Sikhs as threats to the nation, returned to power with its highest-ever parliamentary majority.Footnote 148 Many years later, the BJP government in Gujarat also tried to capitalize on surging majoritarian sentiment by calling state elections barely three months after mass violence had subsided in 2002.Footnote 149 The Indian Election Commission rebuffed this attempt, citing the risk of further violence, but later that year Narendra Modi led the BJP to a comfortable victory in elections for the state legislature.Footnote 150 State elections in Bihar in early 1990, soon after mass violence in Bhagalpur, departed from this pattern. The incumbent Congress party suffered defeat. However, the victorious Rashtriya Janata Dal allowed some of the main perpetrators into its fold, and proceeded to shield them from prosecution.Footnote 151
Ascribing these victories solely to majority triumphalism over attacks on minority groups would be simplistic. Sympathy over Indira Gandhi’s murder contributed to support for the Congress party and for Rajiv Gandhi as a prime ministerial candidate. In Gujarat’s state assembly elections in 2004, Narendra Modi’s reputation for honest efficiency very likely drew voters as much as his invective against Muslims. But, while a single variable cannot explain these electoral wins, it is also clear that abetting a pogrom is not punished at the ballot box in India, and might even be rewarded.
By the same token, democratic elections do not incentivize reparations for victims once target sectarian violence subsides. My discussion in Section 3 indicates that governments that tolerated mass violence in the first place have little commitment to addressing the fall-out. Within India’s first-past-the-post electoral system, the political party or coalition that wins a majority of seats in the legislature leads the government of the day. The ruling party’s dominance in the legislature dilutes legislative scrutiny of the executive branch. Even if opposition parties demand robust support for victims of mass violence, a majority of legislators are likely to defend the government’s record or stay silent. Moreover, so entwined has sectarian discord been with political mobilizing in India that criticism by opposition parties is often weakened by their own past complicity in violence, with legislative debate devolving into whose record is the bloodiest rather than focusing on the needs of victims.
Even within an electoral politics that nurtures hostility based on caste and religion, the four episodes of mass violence discussed in Section 3 were extraordinary in their scale, planning, and ferocity. Nevertheless, a political culture where violence against religious minorities is, to some degree, routine, and yields several legislators, ministers, and officials who view persecution and pogroms as inherently justified. It produces many more who regard identity-based violence as strategically sound—a potent, even if unpleasant, technique for political success. This mix of acceptance, cynicism, and chauvinistic conviction is reflected in the fact that national and state governments have never recognized a legal duty to make reparation to victims of mass violence.
6.2 Slow and Contingent Gains
However, if the government and legislature are structured in ways that fail to contain majoritarianism, other constitutional checks and balances have stepped into the breach after mass violence, albeit weakly and partially. As time has passed after a particular episode of violence, these mechanisms for accountability have gained strength, and it has been possible to seek, and to secure, more support for victims.
6.2.1 Judicial Intervention
The Indian judiciary has expansive powers of review and remedy under the Constitution of IndiaFootnote 152 and is therefore well positioned to redress government neglect of victims of mass violence. Disappointingly, the judiciary’s record in this regard is distinctly mixed, with judicial decisions in the aftermath of violence betraying bias and neglect. As discussed in Section 3, criminal courts in Assam acquiesced without question when the government sought to withdraw all charges related to electoral violence—including charges involving murder—and effectively agreed to an amnesty for the massacre at Nellie. In 1984, the Delhi High Court told the government to keep relief camps running—but only for an additional fortnight—when human rights groups challenged the decision to close camps. A year later, the High Court roundly rejected a petition challenging the Delhi government’s failure to distribute financial aid and investigate the perpetrators of mass violence.Footnote 153 The High Court in Gujarat denied relief camp managers even temporary support, briskly dismissing their petition demanding better services from the government. Despite the startling, extreme violence at issue, in none of these instances did courts reflect upon victims’ constitutional rights to life, equality and freedom of religion.
However, as time passes after a particular episode of mass violence, judges seem less likely to side with the government. Twelve years after mass violence in Delhi, the Delhi High Court declared, in Bhajan Kaur, that the government owed compensation to victims of mass violence because it had failed to fulfil its duty to prevent such violence and protect lives. Not only did the court contradict the executive branch’s rejection of such a duty; it also declared that the money—compensation rather than a favour, in the court’s view—given to Mrs Kaur was insufficient, and increased this amount for her and, by corollary, others in her situation. In Gujarat, the High Court gradually moved from suggesting compromise to upholding the right to equality when responding to an Islamic charity’s petition demanding government aid for repairing religious sites damaged in 2002. While the court suggested resolving the dispute through dialogue for several years, it finally switched to imperative mode in 2012. It held that the constitutional right to equality obligated the government to contribute towards repairing mosques and shrines torn down during violence against Muslims in Gujarat.
6.2.2 Government Compliance
Judicial interventions like these have a material impact. Bhajan Kaur led the government to substantially increase monetary aid to all the victims of mass violence in 1984. This judgment also influenced policy. A year after the court’s decision, India’s Ministry of Home Affairs formulated national Guidelines on Communal Harmony. These guidelines emphasize that governments should provide relief quickly, without discrimination. They direct district administrations to ensure supplies of basic necessities, including security and medical assistance, in relief camps. Most concretely, these guidelines specify assistance of Rs. 300,000—the amount ordered by the Delhi High Court for Bhajan Kaur—to families who lose a close relative to “terrorist or communal violence.”Footnote 154 More recently, in 2008, the national government supplemented the Guidelines on Communal Harmony with a scheme for disbursing financial assistance to victims of sectarian and terrorist violence.Footnote 155
These policy guidelines reflect some of the same flaws manifest in the administrative aid packages discussed earlier. Like those bespoke programmes, these general guidelines provide primarily for monetary aid. They neither use the term “reparation” nor engage with the various types of reparative measures that the Basic Principles advocate. While these policy documents prescribe what governments should do after sectarian or terrorist violence, they do not acknowledge a binding legal duty towards victims. Past experience also suggests that these policy guidelines are easily side-lined: the Gujarat government gave considerably less than the recommended monetary support to victims of mass violence in 2002.
But, despite their limited impact and evasive nomenclature, the Guidelines on Communal Harmony capture the fact that relief and financial aid to victims of mass violence are settled political norms, even if not black-letter legal provisions. Complete defiance of these norms is difficult. In 2002, when the Gujarat government ignored these expectations, withholding emergency services and earmarking a higher sum of money for Hindu victims of violence, it was forced to backpedal by media exposure and criticism. While electoral politics might facilitate identity-based violence, governments—as distinct from political parties—face, and act on, the expectation that “something”—even if minimal and ill-defined—is owed to victims of mass sectarian violence. Political leaders might begrudge that support, but governments cannot entirely withhold it.
The distance between the inclinations of political leaders in power on the one hand and the government’s incentives and constraints on the other widens over time. In the months and years after a large-scale attack on a minority group, the brazen political patronage that facilitated such violence becomes more muted. Politicians in high office cannot publicly condone violent crimes, so support for perpetrators retreats behind closed doors, playing out instead in biased decisions by the police, prosecutors, and junior officials. Political alliances shift, as do individual political fortunes. Over a longer timeframe, power changes hands and successive governments might genuinely seek to redress, at least somewhat, a predecessor’s support for violence. When power changed hands in Bihar in 2005, for example, the new government made an effort to improve reparation for victims of mass violence in Bhagalpur and retry criminal cases that had been jettisoned in the 1990s.
6.2.3 Contingent Improvement
As time elapses, there is increasing room for civil society—religious organizations, human rights NGOs, survivors themselves—to press for accountability. This might mean litigating, negotiating with governments, lobbying legislators, official watchdogs and journalists, and harnessing both competition and concert between these different actors. If a court order compelled the Delhi government to increase aid to victims, the NHRC shamed the Gujarat government out of giving Muslims less money than Hindus, and a conscientious district official in Bhagalpur uncovered mass graves, allowing bodies to be claimed and culprits to be named.
Improvements in one aid package have created opportunities for improvements in another. Once the national government enhanced aid to the Sikh victims of mass violence in Delhi, it would have been difficult to refuse the Bihar government’s demand for similar benefits to victims in Bhagalpur. But electoral motives are always salient. The actual release of central government funds was strategically timed to boost the prospects of the United Progressive Alliance in elections in Bihar the following year.Footnote 156
Securing better support depends upon the political influence that victim groups can wield—their ability to lobby and litigate—and the state’s willingness to engage with them. This, in turn, depends substantially upon how their particular community feeds into the local and state electoral calculus, how wealthy and educated they are, and how easily they can access elite interlocutors.
Muslims in Bihar are a sizeable, electorally significant minority, comprising 17% of the state’s population. Bhagalpur had a slightly higher proportion of Muslims than the state average in 1989, and the Muslim community was relatively well off, with skilled artisans working in the local textile industry. These were very likely important reasons why the Bihar government attended more carefully to helping victims than the governments of Delhi and Assam had done. Many years later, the Janata Dal (United) was consciously courting the support of Bihari Muslims; this meant, inter alia, improving support for victims of mass violence in Bhagalpur. By comparison, Muslims are a smaller minority in Gujarat, comprising 10% of the population. Given the Hindu-majoritarian movement’s deep roots and cultural sway in Gujarat, it has hurt the BJP little—and seems to have helped it—to denigrate the Muslim community in the years following mass violence in 2002.
Only 2% of Indians and only 4% of Delhi’s residents are Sikh. Their relatively small numbers, and resulting electoral insignificance, made it easy for the Congress-led government to mistreat victims of mass violence in 1984. Sikh families were denied protection and basic services, and then forced out of relief camps before the city was safe for them. Many victims of mass violence were of extremely modest means. More generally, however, the Sikhs are one of India’s most prosperous communities. With strong community support, Sikh groups have campaigned for better support for victims and criminal prosecution of those who led attacks in 1984. The Sikh community’s demands for accountability have been eloquently argued in court and in Parliament. While this has not led to effective prosecution of culprits, it has prompted improved reparations.
Unlike aid to victims in Delhi, Bhagalpur, and Gujarat, aid to the Bengali Muslims of Nellie—which was meagre to begin with—has remained static since 1983. The vast majority of people affected by violence in Nellie were extremely poor, and lacked the economic and political capacity of their more economically diverse counterparts in Delhi, Bhagalpur, and Gujarat. In addition, they were simply one set of recipients of the aid given to victims of pre-election violence across Assam. Lobbying for more support in that context would require co-ordination amongst many different groups, some of which might be deeply antagonistic to one another. Nellie’s Bengali Muslims, with tenuous social capital, were poorly positioned to press for better treatment.
Thus, efforts to extract more than minimal reparation from the state are highly contingent. They are shaped not just by numerical strength, social capital, and shifting political fortunes, but also by governments’ efforts to evade other types of accountability for mass violence. In a political culture where the two largest national parties—the Congress and the BJP—have organized mass violence, and many of their regional counterparts have participated in the “institutionalized riot systems” discussed by Brass, leaders across party lines are invested in shielding politicians from prosecution. Governments are not eager to prosecute politicians who instigated violence or discipline bureaucrats who tolerated it, even when the political forces that organized mass violence are no longer in power. As a result, the aid given to victims does not fit within an overarching quest for justice. On the contrary, Indian governments have displayed the tendency identified by Moon of increasing aid while deflecting other dimensions of accountability.
7. Formalizing Norms, Constraining the State
India is a young country, growing richer at a reasonable pace.Footnote 157 With the shifts in opportunity, education, opinions, and demographics this brings, the ritualized, routinized majoritarianism embedded in electoral politics will morph and might diminish. But, regardless of where these longer-term shifts lead, it is important more immediately to improve reparations for victims of mass violence. Governments that tolerated, and even facilitated, such horrific violence bear responsibility for its consequences. On a more utilitarian note, a majority of the victims of such violence are poor, and fall deeper into poverty without considerable state support in the aftermath. The majority of people attacked during the Nellie massacre were living in poverty. The assaults of 1984, 1989, and 2002 may have cut across class, but nevertheless inflicted great damage on less wealthy Sikhs and Muslims. It is easier for attackers to tear down a hovel than a bungalow, just as it is easier for victims to access doctors and lawyers if they can afford to pay privately for these services.
Reparations would be improved in the short term by legislating national standards on relief and reparation. Of course, a statute cannot restructure a politics organized around ethnic and religious chauvinism. But authoritative national standards would provide a formal mechanism to pressure the state after large-scale identity-based violence, at the time when experience shows us that victims are at their most vulnerable, governments at their most hostile, and courts at their most craven.
Obligating the state to provide reparation would shift the political opportunity structure in favour of victims of mass violence. It would allow victims to demand reparation as a matter of right, which strengthens their standing to negotiate with the government and challenge it in court. A law might also embolden citizen-facing officials on the ground to demur when political leaders pressure them to withhold aid from victims. Like a bureaucrat, a judge too would have stronger incentive to protect targeted groups if legal standards on reparation existed, because ignoring those standards would telegraph bias or political compromise, and dilute judicial credibility. Further, legal standards would structure and legitimize judicial intervention in reparations for mass atrocities. If local or state governments refused to assist victims of sectarian violence, a court could simply order recalcitrant officials to fulfil their statutory obligations.
Thus, by making relief and reparations obligatory, a law is likely to improve the administrative response after mass violence, and empower victims to challenge state negligence. The foundation for formal legal standards on reparations already exists in India, though it is partial and uneven. Court decisions, poorly circulated national policy, firmly entrenched relief codes, and past aid packages for mass violence all foreshadow a coherent set of standards on reparation. So, while norms on emergency relief and reparation would not be unfamiliar, legalizing these norms would entail a shift, binding governments and burdening them more than the current, diffuse standards and practices do.
But legalization will be difficult to achieve. My discussion in Section 5 suggests that Indian law-makers would be reluctant to pass a law addressing the consequences of mass sectarian violence. A law such as this might be more politically palatable if it encompasses reparation not just for victims of sectarian violence, but also for victims of other types of mass atrocities. A more inclusive law on reparation would echo the national Guidelines on Communal Harmony, which prescribe financial aid for victims of both terrorist and sectarian violence. At the international level, the Basic Principles exemplify that a common core of measures can be adapted to respond to different types of loss and suffering.
The Basic Principles also offer guidance on what a law might contain. While the Principles are more expansive and ambitious than domestic legislation would appropriately be, they contain procedural norms and a typology of reparative measures that would be instructive when drafting a statute. Most fundamentally, the Basic Principles recognize a right to reparation for individuals who suffer a grave violation of a human right. An Indian law should similarly recognize that victims of state-sponsored or state-supported atrocities have a right to relief and reparation, putting on a statutory footing a principle that already exists in the common law in India. A law should delineate the range of relief reparative measures the state will offer, and set standards and assign duties corresponding to each.
Detailing what a potential law on reparations should include is beyond the scope of this article. However, it is worth noting that some types of reparative measures might not transition well from the realm of discretion to that of duty. For example, unless accompanied by genuine remorse from senior leaders, an official apology or state-sponsored memorial would neither soothe victims nor deter bigots. Remorse and respect cannot be decreed, and to do so might foster pro forma platitudes that feel demeaning to victims. By contrast, the police can guard a relief camp, doctors can treat the injured, state governments can allocate funds for compensation, and local officials can disburse those funds even if the individual functionaries involved harbour bias for victims of mass violence. While these services would be enhanced by respectful, sympathetic delivery, they are not negated by its absence. As a result, they are more amenable to being legislated.
Some elements of relief and reparation are established practice in India (setting up relief camps or offering financial aid, for example). A law should aim to improve past practice by setting minimum substantive thresholds where appropriate, and legislating procedural principles such as nondiscrimination or participatory programme design. Other elements of relief and reparation have been entirely neglected, such as providing rehabilitative services or searching for people who have disappeared. A law should identify the state’s duties with respect to these reparative measures with as much specificity as possible, to guide governments through previously uncharted, but long overdue, territory.
8. Conclusion
The violence I have discussed in this article is often characterized as atavistic, impulsive, a frenzied rupture. In fact, as human rights groups as well as official inquiries have demonstrated, these massacres were co-ordinated by political groups, and allowed to unfold by the state. Mass violence against religious minorities in India has been systematic not just in the ways it was inflicted, but in the kinship it shares with discriminatory and violent political mobilization more generally. The pull of majoritarianism has warped how Indian governments respond to mass violence, and how they make reparation to victims. While violence as widespread and extreme as these four episodes may not recur—one hopes it never does—milder forms of violent persecution against religious minorities are likely to persist in the near and medium term in India. Shifting the political opportunity structure for religious minorities who suffer persecution is urgent and necessary. A law on reparations would constrain state bias and strengthen opportunities for survivors of violence to demand the support that they need.
