Rethink the communal violence bill
Before I develop my critique, let me begin with a word of admiration. The largest part of public commentary thus far has concentrated on whether minorities require special protection against violence, as the bill purports to do. The NAC is right; the critics are wrong.
The question of minority rights has a fraught and embattled history in political thought. The cultural right is often opposed to giving legal recognition to the concept of minorities, saying it prevents the minorities from assimilation, instead promoting communal division, even separatism. This is fundamentally the Hindu nationalist critique of the bill.
But it is not only the cultural right that takes this view. Radical liberals also reject the notion of group rights and protections. For radical liberals, citizenship is an individual right, and individuals should be allowed freely to choose their identities. Group entitlements imprison individuals and societies, inexorably pushing them towards dangerous collective identities.
By comparison, the Hindu nationalist argument is majoritarian, not rooted in individual freedom. For them, it is Hindu values that define India's national culture, and laws should promote assimilation of minorities to the nation's presumed Hindu centre, not perpetuate a minority mindset. Their reasons might be different, but the cultural right and the radical liberals normally have a seemingly unlikely alliance on this matter.
On minority rights, we also have a tradition of moderate liberalism. A moderate liberal asks: Is a person from the minority community simply a citizen like others? Is a rich minority simply an aggregation of rich citizens, or can it still suffer from some serious disabilities?
Not simply in pre-modern history, but also in modern times, it has been easy to turn groups that the mainstream of a society looks down upon, or dislikes, into objects of collective violence. Even richer minorities have suffered such fates. Think of the Tamils in Sri Lanka, the Indians in Uganda, the Chinese in Malaysia and Indonesia, the Sikhs in Delhi after Indira Gandhi's assassination, and Kashmiri Pandits in the Valley in the early 1990s. This realisation — that even modern societies can easily slip into majoritarianism — is the basis for the moderate liberal claim that minorities need special protection.
Does this position imply that minorities can do no wrong? Is the majority community always to blame? By the 1940s, thinking long and hard about this question, Jawaharlal Nehru had started distinguishing between minority communalism and majority communalism. Both were bad, but majority communalism was infinitely worse. Nehru detested Jinnah's communalism, but called Hindu communalism the greatest danger to India. Why?
Unless one carefully watches, majority communalism and nationalism tend to get conflated. Following the logic that the majority community is the prime, or the only, owner of the nation, such seamless morphing is quite common. Many Sinhalese say that they define Sri Lanka, not the Tamil minority; many Malays contend that they are the sons of the soil in Malaysia, not the Chinese; Hitler argued that Jews could never be Germans; supporters of mass murderers in Gujarat (2002) called violence against Muslims an act of patriotism; the killing of Sikhs in 1984 was also described in several circles as an assertion of Indian nationalism. Two Sikhs had after all assassinated the prime minister, the highest elected official, thus desecrating the nation.
Because of these dangers, the key question for moderate liberals is not about the abstract first principles: namely, should democratic rights be individually based or group-based? Rather, we ask a historically necessary practical question: How can one expect minorities to be patriotic, if they are by definition excluded from the joint ownership of the nation, and violence against them is permitted as a legitimate act or, worse, viewed as service to the nation?
This position does not imply that minority communalism ought to be ignored. Nehru had harsh words to say about Muslim organisations and leaders during a Hindu-Muslim riot in Aligarh in 1954, and wanted those organisations punished. My own research in Hyderabad uncovered many instances when Muslim organisations were egregiously complicit in riots. Hyderabad's mass killers came in both hues, Hindu and Muslim; Hindus had no monopoly over rioting. Other researchers came to similar conclusions. Agar Hindu pachees Musalman marenge, said Hyderabad's Muslim wrestlers to Sudhir Kakar, a psychologist who also researched violence, to hum chhabbees Hindu marenge — yeh jo riot hai, woh one-day cricket ki tarah hota hai (if the Hindus kill 25 Muslims, we will kill 26 Hindus — a riot is like a one-day cricket game).
Moderate liberalism does not willfully disregard the violent misconduct of minorities; it only points to the infinitely greater subversive potential of majority communalism. Though real, Hyderabad and Aligarh are less common; the nearly helpless Muslims of Ahmedabad, Baroda, Moradabad, Meerut, slaughtered mercilessly in riots, have been more typical. Riots have indisputably killed many times more Muslims than Hindus in post-1947 India.
In sum, therefore, fundamentally because of the easy, though erroneous, equation of majorities with nationhood, a democracy must protect its minorities from violence. The NAC is right to emphasise the vulnerability of minorities and to stress that the Indian state can behave in a highly majoritarian fashion. But do we need a new law and a new bureaucracy to make that point?
Should civil servants be liable?
The second important aspect of the bill is its insistence that if "communal and targeted violence" against minorities takes place, it will automatically be assumed that the civil servant in charge of law and order has not exercised "lawful authority vested in him or her under law" and he or she "shall be guilty of dereliction of duty". The bill makes civil servants legally liable for riots. They will be fired, demoted or reprimanded, if a riot takes place on their watch.
At one level, this is a welcome insistence. Quite often, IAS or IPS officers have looked at political masters before deciding how to deal with a riot, even though the rule-book makes it clear that the responsibility of law and order at the ground level lies with the civil servant, not with the politician. By making the civil servant liable, the bill seeks to strengthen the civil servant against the politician. The NAC's assumption is that if civil servants were personally liable for riots, there is a greater chance they would act according to the rule-book, not wait for political signals from above.
But this assumption can only be half-right. The NAC has not confronted a factual question. Why has Aligarh been so riot-prone, whereas Bulandshahr, a town next door, has rarely had a communal riot? Why have Meerut and Morabadad been so communally nasty, whereas the neighbouring Muzaffarnagar and Bareilly hardly ever witnessed a communal riot after independence? Did Aligarh, Meerut and Moradabad have riots because the civil servants stationed there ignored, or supported, the killing of Muslims, or is there something about the local relations of Hindus and Muslims in these towns that made them riot-prone?
Indeed, during 1950-95, as I calculated in my book on communal violence, a mere eight cities of India — Mumbai, Ahmedabad, Baroda, Hyderabad, Aligarh, Meerut, Delhi and Kolkata — had nearly 46 per cent of all deaths in Hindu-Muslim riots. Did these eight cities have repeated "dereliction of duty" by civil servants? Did the peaceful cities have especially duty-driven officials and police forces?
To be sure, there are enough cases of officials ignoring their duties during riots. Delhi after Mrs Gandhi's assassination and Gujarat 2002 have already been cited. More examples can easily be added. But it is also worth inquiring whether that is the only reason riots took place.
When I asked why he succeeded in keeping peace in smaller towns of Maharashtra but, in 1984, failed to prevent an awful communal riot in Mumbai and later in Ahmedabad, Julio Ribeiro, an outstanding police officer of post-1947 India whose commitment to duty has never been questioned, said that there was something about how social and economic life was organised in Mumbai and Ahmedabad, which made his task enormously difficult. The same question — why riots in some places, not in others — elicited an identical answer from most police officers and civil administrators I interviewed in my 10-year long study of communal violence. Would the NAC fire a Ribero, and others like him, for their inability to prevent riots?
Riots are jointly produced. They are, in part, an outcome of how the police officials and civil administrators have performed their constitutionally assigned functions. Rioting is also, in part, a result of how social and economic life is organised in a town, whether Hindus and Muslims are segregated or integrated, and what incentives or capacities such local structures have created for politicians, always in search of political gains, to inflame and polarise, or calm and unite, local communities. The same IAS officer who functioned well in Warangal often felt helpless in Hyderabad. The NAC would like to give more powers to the civil servant. But if riots are jointly produced by the state and society, dealing with one side of the equation is surely not enough.
Indeed, the NAC needs to be given another reminder about the limits of state power. Aren't state capacity and governance in the US, Britain and France much higher than in India? Yet the US could not prevent the so-called Rodney King riots in 1992, Britain witnessed racial rioting in the 1980s, and Arab migrants in France rioted in 2005. Los Angeles, Brixton and Paris burned, while the police wielded their batons and even shot to discipline the crowds. If making the state more powerful and/or rule-governed were the solution to rioting, the world would be an easier place to govern.
A new bureaucracy for communal harmony, justice and reparation?
The bill also envisions creation of a new set of state institutions: a National Authority for Communal Harmony, Justice and Reparation, headquartered in Delhi. The National Authority will have seven members, supported by a "Secretary General, who shall be an officer of the rank of the Secretary to the Government of India". Presumably, the members will have the rank of ministers of state and the chairperson will be of full ministerial rank. The National Authority will be given police and investigative staff when necessary; it can investigate the conduct of army officers during riots; it will have the powers of a civil court for inquiry and investigation; and all district magistrates and police commissioners will be required to report to it on matters concerning communal violence. There will be corresponding institutions at the state level, too. A massive bureaucracy will thus be created.
This great institutional proposal invites a basic question: Does the NAC expect the future of India to be as riot-ridden as India's past has been? Massive law-and-order bureaucracies are normally created to deal with a frequently recurring problem, not for something highly infrequent or rare. We need to ask if riots will be occasional episodes, or regular occurrences, in the coming years. If riots are going to be occasional, we can't justify the creation of a huge permanent bureaucracy.
To treat the future as a mechanical extension of the past is almost always an awful mistake. The future is basically uncertain. That is why we use probabilistic, odds-based reasoning about the future.
What can we say about the odds of rioting in India in the next 10, 20 or 30 years? A vast amount of cross-country research on riots and civil wars has been published in recent years. And a large conclusion has emerged. According to worldwide evidence, riots are regular occurrences at low levels of national income, but only occasional episodes at middle and high incomes.
On why this is so, the research is not conclusive. But two hypotheses have been considered reasonable. First, as incomes rapidly rise, popular aspirations change and a desire for material advancement, perhaps always present, becomes more realistic. It is possible to envision a better economic future, if many others around are rising. As a result, a new politics of aspirations emerges, shrinking space for politicians to mobilise groups for communal riots. With rising prosperity, issues in politics begin to change. Communal discontent does not fully disappear, but it begins to take the form of higher-technology terrorism as opposed to low-tech mass riots. Once that happens, a riot- bureaucracy is incapable of handling the problem. Second, at higher national incomes, state capacities also increase and governance improves. The NAC believes that the latter is not happening in India and that may be true. But it shows no understanding of the rising politics of aspirations, which India is beginning to see, as it moves economically forward. Indeed, this could well be an important reason India has witnessed no big riots since 2002.
A preponderant majority of India's riots took place, when India was a low-income country. India has now become a middle-income country and is growing faster than ever before. To the extent we can make predictions on the basis of research, riots will increasingly be a matter of India's past, not its future. While it is not impossible for this prediction to be wrong, it will be a great surprise if communal riots returns to India in a big way, as the nation rises up the income ladder. The basic point is that we can't create a huge bureaucracy with unprecedented powers on the basis of a low-odds scenario.
The NAC appears to be a prisoner of India's past, especially of Gujarat 2002. What happened in Gujarat was a crushing embarrassment for all liberal Indians and every effort should be made to punish the guilty, but to build a new bureaucracy to prevent another Gujarat 2002, which is in any case unlikely in the future, will be a terrible mistake.
Indeed, the creation of such a bureaucracy and law might create perverse incentives. Conflict research shows that many people settle petty personal scores in ethnic violence, pretending that a larger ethnic cause is being served. If public servants are made liable for riots, those opposed to them, for whatever reason, might have an incentive to touch off riots — to punish a civil servant they did not like. Thus, even though the normal tendency is for the incidence of riots to go down at higher levels of income, the creation of a riots bureaucracy might counteract that trend. We could reinvent a problem, which would otherwise naturally decline.
A final point is in order. A distinction needs to be drawn between riots and prejudice. The fact that riots decline at higher levels of income does not mean that prejudice and discrimination necessarily go down. After the 1960s, the US has seen very few riots, but African Americans still end up in jail disproportionately. Rich today, Malaysia has had no big riots since 1969, when it was poor, but discrimination against the Chinese and Indians continues. Prejudice also sometimes takes the form of hate crimes, including those perpetrated by the police, both in the US and Malaysia, but riots are few and far between.
The question for India is obvious. Riots may well decline in India in the future, but will prejudice and discrimination against the Muslims, Dalits and Adivasis also go down? That is a bigger question than riots. But it requires construction of an anti-discrimination law and an equal opportunity commission, not a new bureaucracy to prevent riots.
Varshney teaches political science at Brown University and is the author of 'Ethnic Conflict and Civic Life: Hindus and Muslims in India'