OPINION

The decline of the ‘encounter death’

Praveen Swami

Most police forces are reducing use of lethal force — and shedding communal partisanship.

Six months ago, the police raided an apartment in New Delhi’s Jamia Nagar. Two alleged terrorists and a police officer died. By the standards a violence-scarred nation has become accustomed to, the event was unremarkable. But the Jamia Nagar deaths had an exceptional impact, precipitating charges that police forces across India were operating a large-scale shoot-to-kill policy directed at Muslims: a communal project, it was claimed, that was being camouflaged as counter-terrorism.

Participants at an October 2008 convention in New Delhi, for example, declared that there was “a concerted effort by the Indian police, intelligence agencies and certain political parties to portray all members of the Muslim community as ‘terrorists and extremists’ — to be arbitrarily arrested, tortured and killed in fake encounters.”

Members of the Coordination Committee of Muslim Organisations — an alliance made up of the Jamaat-e-Islami, the All-India Muslim Majlis-e-Mushawarat, the Jamiat Ullema-e-Hind, the All-India Milli Council and the Jamiat Ahl-e-Hadis — went further, demanding that during a “search operation in any Muslim locality, at least one-third of the raiding force must consist of officers belonging to the minority community, and minority elders of the affected area should be taken into confidence.”

Media accounts since have elevated the charge that India’s police officers are trigger-happy bigots to the level of received truth. Little effort has been made, though, to see if the allegations rest on sound empirical foundations. They don’t. In fact, the police are reducing their reliance on lethal force, and shedding communal partisanship. The reason why they do not rely on force helps to explain just why India’s democracy, often reviled by metropolitan elites, is so important to hundreds of millions of voters.

No public-domain documentation exists on the religious identity of individuals killed by the police. Databases maintained by the National Crime Records Bureau set down each incident — but not the religious identity of the victims. The police are obliged to report all lethal force deaths to the National Human Rights Commission. In addition, the Union Home Ministry monitors incidents involving the use of lethal force by the police. For the most part, though, the reporting of incidents by the States is less than comprehensive.

Based on the available Central government documentation, The Hindu was able to examine 750 civilian deaths in police firing which took place between January 2004 and December 2008 — about two-thirds of those estimated to have been killed during this period. Spread across Assam, Delhi, Gujarat, Haryana, Karnataka, Maharashtra, Madhya Pradesh, Orissa, Rajasthan, Tamil Nadu, Uttar Pradesh, Uttarakhand and West Bengal, the data exclude deaths in insurgency and counter-terrorism in the northeast and Jammu and Kashmir. One hundred and forty-six victims, or 19.4 per cent of the sample, were identified by the police as Muslims. Given that Muslims make up 13.5 per cent of the Indian population, it would seem clear that they are disproportionately in danger from the police weapons.

Misleading

A close study of the available data, though, suggests that this conclusion would be misleading. For one, the bulk of the killings have not taken place in the States most often accused of communal bias: Gujarat, Andhra Pradesh, Maharashtra and, more recently, Delhi. Gujarat saw just five police firing deaths in 2005, 16 in 2006 and one in 2007. Delhi registered just eight during the same period. Andhra Pradesh saw high numbers of killings, but mainly of Maoist insurgents of Hindu origin. Instead, an overwhelming majority of killings of Muslims by the police took place in Uttar Pradesh — a State where they make up 18 per cent of the population, not dissimilar to their share of deaths in police firing. The Uttar Pradesh police offensive, targeting violent organised crime, has claimed hundreds of lives in recent years — of Hindus and Muslims. In 2007, the last year for which the NCRB figures are available, the Uttar Pradesh police accounted for 102 of the 250 civilian lethal force fatalities nationwide. By way of contrast, the police fire in Andhra Pradesh led to the loss of 30 lives, while Maharashtra registered 27 deaths. Rajasthan reported 22 fatalities, most of them during caste riots. In 2006, Uttar Pradesh saw 103 fatalities, second only to insurgency-devastated Chhattisgarh. And in 2005, it recorded 42 deaths, placing the State third in police-firing fatalities after Andhra Pradesh and Jammu and Kashmir.

Nationwide, half or a lesser number of civilian fatalities in police firing were the outcome of counter-terrorism operations — and the ratio has been declining steadily. In 2005, counter-terrorism operations accounted for 46.76 per cent of civilian fatalities in police firing. In 2006, the figure rose to 52.12 per cent. The NCRB figures show that in 2007, though, just a quarter of civilian fatalities in police firing — 54 of 252 — were linked to counter-terrorism.

Put simply, there is no evidence to support the claim that there is an increased incidence of extra-judicial executions of Muslims — or, for that matter, Hindus. Even though police forces across India have intensified intelligence-led operations targeting Islamist groups, the NCRB data for 2007 show a sharp decline in the use of lethal force. A large part of the decline came because of a dramatic decline in killings by the police in Chhattisgarh, where fatalities fell to seven. Andhra Pradesh also saw a sharp decline in police killings, from 72 to 45. Only in Uttar Pradesh did deaths caused by the use of lethal force remain at the 2006 levels.

By global standards, the use of lethal force by the police in India is relatively low. Figures published in 1987 show that the police in Dallas, Texas, killed 1.03 people per 1,00,000 population the previous year. San Diego was next, with 0.83 people killed per 100,000, followed by Los Angeles with 0.71 deaths. Far from being trigger-happy, these figures suggest, India’s police forces are extremely cautious in resorting to lethal force.

Communal bias

What these figures point to is a slow but sure process of transformation: for which the social transformation brought about by democracy deserves credit. Less than a decade ago, the police forces across India faced credible charges of communal bias. Reports of judicial commissions, which investigated the 1982 riots in Meerut, the 1978 riots in Aligarh and the 1992-1993 carnage in Mumbai, showed systematic anti-Muslim biases in everything from the use of lethal force and patterns of arrest to the treatment of prisoners.

New studies, though, have thrown up signs of change. In January 2005, the Senior Superintendent of Police, Saharanpur, Safi Rizvi — now an aide to Union Home Minister P. Chidambaram — conducted a study of the district’s prison population. He sought to test the proposition that the police were disproportionately likely to act against Muslims and backward caste suspects. Mr. Rizvi’s study, however, demonstrated that the prison population of Saharanpur closely matched the district’s demographic profile. Hindus made up 58.5 per cent of the jail population, closely mirroring their overall share in the district population. Muslim prisoners accounted for 39 per cent of the jail population, marginally lower than their demographic representation. While Dalits made up 21 per cent of the district population, they constituted just 19 per cent of the prisoners; Brahmins, in a twist, were somewhat over-represented in jail.

Class, more accurate

Rather than religion or caste, Mr. Rizvi concluded, class constituted an accurate marker of which sections of the population were over-represented in prisons. More than 84 per cent of the prison population, he found, was made up of the poor — more than twice their share of the general population, as determined by the National Council for Applied Economic Research. It wasn’t, Mr. Rizvi noted, that the poor were more likely to steal: “the fact is that the poor criminal is promptly sent to jail for stealing 5 pieces of iron from the rail yard, one bicycle or pick-pocketing Rs. 50. He goes to jail for these crimes and stays there — unable to afford a lawyer, sureties or patronage.”

More studies are needed to see if the data from Saharanpur reflect national trends: anecdotal evidence suggests that Muslims are still significantly over-represented in the prison populations of Maharashtra and Gujarat. But if Mr. Rizvi’s findings are borne out by subsequent studies, it would suggest that Muslim and Dalit voters have become adroit at leveraging the political process to avoid victimisation. Police officers, the decline in police-firing deaths also shows, are increasingly sensitive to the costs of the indiscriminate use of force. Large-scale violence, or resort to extra-judicial executions, is no longer possible without inviting protest — and political or judicial censure. By contrast, Uttar Pradesh’s anti-crime killings have continued apace because the police are acting against groups which challenge the influence and authority of mainstream politicians.

Police forces everywhere in the world reflect the biases of the societies which give birth to them. It ought to surprise no one that some police officers in India have communal prejudices. The good news for India is that democracy appears to be making it ever more difficult for bigots in uniform to act on their beliefs.

Recommended for you