Source: Getty

Fighting Insurgency With Politics: The Case of Bihar

Bihar shows how particular political conditions cause states to be poor, weak, and violent—and how careful application of political tactics can reduce violence even in places with few resources and low state capacity.

by Rachel Kleinfeld and Rushda Majeed
Published on June 9, 2016

Poor and weak countries plagued by violence seem to face a chicken-and-egg problem: a lack of resources appears to constrain their ability to fight violence, while violence itself exacerbates poverty. Yet under Chief Minister Nitish Kumar, Bihar, one of India’s poorest states, was able to significantly reduce an insurgency that has plagued the region for over forty years. Bihar shows how particular political conditions cause states to be poor, weak, and violent—and how careful application of political tactics can reduce violence even in places with few resources and low state capacity.

Key Insights From India

  • Like many places facing violence from multiple groups, Bihar’s state weakness, poverty, and violence were all rooted in a power structure that enabled extreme privilege and impunity for an elite few, politicized security forces, and outsourced the monopoly of violence to militia groups. Thus, Maoist insurgents, a multitude of militias, and state security forces were locked in a vicious cycle of retaliatory violence.
  • Other Indian states failed to substantially reduce Maoist violence because counterinsurgency and counterterrorism strategies could not work in politically compromised governments that didn’t address the root problems of power.
  • Bihar successfully combined counterinsurgency and counterterror tactics with a strategy that redressed the power equation.
  • A failure to institutionalize Bihar’s rule of law successes may be threatening gains. Sustainable reform needs a three-part program comprising policy, politics, and institutionalization.

Implications for Counterterrorism and Counterinsurgency

  • Countries facing multiple types of violence simultaneously are often not just weak, but complicit. An illegitimate power structure based in extreme privilege and impunity may be fueling the violence.
  • In cases of such privilege violence, counterterrorism and counterinsurgency policies alone will fail. Counterterrorism assumes the existence of a functional state and can deepen the inequitable order by strengthening otherwise unchanged power relations. Counterinsurgency strategies assume that the poor can be turned toward the state, which fails when the state is seen as illegitimate.
  • A political problem requires a political solution. Without measures to address the causes of violence inherent in power structures, results will be fragile.
  • Elite pacts can be useful instruments to secure short-term reductions in violence. But they only buy breathing room; they are not solutions. The state must use the time gained to institutionalize measures that address state capacity and inequities in access to power; failing that, the state will return to violence.
  • When violence is compounded, addressing one strand of violence at a time will not work. Instead, policymakers must advance a multipronged approach that tackles different strands in conjunction with each other.


On November 14, 2005, nearly 200 Maoist guerrillas stormed a district prison in Bihar, India, just 30 miles from the state capital of Patna. After a fierce gun battle, they freed their counterparts and killed or captured members of the Ranvir Sena, a vigilante militia that had been fighting them for more than a decade.1 Coming on the heels of the 2005 state assembly elections, the attack was a direct challenge to the ruling order. A new chief minister elected on a law-and-order platform was taking office in just two weeks.

Violent Maoist insurgents have plagued India since the late 1960s. By the turn of the millennium, the Maoists, also known as Naxalites for their origins in the village of Naxalbari, were fighting in nine Indian states. The fighting was serious: Maoists killed more than 8,500 people from 2002 to 2016.2 In 2010, the bloodiest year to date, Manmohan Singh, India’s prime minister at the time, declared the Maoists the single biggest internal security challenge the country faced.3

In Bihar, however, Maoist guerrillas were only one among many perpetrators of violence. When Nitish Kumar took the reins as chief minister in 2005, he faced a series of groups locked in a vicious cycle of what could be called compounded violence—multiple types of violence where fighting between armed groups and linkages among political, criminal, and state violence served to entrench bloodshed.

After Maoist violence began, the state decided to allow landholders to arm themselves, privately and legally.4 Thus, it is no surprise that vigilante militias, generally associated with upper- to low-caste landlords, sprang up to fight the Maoists, who were generally landless day laborers. The militias massacred men, women, and children and raped women in areas associated with Maoist violence.

Meanwhile, law enforcement had itself become corrupt and complicit. Police in Bihar tended to assist or offer impunity for militia massacres, while pursuing and arresting Maoists after reprisal murders. So-called encounter killings (extrajudicial executions) substituted for serious investigations in a state in which slow, overburdened courts rarely brought perpetrators to justice.5

The complicity of the state security forces in helping the upper castes maintain their dominance deepened Maoist resolve and recruiting. Tit-for-tat violence among the three groups, and endemic criminal violence, had turned Bihar—a state with a population larger than Germany’s—into a place where people did not venture out after dark. By 2005, the situation had grown so bad that people would keep their guns visible through their car windows to deter attacks, even in the middle of the day.

Yet by 2010, just five years after the Naxalite prison break and Kumar taking office, Maoist violence in Bihar had dropped to less than half the levels of 2001–2005, with a threefold fall in civilians killed, and a vast increase in the state’s recovery of explosives, land mines, and cash (see table 1).6

States facing Maoist violence in India have had different responses to the challenge. The central government has touted the response of Andhra Pradesh as the model to be replicated across India. But its strong counterinsurgency tactics—focused on winning the hearts and minds of the people in order to reduce insurgents’ base of recruits and support so that security forces could fight a smaller, less entrenched enemy—rested on intelligence collection, a well-equipped set of technocrats in the state, and a political class that had already decided to get serious about the fight. Attempts to replicate the security portions of the strategy alone in other states have been unsuccessful.

Jharkhand, a state carved out of what had been Bihar’s southern half, is a better natural experiment—it followed the same policies as the rest of Bihar until the national government forced Bihar and two other states to split in 2000 for reasons of national- and state-level politics after years of competing demands. Like all natural experiments, it is not perfect: for instance, Jharkhand inherited nearly all of Bihar’s tribal population, a poor group that frequently supported Maoists. It also inherited more of the industry that served as bait for Maoist extortion. However, it is not clear that Jharkhand’s problem was much larger than the remaining rump Bihar. What is clear is that, to date, Jharkhand’s strategy for fighting Maoists, a counterterrorism-based effort primarily aimed at killing and capturing terrorists, has failed.

In many international contexts, a poor state such as Bihar would be helped through security assistance to strengthen weak security agencies. In India, the role of external security assistance provider is largely played by the national government, which offers equipment and funds to help different states (see appendix). Bihar benefited from such external assistance. But so did neighboring states that did not perform as well against insurgents. Bihar’s solution was accomplished with fewer or about the same national development and security resources as were expended by neighboring states fighting Maoists, despite an insurgency that was at least as serious as its neighbors were facing.7 The central government provided three pots of funding for fighting Maoists. While accurate numbers are unavailable for the two smaller programs, Bihar’s share of the largest effort, the Security Related Expenditure Scheme, is a fraction of the amount provided to neighboring states facing a Maoist challenge (see figure 1).8 Extra funding was clearly not the decisive factor.

Bihar thus provides an interesting set of lessons for other countries facing insurgent violence where the state security agencies and bureaucracies are weak, complicit, and underresourced. Instead of targeting only the violence of insurgents, Nitish Kumar astutely worked to fight the multiple prongs of Bihar’s compounded violence. His administration used a combination of tactics that

  • employed elite bargains to defang violent vigilantes, removing both their violent militias and a catalyst for Maoist recruitment and reprisal killing;
  • took advantage of the breathing room created by the elite deals to provide greater political power, instead of simply material rewards, to potential supporters of insurgency, giving them a nonviolent means to gain power that cost the state nothing; and
  • addressed the remaining problems through counterterrorism efforts that targeted insurgent fighters and counterinsurgency tactics to further reduce their supporters.

Bihar’s speedy success points to the power of elite pacts to quickly reduce violence. However, it also points to the limited duration of elite negotiated settlements. Indeed, structural problems with Kumar’s political solution may already be enabling insurgent violence and reprisal to rise again, despite sustained ongoing counterinsurgency and counterterrorism tactics.

The success of carefully implemented political tactics in quelling insurgency provided an opening and was a necessary first step in helping a weak state fight violence. Keeping the violence at bay requires moves that Bihar has not yet undertaken: depoliticizing, professionalizing, and improving the ability of the state’s bureaucracies and security agencies. Many international development interventions start with the latter step—missing the importance of getting the power structure right before this capacity building can succeed. Conversely, many diplomatic interventions focus on the political bargains, ignoring the need for shoring up the state. The case of Bihar points to the ways in which these two strategies must interact for success.

Why Counterinsurgency and Counterterrorism So Often Fail

When a state fights an insurgency, the international community tends to see the violent guerrillas as the problem. Foreign assistance is typically predicated on a view that the state’s failure to fight effectively is caused by weakness, poverty, and a lack of training among state security agencies. Thus, foreign governments provide assistance to state security agencies to help them gain the skills and equipment they assume are needed to fight effectively.

But in most cases of insurgency, multiple groups are wielding violence—and the state itself is frequently complicit in extrajudicial brutality. Treating insurgent violence as separate from these complex interaction effects, and treating the state as a neutral and legitimate arbiter, is unlikely to be effective. Capacity building can do some good around the margins, but until the problems with the power structure are solved, it can’t address the central problem.

In most cases of insurgency, multiple groups are wielding violence—and the state itself is frequently complicit in extrajudicial brutality.

Insurgencies do not arise out of thin air: they capitalize on grievances within the state. A large greed versus grievance literature considers which of these causes is primary for successful insurgencies.9 But even those guerrilla movements that are more criminal than ideological depend on grievances to catalyze supporters. They require a portion of the population to see the government as illegitimate. Thus, substantial segments of the population in states confronted with successful insurgencies see the country’s core political settlement—the group that wields power in the state, those who have access to decisionmaking and economic resources—as illegitimate. And that political order problem poses a challenge to both counterterrorism and counterinsurgency techniques.

Bihar is one of many places demonstrating a pattern that can be termed privilege violence, or violence that is enabled by a power structure that supports impunity for highly privileged elites. It typically follows a particular pattern with three parts. First, political and economic elites wish to maintain extreme privilege (often including corruption) and impunity or above-the-law status for themselves and their close families or associates. Second, to do so, they politicize law enforcement to maintain their impunity, and often support violent groups that protect their rule. Police become brutal and capricious, particularly to groups without political protection, while allowing the violence of groups supported by elements of the state. Third, marginalized or unprotected citizens start seeing the state as illegitimate. They turn to violent groups for protection. Bihar is one of many places demonstrating this pattern of privilege violence.

In states where many politicians are corrupt, or where crony-capitalist relationships mean that small numbers of people capture an extreme amount of the state resources, law enforcement tends to be politicized. After all, politicians need to ensure that they can get away with corruption without being arrested by honest, well-resourced, and skilled police investigators. They may want to use the police to harass opponents, or to ensure that the violent goons who help them win elections have impunity for their crimes. And in many countries with high levels of government corruption, politicians use the police as a holding pen for patronage jobs to reward loyal supporters.10 Keeping police weak and subject to party whims, rather than laws, is the easiest way to ensure that political parties, not laws, are empowered. In countries where the military is frequently used to keep domestic order, these realities also tend to affect the armed forces.

But in such countries, counterterrorism tactics to quell insurgency are hard to institute. Corrupt or politicized security agencies that are purposely kept weak and poorly trained can’t execute operations well—no matter how much foreign training and equipment they are given. In Nigeria, military units sell vehicles, bullets, and armaments to the Boko Haram guerrillas they are supposedly fighting. Attempts by the United States to create well-trained specialist units floundered when corruption led to the training of security personnel who paid their way into the course, rather than organic units that would fight together later. In Colombia in the late 1990s, police outposts were repeatedly captured by guerrillas who were better armed and had more will to fight. This occurred despite on-again, off-again security assistance from the United States since the mid-1960s. In Burundi, despite years as a recipient of U.S. funding for training and equipment, the political leadership still threatened to use the police to destroy political opponents in 2015, borrowing language from neighboring Rwanda’s genocide.

Counterterrorism tactics to quell insurgency are hard to institute. Corrupt or politicized security agencies that are purposely kept weak and poorly trained can’t execute operations well.

In India, the problem was no less daunting. The National Police Commission wrote in 1980: “We were repeatedly told by different sections of the public about police malpractices that were becoming increasingly oppressive and extortionist in character.” The police actions it cited included a range of corrupt practices, from fabricating false evidence, to colluding with smugglers, to extorting bribes through threats, searches, and prosecutions. The commission went on to say:

The reward and punishment mechanism of the system has become totally ineffective because of increasing political interference and, therefore, the senior officers, however, determined and committed they might be to the cause of anticorruption work, find themselves unable to deal with corrupt officers who have political contacts and are able to draw political intervention on their behalf whenever anything is attempted to be done to discipline them. The patent inability of a superior officer to deal with a known corrupt subordinate immediately lowers his prestige in the department and induces other subordinates also to seek and develop political contacts as a protective cover to escape punishment for their malpractices. . . . The problem of police corruption cannot, therefore, be satisfactorily tackled unless these [political] links are broken.11

The situation was no different thirty years later. Most state strategies against Maoists have faltered because of misuse of funds, political interference, and poor coordination—hallmarks of a failed political order that cannot implement counterterrorism or counterinsurgency techniques.12 In 2014, the central government began requiring states to provide photographic verification of projects that they had claimed to undertake in order to apply for future funding through the Special Infrastructure Scheme,13 and it cut reimbursements from 100 percent to 75 percent—most likely because of its belief that states were misusing funds.14

Pouring more funding and training from national or foreign sources into security forces simply will not work when the problem is political complicity in violence. Those individuals in weak security forces who happen to acquire skills and training will encounter bureaucratic obstacles placed in their paths. Donated equipment will be sold or hoarded by powerful factions, rather than used to fight. Trained units will be taken off politically sensitive assignments where the government wants to maintain complicity. And brutality will accompany the policies, spawning backlash and deepening insurgency. While attempts at professionalization at upper levels of the officer corps can create seeds that can grow when soil is more fertile, large-scale capacity building will not solve the problem.

But ineptitude and inefficiency may be the least of the problems of politicization. In all countries, politicized police also tend to be brutal police. The problem is not simply a lack of training in human rights laws and equipment to enable better human rights practices and less coerced confessions. When police know that the political class is corrupt and they are powerless to act against it, and that criminals connected to that political class cannot be arrested, they become frustrated and demoralized.15 In countries with alternative employment options, the human capital of the force alters, as those who wish to serve the law steer clear of jobs in the police force, while those willing to play a role in such a system seek law enforcement jobs. For those who retain a sense of duty to enforce the law, brutality can ironically be an outlet for that frustration, and a way of ensuring some sense of retribution for lawbreaking. As the notoriously violent and corrupt Sergeant “Clubber” Williams explained when working under the corrupt Tammany Hall administration in New York City in the early 1900s:

Now, let me tell you something. They may beat you in court, the complainant may not show up, they may jump their bail, politicians may interfere, there are several ways they can beat you, but this [he pointed to the marks and bruises on someone “worked over” by the police] they’ve got, and make no damned mistake about it.16

More than a century later, India showed the same proclivities. The state of Andhra Pradesh, whose counterterrorism tactics have been held up as a model by the Indian government, is embroiled in an investigation of some 285 reported cases of so-called fake encounter deaths allegedly committed by the Andhra police in connection with anti-Naxalite operations.17 Andhra Pradesh’s establishment of an armed vigilante group known as the Green Tigers is a common strategy to create plausible deniability while enabling state-assisted brutality.

Similarly, the state of Chhattisgarh began to arm and support the Salwa Judum vigilante movement in 2005.18 Allegedly, the movement was created by locals tired of Maoist violence, but a member of the Legislative Assembly from the Congress Party led it, hinting at far more politically oriented ties.19 That year, the state government also enacted the Special Public Security Act, which permits detention for up to three years for loosely defined unlawful activities. Human rights groups voiced concerns that the law criminalizes even support provided to Maoists under duress. Predictably, when the paramilitary militia conducted raids and cracked down on Maoists, the result was to herd 150,000 to 300,000 villagers, the vast majority of whom were not Maoists, into displacement camps. In 2011, the Indian Supreme Court ordered the movement banned and its weapons confiscated because of severe human rights abuse. But it has spawned organizations such as the Vikas Sangharsh Samitiformed by the son of the original militia’s principal architectthat are likely to follow in its footsteps.20

If counterterrorism fails because of corruption, politically motivated weakness of law enforcement, and brutality that can enhance insurgent recruitment,21 counterinsurgency fares no better.

Counterinsurgency tactics depend on turning the population toward the state and away from guerrillas. But doing so requires the majority of the population to see the state as more legitimate than the violent actors. That is difficult to do when elites demand impunity and above-the-law status, and when the state provides extreme privilege that clearly serves some of its population better than others.

Fairness is an inherent and deeply coded cross-cultural trait. In experiments repeated since 1978 with different populations throughout the world, people routinely choose to sacrifice potential material gains when an offer to divvy up rewards is perceived as unfair and providing another with a superior amount—even when the alternative is receiving nothing.22 Think of what such a reaction means for counterinsurgency efforts such as development spending and cash-for-work programs in countries where the government is corrupt or the political settlement determining who has power, who gets a seat at the table to make decisions, and who receives state resources is deeply biased toward one group.

The job of turning disaffected, marginalized parts of the population toward the state is hard when politicians and bureaucrats are seen as corrupt and unworthy of legitimacy.

The job of turning disaffected, marginalized parts of the population toward the state is hard when politicians and bureaucrats are seen as corrupt and unworthy of legitimacy. And it becomes nearly impossible when security agencies are brutal. In these cases, insurgents and criminals can paint themselves as modern-day Robin Hoods who are fighting the dominant order. Marginalized parts of the population tend to see these violent actors as at least as worthy of respect as a corrupt, brutal, and inequitable state. At best, they are indifferent between violent insurgents and criminals and the state. At worst, they prefer the insurgents and other violent actors, who may at least provide some succor, material help, or sense of dignity.

It is no surprise, therefore, that Chhattisgarh’s counterinsurgency efforts failed. For instance, one of the tactics the state used was to improve schooling to reduce the insurgents’ recruiting base and win hearts and minds. In a way, it succeeded: after it established residential schools for 12,000 tribal students, the share of out-of-school children ages six to fourteen declined from 50 percent in 2011 to 13 percent in 2013.23 However, as one would predict from the fairness studies, improved schooling does not seem to have had much effect on the insurgency. Maoists continue to hold sway over vast areas. Battles and violence have not been reduced. And the government’s legitimacy has declined following its human rights abuse. Interestingly, Bihar also used the counterinsurgency technique of improving schooling for Dalits. But by allowing local Dalit politicians to distribute educational posts, the government did so in a way that empowered Dalit adults and so affected the core power structure of the state.

In countries facing compounded violence—criminal, political, organized crime, and state brutality—the power structure is nearly always at the root of the problem. And the biased, corrupt, or inequitable structure impedes both counterterrorism and counterinsurgency tactics from working. The issue is not just a lack of resources or knowledge. In fact, ironically, foreign efforts to provide security sector assistance and capacity building can deepen the inequitable structure by strengthening an otherwise unchanged set of power relations. Foreign security assistance can eventually professionalize a force—and, starting with senior officers, it can affect attitudes. But it is rarely able, on its own, to overcome these internal dynamics and create a government able to beat rebels.24

So what can be done in such states to quell violent insurgency? Bihar provides a success story of how carefully applied political tactics can be used to quickly reduce violence in poor states with weak, compromised bureaucracies and security agencies. It also provides a cautionary tale of the need to use the breathing room bought by such political deals to strengthen the state, or risk a violence rebound.

The Vicious Cycle in Bihar: Insurgent, Vigilante, and State Violence

Maoist violence is not new to Bihar. India has grappled with the insurgency since 1967, when violent fighters took root in West Bengal’s Naxalbari village.25 By the early 1970s, the fighting had spread to a number of northern states, including Bihar.26

Like most insurgencies, Maoists in India drew strength from disaffected people who were preyed upon by other citizens and had little recourse from the government. Recruits varied somewhat from state to state: they came from rural masses in Andhra Pradesh and student groups in Kerala. In Bihar, Maoists organized bonded laborers, villagers, and tribal communities that tilled the land. They were attempting to right the vast inequalities they saw resulting from historic patterns of class and caste overlap in Bihar that had allowed a feudal power structure to persist into the twentieth century.

Bihar’s vast inequality and peasant poverty was a centuries-old problem. Since the mid-1700s, the British East India Company’s policy in northern India was to administer indirectly through the use of large landholders, known as zamindars, who were empowered to collect taxes and given judicial powers. After the great famine of 1770 and other social disasters rooted in this policy, the British decided to offer the Permanent Settlement by which zamindars were asked to provide a fixed amount of taxes to the British authorities. The measure was intended to encourage zamindars to improve their land and keep the additional profit, rather than squeezing taxes out of those already destitute. The result, however, was to turn zamindars into a new political class: feudal lords with vast landholdings worked by peasant laborers. The problem was exacerbated by India’s ancient system of caste, a religiously based employment hierarchy, so that upper class and upper caste merged.

After India gained independence in 1947, the country officially ended the zamindari system. But unlike West Bengal and Kerala, Bihar had failed to implement comprehensive land reforms, so most land remained in the hands of the upper-caste lords of the land. And while India officially abolished the practice of untouchability, in reality, Bihar continued to have landlords who exercised near-feudal power over both low-caste and so-called untouchable, or Dalit (whose status is so low that they are officially outside the caste system), laborers traditionally connected to their land.

After India gained independence, Bihar continued to have landlords who exercised near-feudal power over both low-caste and so-called untouchable, or Dalit, laborers.

Subject to caste-based humiliation, rape, and almost ritualized violence meant to enforce the power hierarchy, these low-caste agricultural workers did not share the benefits of democratic India. Despite a 1948 national minimum wage act, in many districts of Bihar wages ranged from a few kilograms of rice (a traditional feudal payment) to up to 35 rupees a day ($0.88), a fraction of the mandated rate.27 Often, workers became bonded laborers when they incurred debts to their landlords that forced them into de facto slavery. They would be required to work to pay off the debts, but their wages were never quite enough to release them from their bond.

In Bihar, the repeal of the zamindari system, along with the Green Revolution, in which technology turned India into an agricultural powerhouse, had the ironic effect of empowering low-caste groups that were not quite as badly off as the Dalits. As politics opened in the 1970s, these low-caste Yadavs and others gained political power, as well as access to landholdings of their own. The Yadavs and other low-caste landholders mimicked the upper castes in enforcing the economic conditions of Dalits through violence. Dalits viewed these landholders as even less legitimate than the upper-caste landlords, whose power at least had religious and traditional sanction. Now Dalits faced multiple caste and class groups intent on keeping them in caste and class misery.

By the 1960s, Bihar’s Dalit laborers had grown tired of being treated as near slaves in a supposedly democratic and independent country. Soon, Bihar was caught in a vicious cycle of violence as the caste hierarchies gave rise to an armed caste conflict. As Maoists began organizing Dalits to stand up for their rights, their goals of equity clashed with landlords’ control over land and resources. In 1971, in Purnea Village, fourteen Dalits were killed and their village destroyed for speaking up to landlords. Allegedly, the family of the speaker of the Bihar Legislative Assembly and member of the then-dominant Congress Party, L. N. Sudhanshu, was involved—a pattern of high-level political support for anti-Maoist paramilitary action that would continue.28 From this early, unorganized pogrom-style violence, landlords formed a variety of militias, or senas, each tied to a different caste. Reprisal violence began.

Naxalite gains at first seemed short-lived. In the shadow of the 1971 India-Pakistan War, paramilitary and police operations crushed the movement in West Bengal. While Maoism grew in other parts of India, the movement was forced underground during Emergency Rule, between 1975 and 1977, when then prime minister Indira Gandhi suspended India’s rule of law and undertook state repression and widespread campaigns of imprisonment. In 1975, Bihar’s government allowed landlords to register firearms for use against Maoists, creating armories that would eventually be used by vigilante militias. The Maoists found succor in neighboring Nepal and Bangladesh (East Pakistan at the time) or in the jungles of the states where they operated, whose tribal peoples were often sympathetic to their cause.

But with the basic political order of the state unchanged, brute repression did not succeed in stamping out the insurgency. When the Janata Dal (People’s Party) beat the Indira Gandhi–led ruling Indian National Congress at the polls and ended the emergency legal order, the released political prisoners regrouped and the insurgency reawakened.29 The restoration of democratic rule of law, absent a change in the structure of which social groups held power, simply reinstated the violence.

By 1994, most of the upper-caste militias had merged to form the dreaded Ranvir Sena. But other militia groups also thrived, supported by the Kurmis and Yadavs (so-called backward castes, traditionally disadvantaged groups a notch or two higher up in the caste hierarchy than Dalits, but still relatively low). These castes formed the Bhoomi Sena and Lorik Sena, respectively, ostensibly to fight Maoists, but often to simply enforce their will on their laborers.

These militias massacred and looted Dalits and other marginalized communities. The Ranvir Sena is believed to have killed more than 400 Dalits in rural Bihar between 1995 and 1999. Maoists slit landlords’ throats and tried to match killing for killing. The tit-for-tat violence against civilians knew few bounds. As one Ranvir Sena leader told Human Rights Watch about the killings: “We kill children because they will grow up to become Naxalites. We kill women because they will give birth to Naxalites.”30

Meanwhile, the state’s response was heavy-handed. The central government of India, as well as state governments, had started to regard Naxalites as an internal security threat and clamped down. Extrajudicial executions, disappearances, and torture of accused Naxalites became common. Security forces targeted sympathetic tribal and Dalit communities, conducted raids on villages and hamlets, and used violence against women. Between 2001 and 2005, 1,309 Naxalite-related violent incidents, 760 civilian deaths, and 141 armed encounters were reported in Bihar.31

Policemen—often from upper-caste communities—routinely refused to lodge complaints from Dalits or open investigations, instead threatening the victims. In some instances, they abetted the senas. In April 1997 in Ekwari, Bhojpur, for example, the police helped the militia break into villagers’ homes and stood by as the massacre unfolded. Although many Naxalites were killed in these so-called encounter killings, no senamember died as a result of police action.32 Police swiftly released militia members who were arrested. Multiple studies, including one by Human Rights Watch, have documented a pattern of state collusion and police complicity in the sena attacks.33

Policemen—often from upper-caste communities—routinely refused to lodge complaints from Dalits or open investigations, instead threatening the victims.

For decades, politics had inflamed the vicious cycle of caste violence in Bihar. In its early decades, India was ruled by the Congress Party, which claimed to govern on behalf of all of India but was dominated by upper castes. In 1977, Emergency Rule was brought to an end, and the national government, Bihar, and other states in India elected their first low-caste leaders. But in Bihar, the two-year experiment with low-caste leadership inflamed upper-caste members’ concerns—which they used violence to settle. That year, landlords burned to death fourteen Dalits over a land dispute, and landlord-led militias began to consolidate and deepen their organization.34

Tacit political support allowed the Ranvir Sena to operate with impunity even after the central government labeled it a terrorist organization and banned it in 1995. Its leaders boasted of carrying out massacres after the ban.35 A well-known party activist with the Bharatiya Janata Party (BJP) moonlighted as the head of the Ranvir Sena in Bihar.36

The low-caste and Dalit population hoped that succor had arrived in 1990, when Lalu Prasad Yadav, a leader of the Yadavs, was elected as Bihar’s chief minister on a social justice platform. However, rather than rebuilding a more just state, he chose to remain politically relevant by stoking caste tensions. Yadav antagonized the upper castes by chipping away at their traditional powers. He didn’t trust the bureaucracy, which was dominated by the upper caste. His response was to reduce the powers of the traditional upper-caste police by refusing to hire for open positions, since they would generally have gone to better-educated upper castes. In so doing, he rendered the police less and less functional. A thoughtful paper on the Yadav period was aptly titled “State Incapacity by Design: Understanding the Bihar Story.”37

Meanwhile, perhaps to keep caste tensions high to help his party, Yadav deliberately failed to act against the Ranvir Sena for violence against Dalits, enabling the Naxalites to continue to capitalize on caste grievances. While the state did little to pursue the militias, it actively fought the retaliatory violence inflicted on landlords by the Maoists, including invoking the now-repealed Terrorist and Disruptive Activities (Prevention) Act.38 Local analysts argue that Yadav’s unwillingness to protect the Dalits could be attributed to his allegiance to the Yadav landowning backward castes, on whose electoral support he depended.39 Moreover, tolerating upper-caste massacres of Dalits helped him keep the low castes angry at the upper castes, and therefore built support for his brand of caste-based politics.

What Worked in Bihar

Nitish Kumar inherited this situation when he became chief minister of Bihar in 2005 on the promise of restoring law and order. His solution included counterterrorism and counterinsurgency elements, which seem necessary to solve problems of insurgency but by themselves are not sufficient. As befits a problem whose roots lie in the power structure of the state, he also chose to address political factors.

Kumar benefited from some good timing. First, rural wages were set to increase just as he came to power, as a result of a central government program of guaranteed rural employment.40 Rural areas experienced higher agricultural productivity, leading to more demand for labor. Because of the consequent labor shortages, rural wages rose above the legal minimum wage.41 Now people in economically depressed zones had an alternative to the economic incentives offered by Naxalites.

Second, Naxalites had overstepped, costing them influence and support. As they increased the use of extortion and kidnapping to make money, they reduced the ideological clarion call that had helped them win recruits. Dalits began to resent the violence and extortion. The Maoist support base wavered—unable, yet, to turn toward a state that appeared illegitimate and corrupt, but no longer enamored of the solution offered by Naxalite cadres.

Nitish Kumar’s plan to restore law and order in Bihar included counterterrorism and counterinsurgency elements. He also chose to address political factors.

Kumar had a smaller problem on his hands, thanks to the national government’s decision to split the state into two portions. Jharkhand, formerly the southern part of Bihar, inherited a greater share of the tribal population, jungle territory, and industry (providing extortion opportunities), and therefore took a significant part of the Naxalite threat with it. However, the partition had been a reality since 2000 with no decline in violence in Bihar.42 Clearly, that alone did not account for the sudden reduction of violence.

What Kumar did differently was recognize this window of opportunity and capitalize on it. Crucially, Kumar realized he was fighting not just an insurgency. He knew he would have to clamp down on all three variables in the equation at once: addressing the violence of the landlords’ senas and of the state as well as that of the Naxalites to reduce the bloodshed caused by the latter. By decreasing the sena and state violence that provided an excuse for Maoist violence, Kumar focused on an element often ignored in both counterterrorism and counterinsurgency—that those caught in the middle may be facing violence from many sides. Naxalites provided some sense of security (or, at least, retributive justice) to Dalits and tribal peoples. The chief minister needed to get the senas and law enforcement under control so that the Naxalites could not pose as saviors who could retaliate in the face of an indifferent state.

Effective Policies

The usual way to tell the story of the “Bihar miracle,” as the press soon dubbed Kumar’s administration, is by listing the state’s successful policies to improve law and order and simultaneously spur development. Those policies were well implemented, thanks to a reformist leader who cared about change, wanted to make it happen, and was a solid administrator who demanded accountability. And the policies were important. In addition to their importance individually, they relied on being implemented all at once, in a manner that was holistic in addressing problems, and, equally important, kept insurgents on the defensive and unable to quickly regroup before they were fatally weakened.

Counterterrorism: Law and Order

Speedy, Certain Justice

Kumar’s team focused on quick wins first. The team made it clear to police, prosecutors, and judges that he wanted to reduce crime, and through a vigilant daily focus on outcomes, it helped different participants in the legal apparatus who rarely made common cause find ways to work together to improve justice. The innovations that the police, prosecutors, and judges devised together required no changes to laws, simply a new standard method of operating and leaders who forced greater accountability in each bureaucracy—particularly the police. Thus, the changes could begin immediately.

One of the changes took aim at firearms. In 2000, Bihar had just 8 percent of India’s population but accounted for 26 percent of India’s total murders with the use of firearms, and the relatively small city of Patna accounted for 40 percent of all murders with the use of firearms in Indian cities.43 The new administration decided to utilize the Arms Act, a little-known clause from 1959, which banned illegal possession of firearms. People caught with unlicensed arms could be convicted quickly because witnesses were almost always policemen, who were less likely to be intimidated and were more likely to testify. By using the Arms Act en masse, police, judges, and prosecutors became skilled at building simple, strong cases that allowed for speedy trials and quick convictions. (Police could, of course, plant false evidence, which likely occurred, especially given the focus on conviction numbers. However, false evidence and false police testimony were occurring well before the use of these speedy trial procedures in order to enrich individual police. It is not clear that the extent of misuse was increased through the speedy trials.)

Members of various militias—including those that enjoyed the protection of the ruling parties—were put on trial, as were Naxalites. In a short time, Bihar started achieving an average of 1,000 convictions a month, including 109 Maoist convictions in 2006 and 2007.44 By 2012, more than 80,000 criminals had been convicted under the Arms Act or other offenses. Most of those convicted faced quite short sentences compared to the extent of their crimes. But the sentences came swiftly after arrest, and those convicted were soon in jail, providing a deterrent that was not possible when long trials and multiple appeals allowed known criminals to roam free for decades. Meanwhile, some convictions were more significant: a few criminals proved to have engaged in massacres were handed death sentences.45

Surendra Kishore, editorial adviser to the newspaper Dainik Bhaskar, explained the importance of these Arms Act trials to the fight against Maoists: the mass killings stopped, he explained in our interview, “because of the convictions of both Maoists and the sena people . . . the top sena people and Maoists didn’t carry out the killings themselves but hired people [to do it]. If [most of] the killers are convicted, how are they going to find new people?”46

Wiretapping and Surveillance

The government made efforts to improve intelligence gathering as well.47 Previously, most intelligence came from urban centers, since police penetration into Naxalite areas was low. But now, carefully placed cell towers in Naxalite areas helped collect intelligence on the movement of cadres, which the police used to track and deter Naxalite attacks. Over time, with the help of intercepts and people who were now willing to turn informers, the police arrested top Naxalite leaders in Jehanabad, Aurangabad, and Gaya.48 As the counterinsurgency strategies bore fruit, people were more willing to turn to the state against Maoists and provide human intelligence.

As the counterinsurgency strategies bore fruit, people were more willing to turn to the state against Maoists and provide human intelligence.
Tracking Assets

The government also confiscated assets of Maoist leaders. Former director general of police, Abhayanand (who uses only one name), explained when interviewed, “Naxalites have an economy of their own. They collect revenue and spend a lot of money. They have an intricate [financial] web, no less complicated than the government machinery.” The police focused on implementing the Unlawful Activities (Prevention) Act, which included provisions for confiscating assets of Naxalite leaders. “Bihar was the first state to implement this provision of the act. [The Naxalites] were hassled more by its implementation rather than the sound of a gun,” Abhayanand said.49

The police tracked down other sources of revenue. For instance, in areas under their control, Naxalites extorted levies and taxes from brick shop owners and road contractors. The police closely tracked suspects and victims, making a concerted effort to check extortion.


Road projects bridged counterterrorism and counterinsurgency techniques. Bihar’s towns and villages were poorly connected, and it was difficult to reach many Naxalite areas from Patna. Contractors who took on public infrastructure projects in these areas had to pay up to 20 percent of the total project’s cost as levies to the insurgents. Even if they were able to complete projects on time, there were no guarantees that the Naxalites would not blow up a newly built road or bridge.

With funds from the national government pouring in for road development, the state roads department focused on choosing competent contractors through a competitive online process, and it monitored project milestones and timelines. The bridge construction agency built new bridges in Naxalite areas. Critical road infrastructure soon materialized across the state, connecting previously Naxalite-heavy areas. High-quality highways allowed security services to quickly reach Naxalite areas. They also allowed economic development and a civilian government presence to take hold in impoverished areas.

Counterinsurgency: Development

Kumar’s government initiated a series of textbook counterinsurgency measures. For instance, he undertook the Apki Sarkar Apke Dwar—Your Government at Your Doorstep—program in all 65 panchayats, or village councils, in Naxalite areas to improve public services and infrastructure in these villages. As Chanchal Kumar, the secretary to the chief minister, described it in our interview, “These areas were flooded with development schemes, and all departments were told to pour in money.”50

The Home Department first rolled out the program in Sikarayi, an area with top-ranking Naxalite leaders. Building roads, schools, health clinics, bridges, irrigation canals, even houses, for villagers were a high priority. The government also constructed panchayat buildings (basically town halls for each village) to serve as one-stop shops for public services. It ensured that civil servants would be available to hear grievances and offer solutions. And it emphasized the proper implementation of state and central government programs, such as the national employment guarantee scheme. Arun Kumar (no relation to others), inspector general of the Central Reserve Police Force in Bihar, said during our interview, “It is a credit to the state government that it has really enforced the Government of India programs. These programs are going down the chain to the local levels.”51

Kumar undertook projects with high symbolism and photo-op value that showed change was happening. His program to award cash for buying bicycles to schoolgirls in grades 9 and 10 with high grades was extremely popular and effective.52 It became one of the world’s largest cash transfer programs of its time, with cash for more than 900,000 cycles disbursed from 2006 to 2010.53 The government also provided school uniforms to children and expanded the central government’s midday meal program to encourage enrollment and attendance. But the bicycle program did more than just transfer funds: by sending legions of uniformed schoolgirls through the streets each day, it was a highly visible change that reassured people that areas were safer, and, in a virtuous cycle, made them safer by having more people on the roads and eyes on the streets.54

The highly symbolic, visual, and concrete programs meant that people in the poorest regions with the least infrastructure could see their government working for them. This built legitimacy for the state. With people angry at the Maoists for their extortion and level of violence, the state was now giving them reasons to trust the government. Thus, the counterinsurgency tactics helped the government gain informers who were willing to turn on the Maoists. Police and intelligence services were therefore able to infiltrate Maoist groups and weaken their hold.

Effective Politics

It is easy to tell the story of Bihar’s success with this mixture of counterterrorism and counterinsurgency policies—and clearly, they played a role. However, they don’t tell the whole story. These well-implemented tactics rested on political measures that were equally crucial.

Nitish Kumar had been in politics for decades without winning office. His problem, in caste-ridden Bihar, was that he lacked his own caste base: his group, the Kurmis, were too small to form a powerful voting bloc, and castes tended to vote for their own members. So Kumar decided to construct a power base by putting together a savvy mix of the highest and lowest castes of voters. He kept his electoral promise to the lower castes by creating a new group of Mahadalits—the poorest of the lowest castes—to receive state benefits. In doing so, he carved out a new constituency on the extreme low end of the caste ladder that formed the recruitment pool for Maoists. He married this group with an alliance with the BJP, which controlled an upper-caste Hindu constituency that was aggrieved by Lalu Yadav’s government’s help to his own caste members and their criminal impunity. Politically, Kumar had created a base that included all the castes except those corralled by his main opponent, Lalu Yadav and the Yadav caste. Thus, Kumar had a strong political incentive to curb violence: his voting constituencies demanded it, giving him the political structural support for his individual political will.55

His first steps were classic examples of elite-pact politics: buying off violent actors to persuade them to stop fighting. He immediately kept highly symbolic promises to his upper-caste Hindu base. In 2006, soon after taking office, he dissolved the Amir Das Commission, a body appointed in the 1990s to investigate the political ties of the senas that were preying on Dalits.56 He continued to reassure and appease his upper-caste voters two years later by suspending a commission he himself had created in 2006, which was meant to address land distribution grievances.57

In other words, Kumar tacitly offered impunity to the senas by wiping away past sins and reassured the upper castes that land reform was off the table.58 These moves encouraged the upper castes, which believed they now had a government in power that was responsive to their concerns—and therefore did not need to resort to violence to ensure their status and power.

Kumar tacitly offered impunity to the militias by wiping away past sins and reassured the upper castes that land reform was off the table.

As further incentive to lay down arms, Kumar provided some criminal-political leaders with political opportunities on his party ticket, and incentivized them to end their violent criminal activity in exchange for opportunities to make money off the state.59 As the state’s economic fortunes rose under Kumar’s law and order success and good public relations, former violent leaders diverted their focus to other lucrative businesses, such as real estate and construction, in which state regulations and contracts could help. Those roads built to help counterinsurgency and counterterrorism goals were constructed largely by upper castes that won state contracts and became extremely wealthy as a result.60 Kumar was replacing one set of rents with another—but in doing so, he was bringing down the amount of violence.

As Ashwani Kumar (no relation to others), professor of development studies at Mumbai’s Tata Institute of Social Sciences, said in an interview, “When Nitish Kumar came into power, these [senas] disappeared. Though occasionally former members of Sena flex their muscle power and display ideological hostility to [Naxalites] and Dalits, the Ranvir Sena has lost its appeal completely. Bhumihars said, ‘We are in power. Why do we need the Senas as “surrogate arms of the state”?’ And you don’t have to use violence, just use accommodation to make the senas disappear. Kumar is an astute politician because he carved out viable rainbow caste coalitions for governance and he also did not disturb continuing dominance of rural elites. Interestingly, the ‘sena warriors’ and their criminal elements found alternative employment and profit opportunities in booming road construction and its allied industry. Thus, a ‘weak state’ like Bihar achieved a rare success in crime reduction and containment of Senas without any major military or armed operation or social chaos.”61

If Kumar’s elite-pact activities had stopped here, he would have potentially reduced militia violence—but at the cost of state legitimacy. In other insurgencies, this plays out in a predictable manner: insurgent violence grows, often with a new group of insurgents who have been cut out of the deals. That in turn creates a new round of paramilitary militias with tacit state support. Unfortunately, this is precisely where many negotiated settlements to end insurgency and violence do stop—with agreements that placate the violence-wielding elites, warlords, and militia leaders of a country or region but that deepen the underlying political settlement that created grievances in the first place.

Instead, Kumar used the elite pacts to buy breathing room, giving him the political space to address some of the deeper inequities in the political order that drove Naxalite recruitment. Although land reform and naming those involved in sena violence were off the table, there were other ways Kumar could address low-caste grievances. In 2006, he introduced a bill reserving 20 percent of seats in local government for the extremely backward castes—the lowest of the low who formed the recruitment base of Naxalites and also, not coincidentally, were part of his voting base. Instead of relying on state giveaways at the whim of politicians at the top, the reservation allowed Dalits to gain political power and break the hold of the many castes above them in the traditional power structures without resorting to violence.62

Because state programs in India are frequently distributed through local governments, political power also means economic power. But crucially, these policies were not just economic giveaways operating through the same political settlement—they redistributed power. Some low-caste leaders would clearly benefit from this more than others by gaining a new set of patron-client relationships that ran through the lowest castes.

Now Dalits had a spot at the state trough as well and could dole out patronage posts and government contracts to benefit their group. For instance, Nitish Kumar wanted to improve education and healthcare. To tackle the shortage of public school teachers, as a temporary measure, he empowered the elected local government leaders to hire new teachers on short-term contracts, instead of recruiting them through public examinations, which would take time. The state began to spend 20–25 percent of its budget on education. Thus, the new, low-caste local leaders elected through the reservation system had corruption and patronage opportunities to give to members of their group.

The outcome of the educational and health spending was mixed on its merits. During Nitish Kumar’s first term, local governments hired more than 200,000 teachers and built 10,000 additional classrooms.63 Literacy rose 17 percent by 2011, while enrollment increased to more than 95 percent.64 Similarly, with better-staffed health clinics, the number of patients treated at a clinic rose from an average of 39 per month to 4,000.65 However, the locally chosen teachers, not surprisingly, turned out to be fairly poor at providing education. While enrollment was high, school attendance was low, as was comprehension. Meanwhile, health clinics provided mixed outcomes.66 Now, the state is turning to a second set of programs to better train those teaching and providing healthcare.67

The outcome in terms of political and economic empowerment for some Dalits was an unmitigated success, and one that brought this disenfranchised group into a new political settlement with the state. As Shaibal Gupta of the Patna-based Asian Development Research Institute said when interviewed, “Nitish Kumar started mainstreaming people at the margins. So, with this positive discrimination, the social base of the [Naxalites] started joining the [state] system at the lowest level.”68 And the new political power gained through democratic means resulted in fewer recruits for the Naxalites. Arun Kumar of the Central Reserve Police Force noted in his interview, the “reservation in panchayats has helped a lot. The Scheduled Castes and Scheduled Tribes were a chunk of the cadres, and that has really helped stymie progress and support for [Naxalites].”69

Missed Opportunities on Disarmament and Security Sector Reform

In most cases of insurgency, international assistance pours in to help with disarmament, demobilization, and rehabilitation, as well as security sector reform. Both are important to a lasting settlement. Neither took root in Bihar.

Many attempts to fight insurgency include amnesty or disarmament, demobilization, and rehabilitation programs. Strategies that encourage surrender are important for multiple reasons: they enable intelligence to be gained from departing cadres, which allows remaining guerrillas to be better targeted—a key element in many successful counterinsurgencies, such as Colombia’s. They also affect guerrilla morale, as per the old military strategy of leaving one flank open (an enemy enclosed on four flanks will fight, given no other option, but when pressed hard militarily and allowed an exit, many will choose to leave, reducing the appeal for those remaining). Andhra Pradesh used that strategy to offer attractive surrender and resettlement policies.

The outcome in terms of political and economic empowerment for some Dalits was an unmitigated success, and one that brought this disenfranchised group into a new political settlement with the state.

Bihar had a halfhearted surrender program that was not implemented seriously. The incentives offered were so low monetarily, and so poor at providing any impunity from jail time, that few cadres took part. The loss of this opportunity may be behind Bihar’s recent announcement increasing incentives for Naxalites who surrender.70 The failure to implement a serious disarmament, demobilization, and rehabilitation strategy, when funds were available from the national government and better programs were forged in neighboring states, is inexplicable.

Another lost opportunity was Bihar’s failure under Nitish Kumar to reform police behavior. Bihar certainly needed deep change in that arena. For decades, the police had failed to prevent kidnappings, extortion, murders, and caste wars. Lalu Yadav had not trusted the upper-caste police, and as a result, the police faced overwhelming capacity constraints. The government had not hired new recruits for the past fifteen years, and the force thus faced a shortage of more than 30,000 personnel. While the national average was low at 122 police officers per 100,000 people (the United Nations recommends 222),71 Bihar had only 57 per 100,000, though by 2015 Nitish Kumar was noting the number had expanded to 88.72 The average age of a police constable was thirty-eight years. Hiring would be time-consuming, however, because the state would have to administer public examinations and train recruits. Since Bihar had lost its police training center to Jharkhand in 2000, reconstruction of an entire educational apparatus would be required.

Kumar’s team hit upon a temporary measure to fill the gap. It began hiring retired army personnel on one-year contracts. This State Auxiliary Force was particularly intended to combat Naxalites in hard-to-reach areas. By 2015, the state had recruited 11,000 more constables, with additional hiring in the works to bring numbers in line with the national average. Kumar also increased the police budget and sanctioned extra funds for police stations. He focused on modernizing equipment and uniforms, and, in 2008, he approved funds for constructing a new police academy. In addition, Kumar’s team worked with senior officers to train and professionalize the police force.

The State Auxiliary Force (SAP) seems to have been helpful. As P. K. Thakur, the director general of police, put it during an interview, “We took them [ex-army men] in, and they worked out well. It gave us time to build up our own resources, and that helped us tremendously. It brought the system on track.”73 But reports were not universally positive. The Central Reserve Police Force’s (CRPF’s) Arun Kumar claimed that “initially, the [State Auxiliary Force] did fight Maoists, but their camps were looted and there was a huge loss of arms and ammunitions. So the state government fell back on the CRPF.”74 Anup Mukerji, the former chief secretary and permanent secretary for rural development, felt that the police “would rely on the SAP a lot. If they had to do an operation, they would keep the SAP with them. But the SAP also had a problem with indiscipline.”75 Yet the SAP was a stopgap. More permanent reforms did not take place. By 2015, the police school was not yet functioning. Most police stations were still not built in Dalit areas where they were most needed. Instead they were built in upper-caste parts of villages where Dalits were socially sanctioned from walking.76

While the failures on disarmament, demobilization, and rehabilitation and security sector reform have reduced sustainability, Nitish Kumar showed what leadership alone could achieve. He required a reduction in encounter killings and an upholding of human rights, for instance.77 Police may not have become more efficient or less corrupt. And they may have continued to look down on Dalits. But the Kumar administration insisted on reduced brutality and achieved it. It is always difficult to take reported cases at face value, as often reports rise when a more sympathetic or effective administration is in place and citizens believe it worth reporting. Nevertheless, the statistics on abuse appear to support a trend of improvement under Kumar (see figure 2).

Moreover, the increase in political representation likely meant that the police started to take reports of crimes against Dalits and women more seriously. Reports of rape have risen significantly, which many attribute to greater police attention to the problem and more willingness to report, as opposed to a greater incidence of rapes.78

For lasting change, the security apparatus will need to reduce corruption and improve efficiency. Professional police who uphold the law rather than the political whims of those in power are essential to locking in reduced violence. Without this professionalization, new political leadership can simply reverse the changes.

But Bihar offers hope for poor countries with weak bureaucracies whose political leadership want to make change: improvements in violence could begin and get quite far with the same police force under new leaders who insisted on results, maintained accountability, and were not corrupt themselves. And while Bihar was helped by national funds, these funds and programs were available to nearby states that did not perform as well. Funding and support may have been necessary, but were certainly not sufficient, to achieve a positive outcome.


Many pundits are starting to claim that Bihar’s fight against Maoism is faltering thanks to new political factors. Largely because of political differences with the current prime minister, Narendra Modi, Nitish Kumar jettisoned his alliance with the BJP, which cost him support among the upper castes. He therefore had to turn to his former colleague turned nemesis, Lalu Yadav, to forge a winning coalition. While it is too early to tell, if Yadav further politicizes the police force and provides impunity for his criminal cronies, the increasing delegitimization of the state may bring back increased violence.

This potentially unfortunate conclusion points to yet another lesson. Changing the power structure was a crucial way to open the door to reform in Bihar, but sustaining the changes by building up a weak state is equally important. Without institutionalization, political changes are subject to the caprices of individuals, no matter how committed and charismatic any individual politician may be. Many believe Kumar tried to institutionalize change during his decade in power—and now that he is in his third term, he could deepen his past reforms. At the same time, Yadav’s push for a greater say in governance in the coalition may make it difficult for the state to maintain legitimacy or carry forward the reforms. Only time will tell which force will be stronger.

Thus, Bihar provides a useful lesson to other countries. A vast amount of violence reduction can be accomplished while the government apparatus remains weak. But it also supplies a cautionary tale. Changes to politics provide a window of opportunity. Lasting change requires capitalizing on the window created by a new political deal, to rebuild a functioning state and a strong, meritocratic bureaucracy.

Appendix: Security Policy in India

In India, security is shared between the federal and state levels. The national government wants a uniform policy, so that insurgents cannot just cross state borders to avoid capture. Policies must be implemented, however, by states. And highly differential outcomes in different states show just how much depends on implementation and leadership, rather than available funds alone.

India’s national government has a twofold strategy that mixes counterterrorism law enforcement with counterinsurgency techniques based on improving economic development in Maoist-hit areas. In 2006, the national government created a department at the Ministry of Home Affairs to coordinate a bevy of overlapping programs and governance structures intended to tackle left-wing extremism in nine states, including Bihar. The unit implements security-related programs, coordinates the development programs of other ministries, and tracks progress across states.

On the counterterrorism side, the national government deploys the Central Armed Police Forces in the nine states.79 It also assists states with operational infrastructure and logistics. For instance, it provides funding for the insurance, training, and operational needs of state security forces; community policing; and security-related infrastructure, including funds to strengthen police stations in each state. Finally, it provides funds to rehabilitate Naxalites who surrender in accordance with state policies, and it offers financial support to the victims of Naxalite violence.

To support a counterinsurgency effort, the national government has an Integrated Action Plan meant to improve infrastructure and public services in 82 tribal and backward districts. Its civic action program funds small development schemes, along the lines of Afghanistan’s National Solidarity Program.80 And it coordinates the development programs under the 2006 Scheduled Tribes and Other Traditional Forest Dwellers (Recognition of Forest Rights) Act.

Finally, to help achieve both goals, the national government provides funds for roads to develop poor areas and to help security services reach Maoists so that they cannot regroup in far-flung areas. The 2009 Road Requirement Plan finances roads in 24 districts of the affected states, while the Special Infrastructure Scheme (2007–2012) includes 500 crore rupees ($113 million) to build and upgrade roads and bridges in Naxalite areas.81


1 Arun Sinha, Nitish Kumar and the Rise of Bihar (New Delhi: Penguin Books, 2011), 155.

2 Statistics on Maoist murders are from the South Asia Terrorist Portal, “Annual Fatalities Maoist in Graph: 2016,” last updated April 24, 2016, accessed May 2, 2016, The nine Indian states where the Maoists were fighting are Andhra Pradesh, Bihar, Chhattisgarh, Jharkhand, Maharashtra, Madhya Pradesh, Odisha, Uttar Pradesh, and West Bengal. In the core states of Bihar, Jharkhand, Chhattisgarh, and Andhra Pradesh, 90 percent of Maoist-related killings occurred in 2001–2008, according to statistics collected in a new data set by Devesh Kapur, Kishore Gawande, and Shanker Satyanath, “Renewable Resource Shocks and Conflict in India’s Maoist Belt,” Center for Global Development, Working Paper no. 302, August 2012.

3 “India: Ending the Red Terror,” Economist, February 25, 2010, accessed on April 5, 2016,

4 Avinash Kumar, “Illegitimacy of the State in Bihar,” Economic and Political Weekly 44, no. 44 (October 31, 2009): 8.

5 Smita Narula, “The Pattern of Abuse: Rural Violence in Bihar and the State’s Response,” chap. 4 in Broken People: Caste Violence Against India’s “Untouchables” (New York: Human Rights Watch, 1999),

6 “Can Growth Tame Maoists?,” Economic Times, April 1, 2013, A comparison of Naxalite-related incidents in 2001–2005 and Kumar’s first term (2005–2010) shows that violent incidents and the number of civilians killed fell dramatically. In addition, the government recovered a large amount of explosives and levy money. The number of Maoists killed did not change much, but five times as many Maoists surrendered.

7 Among the states addressing Maoist violence, Bihar had 22 districts considered hard-hit enough to qualify under the national scheme. Andhra Pradesh had only eight districts covered under the national scheme. Jharkhand, just to Bihar’s south, had 21 districts covered, and Chhattisgarh had sixteen. Central Paramilitary Forces were also unevenly distributed, with sixteen battalions in Chhattisgarh, six in Jharkhand, and just four in Bihar. See R. P. N. Singh, “Financial Grants to Naxal Affected States,” Ministry of Home Affairs, August 14, 2013,

8 Security Related Expenditure Scheme figures are compiled based on data from Indiastat, “Selected State-Wise Funds Released by Central Government Under Security Related Expenditure Scheme (SRE) in India (2002–2003 to 2009–2010),” accessed March 30, 2016,; and Kiren Rijiju, “Funds Under SRE to Maoist Affected States,” Ministry of Home Affairs, August 13, 2014,

9 Grievances are necessary to mobilize insurgents—but they are not sufficient, of course. Leadership, organization, terrain, and access to funding are required to keep the fight going. A few argue that grievances are simply window dressing for motivations that are primarily economic. See Paul Collierand Anke Hoeffler, Greed and Grievance in Civil War, World Bank, Policy Research Working Paper no. 2355, May 2000, However, most researchers who closely study guerrilla warfare, rather than economists conducting large n-table studies, believe Collier’s data may oversample from sub-Saharan African conflicts where grievances are ever-present and insurgencies are more criminal in nature. See, for instance, David Keen, “Incentives and Disincentives for Violence,” in Greed and Grievance: Economic Agendas in Civil Wars, eds. Mats Berdal and David M. Malone (Boulder, CO: Lynne Rienner, 2000), 19–43; and the collection of cases in Marianne Heiberg, Brendan O’Leary, and John Tirman, eds., Terror, Insurgency, and the State: Ending Protracted Conflicts (Philadelphia: University of Pennsylvania Press, 2007).

10 Although the police in India are part of a national civil service, law and order is a state function and the police are under state leadership, with the exception of senior-ranking officers who are under dual control. India’s police forces have been politicized since then prime minister Indira Gandhi’s Emergency Rule in the 1970s.

11 Indian National Police Commission, Third Report (New Delhi: Government of India, 1980), chapters 22.6–22.7,

12 Sameer Lalwani, “India’s Approach to Counterinsurgency and the Naxalite Problem,” CTC Sentinel 4, no. 10 (October 2011).

13 Through the Special Infrastructure Scheme (2007–2012), the central government allocated 500 crore rupees ($113 million) to build and upgrade roads and bridges in Naxalite-hit areas.

14 Vivek Deshpande, “Central Schemes: No Major Change for States Hit by Left Wing Extremism,” Indian Express, August 7, 2015,

15 Based on Rachel Kleinfeld’s forthcoming book, Back From the Brink (2017).

16 Cited in Thomas Reppetto, American Police: A History 1845-1945 (New York: Enigma Books, 2010).

17, “Naxalite,” last updated December 30, 2015,

18 Kapur, Gawande, and Satyanath, “Renewable Resource Shocks,” 35.

19 Manisha Priyam, Sanjeev Chopra, and Om Prakash Chaudhary, “Against the Odds: Educational Initiatives in Dantewada,” Department of Administrative Reforms and Public Grievances, 3, accessed May 20, 2016,

20 Pavan Dahat, “Salwa Judum-2 Is Born in Bastar,” Hindu, May 5, 2015,

21 Brendan O’Leary and Andrew Silke, “Understanding and Ending Persistent Conflicts: Bridging Research and Policy,” in Terror, Insurgency, and the State,401, 408.

22 The literature on what are known as ultimatum games suggests that offers of less than 30 percent of a given amount are typically rejected, a finding that has held across vastly different cultures studied, with some exceptions. The findings appear to have deep biological roots, as even chimpanzees and dogs have been found to exhibit similar behavior, giving up a potential reward to prevent another from receiving what is perceived to be an unfairly large benefit. See, for instance, David G. Rand, et al., “Evolution of Fairness in the One-Shot Anonymous Ultimatum Game,” Proceedings of the National Academy of Sciences of the United States of America 110, no. 7 (February 12, 2013).

23 Priyam, Chopra, and Chaudhary, “Against the Odds,” 11; Nitin Mahajan, “Chattisgarh Plans Residential Schools in Naxal Hotbed,” Indian Express, January 2, 2008,

24 See Monica Duffy Toft, Securing the Peace: The Durable Settlement of Civil Wars (Princeton, NJ: Princeton University Press, 2009) on the failures of foreign-backed governments to beat rebel movements. Foreign efforts have been successful at creating elite enclave units that can address small, specific problems—such as U.S. efforts to train Georgian battalions to fight terrorists in the Pankasi Gorge. But when an insurgency is more general, enclave units are rarely enough to succeed.

25 Manish K. Jha and Pushpendra, “Governing Caste and Managing Conflicts: Bihar, 1990-2011,” Policies and Practices 48 (March 2012), 12.

26 Sanjay K. Jha, “Naxalite Movement in Bihar and Jharkhand,” Dialogue 6, no. 4 (April–June 2005).

27 Smita Narula, Broken People. India’s minimum wage is set by type of job and is not easily summed up in a single number. However, the minimum wages for the roles generally filled by Dalits in agricultural work in Bihar are five to six times this amount.

28 Avinash Kumar, “Illegitimacy of the State in Bihar,” Economic and Political Weekly, October 31, 2009, 8.

29 The situation of a reawakened insurgency in some ways paralleled that of Colombia in the mid-1960s, when Plan Lazo, the counterterrorism mission assisted by the United States, fractured a movement of communist guerrillas and reduced their number to just a few hundred. But again, absent a more just political settlement, the guerrillas regrouped after president César Gaviria Trujillo’s authoritarian rule and eventually grew to more than 18,000 by the turn of the millennium. The Irish Republican Army faced a parallel near-destruction militarily by the mid-1940s, after losing the Irish Civil War of 1922–1923 and military fights with the UK in 1939 and with the Northern Irish government in 1945. And similarly, with grievances unaddressed, it was able to revive in 1969 and go on to its period of greatest strength.

30 Narula, Broken People.

31 “Can Growth Tame Maoists?,” Economic Times.

32 Narula, Broken People.

33 Narula, “The Pattern of Abuse: Rural Violence in Bihar and the State’s Response,” in Broken People.

34 South Asia Terrorism Portal, “Chronology of Massacres in Central Bihar (1977-2001),” accessed March 8, 2016,; see also Santosh Singh, “A Lasting Signature on Bihar’s Most Violent Years,” Times of India, June 3, 2012,

35 Narula, Broken People.

36 Giridhar Jha, “Brahmeshwar Singh: The Butcher Behind the School Teacher-Like Looks,” India Today,June 1, 2012,

37 Santhosh Mathew and Mick Moore, “State Incapacity by Design: Understanding the Bihar Story,” Institute of Development Studies Working Paper 2011, no. 366, May 2011.

38 Narula, Broken People.

39 Kumar, “Illegitimacy of the State in Bihar,” 8–11. Also see Ashwani Kumar, Community Warriors: State, Peasants and Caste Armies in Bihar (London: Anthem, 2008).

40 The 2005 Mahatma Gandhi National Rural Employment Guarantee Act, implemented in February 2006, guarantees livelihood security in rural areas by providing employment to unskilled laborers for at least one hundred days in a year.

41 Swaminathan S. Anklesaria Aiyar, “Bihar: How the Poorest, Worst-Governed State Attained Double-Digit Growth and Became a Role Model,” in Economic Freedom of the States of India: 2013, eds. Bibek Debroy, Laveesh Bhandari, and Swaminathan S. Anklesaria Aiyar (New Delhi: Academic Foundation, 2014), 54–55.

42 The number of Maoist incidents in Bihar did decline before Kumar’s election. However, each incident was more violent and resulted in more deaths. That trend tends to be characteristic of insurgencies that are strengthening and able to carry out more large-scale violent acts.

43 Jeffrey Witsoe, Democracy Against Development (Chicago, University of Chicago Press, 2010), 9.

44 Press Trust of India, “86 Police Men Killed by Naxals in Bihar Since 2005: Statistics,” Economic Times, September 5, 2010,

45 Press Trust of India, “Five Naxals Get Death Sentence for Killing Bihar Cops,” Times of India, December 6, 2007,

46 Surendra Kishore, interview with Rushda Majeed on July 1, 2015, in Patna, Bihar.

47 Intelligence gathering also played a key role in the success of Andhra Pradesh, which the Indian government has used as a model of development. The state successfully used a ceasefire period to infiltrate Naxalite cells and collect intelligence, a down payment that yielded significant dividends when the government began using more force against the Maoists after the ceasefire.

48 Press Trust of India, “Maoist Area Commander and Six Others Arrested in Bihar,” India Today, February 18, 2015,

49 Abhayanand, interview with Rushda Majeed on June 29, 2015, in Patna, Bihar.

50 Chanchal Kumar, interview with Rushda Majeed on July 3, 2015, in Patna, Bihar.

51 Arun Kumar, interview with Rushda Majeed on July 4, 2015, in Patna, Bihar.

52 Karthik Muralidharan and Nishith Prakash, “Cycling to School: Increasing Secondary School Enrollment for Girls in India” (working paper, University of California, San Diego, January 2, 2016),

53 Aiyar, “Bihar: How the Poorest,” 55.

54 See Jane Jacobs, The Death and Life of Great American Cities (New York, Random House, 1961).

55 For a terrific unpacking of the black box of political will, see Diane de Gramont, “Beyond Magic Bullets in Governance Reform,” Carnegie Endowment for International Peace, November 2014.

56 Neyaz Farooquee, “A Dalit Village Waits Uneasily for Justice,” India Ink (blog), New York Times, December 2, 2013,

57 Rakesh Choudhary, “Land Reforms Are Critical for Bihar, So Why the Silence During Polls?,” Huffington Post, November 23, 2015,

58 Dipankar Bhattacharya, “Ranvir Sena Revisted: Feudal-Kulak Power and Lalu-Nitish Continuum,” Economic and Political Weekly, July 28, 2012.

59 For a careful analysis of the criminal links across Bihari political parties, see the Association for Democratic Reforms Bihar page, which contains links to the white papers “Analysis of Criminal and Financial Background of MLAs and Candidates” 2005, 2010, and 2015, jointly published by Bihar Election Watch and the Association for Democratic Reforms. Analysis of the political reasoning behind these alliances stems from research discussions. “Bihar Assembly Elections 2015,” Association for Democratic Reforms, November 20, 2015,

60 Supriya Sharma, “Caste Politics: Why the Bihar Election Is Tragic, Regardless of the Outcome,”, September 29, 2015,

61 Ashwani Kumar, interviews with Rushda Majeed by Skype on July 6, 2015, and by e-mail on May 1, 2016.

62 Political reservations of government seats in India have been shown to have significant positive effects for lower castes, who face reduced humiliation and violence and greater political access as a result, even if the reservations may have little affect on traditional stereotypes. See Simon Chauchard, “Can Descriptive Representation Change Beliefs About a Stigmatized Group? Evidence From Rural India,” American Political Science Review 108, no. 2 (May 2014).

63 Aiyar, “Bihar: How the Poorest,” 55.

64 Ibid.

65 Ibid. See also ASER Center, Annual Status of Education Report 2007 (New Delhi: ASER Center, 2008), and ASER Center, Annual Status of Education Report 2012 (New Delhi: ASER Center, 2013).

66 Arnab Mukherji and Anjan Mukherji, “Bihar: What Went Wrong? And What Changed?,” National Institute of Public Finance and Policy, Working Paper no. 2012-107, September 2012, 37–38. However, under Lalu Yadav, staff absentee rates of 58 percent in primary health centers topped the nation, and more than one-quarter of primary school teachers were always absent, the third-worst rate in the country. Rajesh Chakrabarti, Bihar Breakthrough: The Turnaround of a Beleaguered State (Kolkata: Rupa Publications, 2013), 14.

67 Press Trust of India, “Quality Education Still a Challenge, Says Nitish Kumar,” News18, July 16, 2015,

68 Shaibal Gupta, interview with Rushda Majeed on July 7, 2015, in Patna, Bihar.

69 Arun Kumar, interview with Rushda Majeed on July 3, 2015, in Patna, Bihar.

70 Sayantanee Choudhury, “Bihar Government Releases Rs. 1.06 Crore for Rehabilitation of Maoists,” Times of India, December 11, 2015,

71 Deeptiman Tiwary, “Only 106 Cops per 1 Lakh Indians,” Times of India,February 22, 2014,; and National Crime Records Bureau, “Police Strength, Expenditure and Infrastructure,” chap. 17 in Crime in India 2005 (New Delhi: Ministry of Home Affairs, 2006),

72 Press Trust of India, “Nitish Claims Credit for Bihar’s Growth Story During His Rule,” News18, July 27, 2015,

73 P. K. Thakur, interview with Rushda Majeed on July 3, 2015, in Patna, Bihar.

74 Arun Kumar, interview with Rushda Majeed on July 3, 2015, in Patna, Bihar.

75 Anup Mukerji, interview with Rushda Majeed on July 11, 2015, in Patna, Bihar.

76 Narula, Broken People.

77 While statistics on police brutality are not available, Julu Katticaran researched the number of complaints against encounter killings and other human rights abuses, and custodial deaths. Her findings are in the appendix.

78 “Crime Against Women in Bihar Increased by 11 Per Cent in 2013,” India Today, January 13, 2014,

79 Ministry of Home Affairs,“Left Wing Extremism (LWE) Division,”Government of India, accessed August 14, 2015,

80 Compiled from the website of the Ministry of Home Affairs, Government of India.

81 Ministry of Home Affairs, “Left Wing Extremism (LWE) Division.”

Carnegie does not take institutional positions on public policy issues; the views represented herein are those of the author(s) and do not necessarily reflect the views of Carnegie, its staff, or its trustees.