MONTHS AGO an important Department of Justice (DOJ) White Paper was leaked to the press. This document outlined, in concise and prescriptive wording, when the DOJ believes that the government of the United States has the legal authority to kill an American citizen.

Yet as much as this appeared an overstep in state power, the White Paper was nothing but business as usual—at least in theoretical terms. Across political and social thinking, the idea that the state has the right to kill its own citizens is rarely contested. From Hobbes to Weber and Mbembe, it is explicit or implied that states decide the conditions under which citizens can, and indeed should, die in order to preserve sovereignty. These conditions range in both scale and act, from when governments make declarations of war, to the process and finality of capital punishment, and the open secret that certain populations within states are left to die, or left to kill each other, because states deem them less important. In practical terms such state violence ranges from concealed and diffuse, as in the American Ghetto¹, to shocking and acute, as in Guantanamo².

If the DOJ proclamation raised concerns about too much central power, it can also help us think about states with the opposite problem: a dearth of central power. Much closer to home, many states have little ability to make these kinds of proclamations or to back them up, a fact that raises big theoretical questions about sovereignty in the contemporary world. Many states that we would never otherwise declare failed fit within a new category of states where central authority is plastic or "negotiated"³.

At the heart of this empirical problem are major cities. In recent decades, many major cities in the developing world, from Rio de Janeiro to Johannesburg have been torn apart by violence. Market-oriented and non-revolutionary organized armed groups—violent entrepreneurs—have exploded across borders, following illicit supply chains and feeding off easy access to guns in order to fortify urban territories. Cities as disparate as Caracas, Nairobi and Jakarta are wrapped up in the throes of vertiginous urban violence.

A DECENTRALIZED RIGHT TO KILL

The nature of sovereignty and these conditions of violence can be made legible by looking at who, other than the state, has the ability to regulate violence. Recent experiences in Mexico, El Salvador and Brazil, have shown decisively that some states have little or no capacity to decide who should live and who can die within their borders. Instead, as my own dissertation fieldwork in São Paulo, Brazil, examines, non-state armed groups often regulate the conditions of life and death. These are groups like El Salvador's Maras, São Paulo's Primeiro Comando da Capital, Mexico's Zetas, and the Sixth Division under Don Berna in Medellin. As these groups consolidated control or established truces with other groups or the police, we observe radical and cyclical oscillations in the respective homicide rates of cities.

This problem is not far from the American doorstep. In March of 2012, the two dominant and transnational Mara groups in El Salvador—the Salvatrucha and Barrio 18—agreed to end decades of bloodshed that had engulfed not only both groups, but society in general. The leadership of the two organizations, both of which are located within the prison system, sought concessions from politicians for better prison conditions as a bargaining chip. One year later, the result has been astounding. The homicide rate has declined by around 60%—no small feat for a place that had a homicide rate of 14 per day in 2009⁵.

It is now public knowledge that these two Mara groups are behind the relative peace. They have spoken openly about it and, in contrast to other countries, public security figureheads have struggled to take credit for the decline. Most understand that the sustainability of this relative peace depends very little, if at all, on public policy, good policing or state willingness. That El Salvador is now somehow safer'—even as other types of violence, including kidnapping, extortion and other economic crime continue to increase because they are the economic life line of the Maras—has come to passin spite of the public security system, not because of it. In some sense, peace depends on non-state armed groups agreeing to be civil.

REGULATING LIFE AND DEATH: THE THREE 'LOGICS' OF SÃO PAULO

In São Paulo, Brazil, the dynamics are similar, if more obscured. The existence of the Primeiro Comando da Capital (PCC)—a powerful prison-based non-state armed group—has redefined how, and how many, people die in this mega-city of nearly 20 million residents. This is much to the chagrin of a public security system and its leaders who, until December of last year, were loathe to admit that the organization existed at all. But the street-level influences of the PCC are inescapable, particularly for those who work within the public security system—the police themselves.

Based on ethnographic fieldwork carried out in São Paulo with police homicide and other specialized detectives since 2009, my research examines the decentralization of organized violence that exists in the absence of a state monopoly on the right to kill in the city. I identify and trace three distinct and antagonistic logic' of death that exist in São Paulo, as channeled through three groups: The PCC, police who kill civilians, and the police that investigate both of them. Each of these three groups, I argue, regulate death in contradictory terms.

POLICE WHO KILL CIVILIANS

In this city, as in other Brazilian cities, police kill civilians at a rate of at least one per day. They are at once juries, judges and executioners. As jury and judge, often in very violent and uncertain circumstances, beat cops make snap decisions about when to kill. As executioners, they carry out that sentence with their own hands, as an act of unbridled state power. These killings are systemic, being positioned within the public security apparatus as an important, if not necessary, facet of everyday police work. Until just recently, the law gave these police the decisive upper hand. Individuals killed by police were formally categorized post-mortem as having committed a crime against both society and the state. The act of resisting arrest legitimized deadly violence on the part of police, who could position their actions as necessitated by resistance to the state, very broadly and subjectively defined. In death, the 'criminal' was eternally implicated, justifying the actions of the police who themselves became the victim.

Many police see these killings as central to the practice of policing in a city where the monopoly on violence has fragmented. Immediately following such a killing, the structural logic of this violence echoes in statements like "menos um" (one less), "bandido bom é bandido morto" (a good thief is a dead thief) and, to the police who killed, "parabéns!" (congratulations!) with pats on the back. Chronically burdened by a never-ending tide of crime, many police see all 'criminals' as an irrevocable scourge to be permanently eliminated, at any cost.

One of the reasons that police feel so strongly, however, is because the influence of crime—with the PCC as its masthead—threatens their own lives. Police come from and often continue to live in the same kinds of subaltern urban spaces that the PCC now controls. Many have grown up alongside today's criminal element, continuing to live on the same streets and sharing the same grocery stores, bars and corner markets. Worse still, in these communities the law of the PCC supersedes the law of the state, meaning that police are forced to set aside their identities and roles as police officers if they hope to survive. For police themselves, there is no doubting why community members dramatically underreport crime to the police stations in these areas—there is a much more localized and improbably 'functional' form of order.

THE PCC

As others and I have detailed elsewhere, the PCC is a formidable force of order in the city of São Paulo⁶. Emergent from the violence of the São Paulo prison system, the organization formed as a means of self-protection for members. In its genesis, and before it was the drug-oriented urban-scale protection racket of today, the sole objective of the PCC was to eliminate what it described as the violence and injustice of the medieval conditions within the prison system. In doing this, it succeeded. The PCC today controls 135 of the 152 prisons in São Paulo state with its own form of governance and morality. Sexual violence and the use of drugs that 'destabilize', such as crack, are prohibited. A self-protection rationale subsists deeply within an organization that values security at all costs. Even as it made the leap from the prison system to the poorer parts of São Paulo ten years ago, and, today, as it continues to spread to other Brazilian states and even into Paraguay and Bolivia, this rationale remains a powerful backdrop.

Part of ensuring security for the organization's 'own' means punishing people for doing things that threaten the organization and its affiliates. In my fieldwork I have collected a variety of data from sources such as seized PCC notebooks and pen drives, homicide cases files and the process of investigation of those cases, police conversations with PCC members, as well as my own conversations with residents of PCC controlled communities. This data points to a particular form of authority and punishment, carried out via local and ad-hoc 'tribunals'. Punishment can be differentiated into two categories: Punishment for 'baptized' members, and punishment for residents of the communities controlled by the organization. Members can be killed for any number of reasons, from failure to pay monthly dues, to stealing money or drugs, or for lying about important matters. The punishments are tabulated in notebooks, and include a host of personal details and include justifications for the reasons why. In the community, however, residents are also punished for committing crimes without the explicit approval of the organization. Community members seen to have carried out a crime without the blessing—implicit or explicit—of the PCC are punished by death. Murder occurs when the organization says so. An illegitimate killing is punishable in the same terms.

HOMICIDE DETECTIVES

Make no mistake, São Paulo's public security system has its peculiarities. One such peculiarity is that homicide detectives are responsible for investigating two types of deaths: Intentional homicides and police killings of civilians. In terms of homicides, detectives carry out investigations of thousands of incidents every year, ranging from crimes of passion to bloody multiple homicides. Most of these killings, though, are of young men from poorer parts of the city. A large majority of these are found in public spaces with gunshot wounds. Police detectives have little doubt that these kinds of homicides are deeply intertwined with the influences of the PCC. Even cases that appear unrelated are often traced to things like used car sales, which has become a key money laundering racket for the organization in recent years. A 2012 report from the public prosecutor pointed out that the resolution rate—cases closed but not necessarily proceeding to trial—was 29.5%. Put differently, much less than one in three cases ever make it from body on the street to defendant in the courtroom.

When it comes to police killings of civilians, these detectives are responsible for investigating and arresting police that have killed illegitimately. This illegitimacy is of course subjective, but the routine nature of these killings makes this work predictable. Squeezed between the two other logics of death, these detectives search for when the story doesn't add up. When they arrest police, which happens many times each year, they often do so because witnesses' versions are out of sync with the police story, the person killed isn't viewed to be a criminal or they come across video evidence. While these police don't themselves pull the trigger, they have their own normative outlook on when killing is appropriate. In an environment lacking technology and accountability, this outlook drives their deductive reasoning. Detectives have much discretion, allowing them to lean heavily on their own moral borderlines to make sense of when death is legitimate or not. Hunches, backed up by evidence, come to define when someone must be held to account for killing another. These borderlines are strong and informed by both their own identities as police and their desire to centralize violence and mitigate their own insecurity. As one detective told me just after arresting three police for executing a man on the side of the highway, "Corrupt police make it harder for the rest of us."

BEYOND THE MONOPOLY

David Simon, a journalist and co-creator of the HBO series "The Wire," is one of the only researchers to have studied homicide detectives. He describes a much different environment in Baltimore, a city with its own major struggles with violence. Yet even in this American city, troubled as it is, there is no rivalry between the state and other notions of legitimate death. In the end, he writes "…only a cop has the right to kill as an act of personal deliberation and action⁷."

Here we can make a clear distinction. When it comes to the state's right to kill, there are states that can decide, more or less unilaterally, when to kill their own citizens. Rightly or wrongly, they carry out this death through a number of institutions, each with their own checks and balances. The DOJ White Paper is a reflection of such a reality. But, on the other hand, there are many states in the world today where the right to kill is fractured. In these places the state rarely has the final (or even primary) say in who can and cannot die. Who must die and who can live is a product of different forms of governance and security in a polity. It is in these kinds of states—places that are not subsumed by conventional notions of how states should work but are far from failed—that I seek to advance a new way of thinking about security, democracy and the very integrity of the state itself.

REFERENCES

1 Wacquant, Loic. (2001). Deadly Symbiosis: When Ghetto and Prison Meet and Mesh. Punishment and Society, 3(1), 95-134.

2 Moqbel, Samir Naji al Hasan. (2013 – April 14). Gitmo is Killing Me. Opinion Editorial, New York Times. Retrieved from: http://www.nytimes.com/2013/04/15/opinion/hunger-striking-at-guantanamo-bay.html

3 Müller, Markus-Michael. (2012). Public Security in the Negotiated State: Policing in Latin American and Beyond. Basingstoke: Palgrave Macmillan. 

4 Blok, Anton. (1974). The Mafia of a Sicilian Village, 1860-1960: A Study of Violent Peasant Entrepreneurs. Cambridge: Waveland Press. 

5 Carlos, Juan and Jessica Bennett. (2012). In El Salvador, A Gang Truce Can't Stop the Violence. Mother Jones online. Retrieved from: http://www.motherjones.com/photoessays/2013/03/el-salvador-gang-truce-mara-salvatrucha-barrio-service.

6 Feltran Gabriel. (2011) Fronteiras de tensão: Política e violência nas periferias de São Paulo. São Paulo: UNESP. 

7 Simon, David. (1991): Homicide: A Year on the Killing Streets. New York: Holt.