Читать книгу Police Power and Race Riots - Cathy Lisa Schneider - Страница 8

Introduction

A riot is somebody talking. A riot is a man crying out “Listen to me mister. There’s something I’ve been trying to tell you and you are not listening.”

—Federal Communications Commissioner Nicholas Johnson (speaking after the 1968 riots in Washington, D.C.)¹

On the night of July 18, 1964, three weeks after Lyndon Johnson signed the Civil Rights Act, New York City police lieutenant Thomas Gilligan shot and killed James Powell, a fifteen-year-old black student, outside his high school in upper Manhattan. An altercation had ensued when Patrick Lynch, the white janitor of a nearby building, sprayed black high school students with a garden hose as they left the school. Lynch shouted racial epithets at the boys, and Powell and two other high school students chased him back to his building. Officer Gilligan arrived on the scene, pivoted, and shot Powell three times. The high school students screamed and cursed at the police, some throwing bottles and cans. “Come on,” they taunted Gilligan, “shoot another nigger.”² Dozens more police arrived on the scene.

The following day hundreds of Harlem residents gathered in front of police stations but were met with walls of tactical police. Scuffles broke out as police wielded batons and shot into the crowds. Meanwhile residents threw bricks and bottles, pulled fire alarms, overturned cars, looted stores, and occasionally attacked white bystanders. After three days the disorder spread from Harlem to Bedford-Stuyvesant in Brooklyn. More than six thousand officers were called in to quell the disturbances. By the sixth day of what would become the first major urban uprising of the 1960s, more than five hundred people had been injured, including thirty-five policemen, and another black man was dead, of gunshot wounds. Property damages were assessed at several million dollars.

In the weeks that followed, similar outbreaks occurred in Rochester, Philadelphia, and several smaller northeastern cities. The following year violence in the Watts ghetto in Los Angeles exploded, leaving thirty-four dead. Riots in Chicago, Cleveland, Dayton, San Francisco, Atlanta, and Omaha erupted in 1966. In 1967 Puerto Rican riots broke out in New York, concentrating in East Harlem and South Bronx. Black neighborhoods in Boston, Nashville, Cincinnati, Newark, and Milwaukee followed suit. Detroit experienced the largest and most deadly disturbance; there police violence left forty-three dead: “By one count the 1964–1968 period produced 329 important riots in 257 cities, with 52,629 persons being arrested for riot-related offenses, 8,371 injured, and 220 killed—mostly black civilians.”³ Only twice in the next five decades (excluding local and relatively isolated, if significant, neighborhood skirmishes and confrontations between residents and police) would minority residents set a major American city aflame.

On October 27, 2005, in Clichy-sous-Bois, outside Paris, police chased one black Mauritanian and two North African (Tunisian and Kurdish) teenage boys, Bouna Traoré, Zyad Benna, and Muhittan Altun, into an electrical substation outside Paris. The boys had tried to avoid an identity check and were pursued by officers carrying stun guns. Cornered, the boys scaled an eight-foot wall covered with barbed wire and skull and crossbones warning of the dangers of electricity. Seeing the boys inside the generator, the commanding officer notified headquarters that reinforcements would not be needed, the boys would not live long now.⁴ The police officers then abandoned the site. Inside the grid, the terrified boys clung to each other for eleven minutes, weaving back and forth and looking for a way out. When Bouna accidentally hit the generator, he and Zyad, who was holding his hand, died instantly. Muhittan, although holding Zyad’s hand, was saved by the power surge. Severely burned, he retraced his steps, rescaled the wall, and ran, crying hysterically, into the arms of Bouna’s stunned and unprepared older brother, Siyakha, who was on his way to buy food to end the family’s Ramadan fast.

In the days that followed, young men from the town marched through the streets cursing and throwing garbage at the police. A massive nonviolent march was followed by more police actions including the shooting of tear gas into a mosque filled with women and children. Following the tear gas incident, then minister of the interior Nicolas Sarkozy, having unjustly accused the three teenagers of criminality two days before, gave an inflammatory radio address denying the responsibility of the police or the need for investigation. Hurt and angry, young people retaliated, setting cars and buildings aflame. Fires spread from one poor minority suburb (banlieue) to another, until 280 suburbs, cities, and towns across France were ablaze. For three weeks youths of predominantly North African and African descent set fire to an estimated 10,000 cars and attacked 255 schools, 233 other public buildings, and scores of private businesses. Despite the arrest of nearly 5,000 people, the police subdued the burning banlieues only after the government imposed a curfew—a means last used against Algerian Frenchmen during the Algerian war.⁵

These two series of events, taking place on opposite sides of the Atlantic, in strikingly different settings and half a century apart, nonetheless display some startling similarities. In both, police violence resulted in the death of one or more minority teenagers. In both, the ensuing uprisings began where the boys had been killed but soon spread to neighboring sites and then distant towns and cities. Following both events, the respective states used the riots to justify harsh new criminal policies targeted at poor minority youths—launching a war on drugs in the United States and harsh new criminal policies in France.

These parallels raise three critical questions. First, why did police forces in such dramatically different settings interact with distinct minorities in such similar ways? Second, why did these interactions lead to riots in New York and hundreds of cities across the United States during the 1960s and in Paris and across France in 2005? Third, why have riots been so rare in New York and most of the United States since the 1970s, even where police brutality and racial profiling have grown? This puzzle is all the more striking since the socioeconomic divide in American cities, like New York, ranks among the world’s widest, whereas the divide in French cities, like Paris, ranks among the narrower. As Michael Katz observes, “Today the most intriguing questions are not why the riots [of the 1960s] occurred but why they have not reoccurred…. With the exception of Liberty City Miami in 1980 and South-central Los Angeles in early 1992, American cities have not burned since the early 1970s. Even the botched response to Hurricane Katrina did not provoke civil violence. The question becomes all the more intriguing, in the light of October 2005, when riots erupted in at least three hundred towns across France.”⁶

How can we explain the conundrum that Katz identifies? The short answer to this puzzle is that many of the factors that make life unbearable in American slums have little to do with riots. Misery is ubiquitous; riots are exceedingly rare. This is not surprising as a) poverty and unemployment dishearten rather than mobilize and b) people often blame themselves for economic failure. Police violence, in contrast, and the killing of unarmed minority youths by white police in particular spark moral outrage, activate racial boundaries, crystallize grievances along a single us/them boundary, and provoke riots.

No feature of a racially divided society is a more potent symbol of racial domination or instills the message of subjugation more forcefully than police. The frequent identity checks, the stop-and-frisks, the disrespect and brutal manner with which police address minority youths, and, worst of all, the utter impunity that allows the most racist and sadistic officers to commit gross violations of human rights and homicide: all these constantly and painfully remind youths of their subordinate status. As the French rapper Monsieur R notes, police “just by the way they look at you they give you the feeling that you are a second-class citizen, even if you were born here. Children are stopped for inspection five times, just on the way from their home to the metro! And I’m talking of a walking distance of less than 10 minutes…. Today in France the police logic is simple…. Here, if you’re black or Arab, it doesn’t matter if you have money or a good job, you’ll remain black or Arab your whole life.”⁷

In unequal, racially divided societies political elites rely on police to enforce categorical boundaries. When political rhetoric, distorted media attention, and public policy activate⁸ those boundaries, police violence against subordinate groups intensifies. Police violence further polarizes social relations around an us/them divide. Under such conditions an egregious incident of police violence may trigger riots and urban unrest. If social movements, courts, or other institutions offer alternative paths to justice, no matter how limited, riots are rare. Riots are the last resort for those who find all other paths to justice blocked.

Policing Racial Boundaries

All states draw boundaries that, as Eric Wolf puts it, “define rights to membership, construct justificatory ontogenies for their cadres and lay down criteria for denying participation and benefits to groups deemed unwelcome, unworthy or deleterious.”⁹ State building, by definition, entails the creation of unequal, bounded categories—Frenchman/German, citizen/noncitizen, and national/non-national. Extreme categorical inequality results when powerful groups conquer less powerful groups and force them into submission. During and following such historical processes members of powerful groups tell stories about members of less powerful groups to justify their own privileged position. As Adolph Reed notes, “Ascriptive ideologies are just-so stories with the potential to become self-fulfilling prophecies. They emerge from self-interested common sense as folk knowledge: they are ‘known’ to be true unreflectively because they seem to comport with the evidence of quotidian experience. They are likely to become generally assumed as self-evident truth, and imposed as such by law and custom, when they converge with and reinforce the interests of powerful strata in the society.”¹⁰

Categories simplify and facilitate exploitation (the expropriation of profits, labor power, and resources) and opportunity hoarding (the exclusion of others from access to valuable resources and opportunities). They “lock such differences into place by delivering greater rewards to occupants of the ostensibly superior category,” notes Charles Tilly.¹¹ Because the categories solve pressing organizational problems and are costly to change, they often outlast their original purpose. Categories constructed for the purposes of slavery, conquest, or colonialism can later be used to reinforce unequal systems of remuneration, for example, “assisting employers in assigning workers to jobs for which they were racially suited.”¹² One Pittsburgh company in the 1920s, Reed notes, classified thirty-six different racial groups by their supposed capacity for twenty-two distinct jobs, each demanding different atmospheric conditions, levels of speed and precision, and day- or night-shift work. Similarly, Anibal Quijano and Ramon Grosfoguel and Chloe S. Georas describe “colonialities of power,” in which social practices are “implicated in relationships among people even when the colonial relationships have been eradicated.”¹³

Adaptation reinforces categorical inequality when social networks form on either side of a categorical boundary. Those on one side of the boundary “claim solidarity with others on the same side … and invoke a certain sort of relationship to those on the opposite side,” notes Tilly.¹⁴ Over time both sides “attribute hard and durable and even genetic reality to the categories they inscribe. Wherever they come from the categories have serious social consequences.”¹⁵ That is not to say that racial categorizations are static and ahistorical, or that racism, as David Goldberg puts it, “is singular and monolithic, simply the same attitude complex manifested in varying circumstances.”¹⁶ Rather preexisting, yet malleable categories are used by a wide array of actors to navigate or control a complicated and evolving environment.

Several problems remain for state authorities. First, polarization can have “the unintended consequence of defining, legitimating and provoking group identity and mobilization, forging struggles for inclusion between state agents and emerging political actors,” as Anthony Marx notes.¹⁷ Second, racial, ethnic, or religious segregation may unintentionally provide a space for resistance, as Niall O Dochartaigh and Lorenzo Bosi observe in their study of the civil rights movement in Northern Ireland: “[B]oundaries drawn and maintained by states can act as a source of power for oppositional forces by creating and consolidating homogenous bounded spaces that can act as resources for resistance.”¹⁸ It may also, as Sebastien Roché observes, contribute to the diffusion of riots: “When the minority at the national level becomes a majority in a neighborhood, the balance of power tends to be inverted.”¹⁹ Faced with such challenges, “privileged groups may indirectly encourage police killings,” David Jacobs and Robert M. O’Brian note, “by demanding that restraints be removed so the police can maintain order.”²⁰

The control of racial and spatial boundaries is an essential component of police work. “The desire to render society transparent, to know everything about everyone extends back at least as far as the European Enlightenment,” notes Clifford Rosenberg. “Faced with the impossibility of realizing these ambitions,” he continues, “police forces have had to give priority to some categories of people or places to watch over…. police forces have always worried about foreigners, aliens, outsiders and marginal types.”²¹ Governments and dominant groups expect police to make these distinctions and pass laws to facilitate this process. In 1925, for instance, the French minister of the interior told the prefect of police, “[W]e want to weed out the bad ones, but we cannot give them the impression that we are treating them like foreigners, undesirable foreigners at that, and requiring passports for that reason…. We can demand passports under other pretexts, but not to protect ourselves from them; we must appear to be protecting them.”²²

Police forces do not invent these categories. Rather they use categories that mirror those of the society in which they are embedded. They commonly match preexisting exterior categories such as race or ethnicity to interior categories such as citizenship rights, obligations, and forbidden behaviors. These classifications serve as shorthand, allowing officers to summarize “complex and ambiguous situations in a short period of time.”²³ Some neighborhoods are designated as in need of protection and others as foreign or dangerous. “Both the racial characteristics of the suspect and the suspect’s neighborhood influence police decisions to stop, search or arrest a suspect,” Jeffrey Fagan, Amanda Geller, Garth Davies, and Valerie West observe: race and neighborhood interactively “animate the formation of suspicion among police officers.”²⁴

In racially divided societies, police, like occupying armies, mark stigmatized minority neighborhoods as enemy territory. “Enforcement through bounding,” note O Dochartaigh and Bosi, “has the key advantage of simplicity…. Boundaries provide a simple and powerful method of communication, pouring huge and often complex information into a single symbol that distinguishes between the vast and complex variety of the internal and similarly complex external. To bound space is to communicate huge volumes of information in a single symbol.”²⁵

The categories such boundaries inscribe are largely fungible. Tilly notes, “[S]eemingly contradictory categorical principles such as age, race, gender and ethnicity operate in similar ways and can be organizationally combined or substituted within limits set by previously established scripting and local knowledge.”²⁶ In New York and Paris, police paint blacks, Puerto Ricans, Arabs, and Berbers as racialized others. One French officer who railed against “little jerkoffs,” for instance, told Didier Fassin, “The blacks are just like the Arabs, except they’ve no brain.”²⁷ Similarly two high-ranking members of a French left-wing police union told me a story that linked both blacks and Arabs to a host of ominous threats:

The problem here began after the war in Algeria. The Arabs could not stay there because they had collaborated with France. They ended up congregating in small areas. Housing prices went down and black immigrants now found it very cheap and gathered there too. This concentration created an underground world…. Yet for us as police it is good, it is easier to bust. If they were spread around the city it would be difficult to police. If you let them live together, you do not even have to go into the cité. You can put police at either end and close it. It is a way to localize and crystallize delinquency in a single place. But, if you live there and see all blacks and Arabs on the pavement you can imagine people say “what are the police doing?”²⁸

“Only in extremely unequal societies where particular groups are denied full membership,” notes Pieter Spierenburg, “do police act in a disrespectful and brutal manner with unarmed citizenry.”²⁹ Paul Chevigny concurs: “The term punishment is never used unless the person on whom the penalty is inflicted is clearly subordinate to the one imposing the penal act.”³⁰ Fassin observes,

Police violence, whether physical or moral, is exercised in a radically and institutionally unequal manner. On one side are individuals who have not only the monopoly of the legitimate use of force, but also exclusive access to effective use of it given the circumstances. On the other are individuals who are doubly captive, owing to both the physical coercion they undergo and the latent threat weighing on them if ever they should have the bad idea of talking back. Whether detained, handcuffed or simply surrounded by officers, the person exposed to their power is rendered structurally inferior: he is bound to submit, and any protest or rebellion can only lead to even greater submission. Violence is therefore almost always strictly unilateral. But it is also targeted. It is not applied to all.³¹

Comparative studies of policing tend to emphasize the impact of different state structures and institutional cultures on policing strategies. “Policing styles are influenced by the political system,” argue Donatella Della Porta and Herbert Reiter. “Institutional features such as police organization, the nature of the judiciary, law codes and constitutional rights—may play a role in defining the opportunities and constraints.”³² But if institutional structure and culture were key, then police in France—a country with a strong central state apparatus, a national police force, common law tradition, and elimination of racial categories in census and law—should interact with black and Arab residents very differently than American police—locally controlled, in a federal state with statutory law and multicultural traditions—interact with blacks and Latinos.

Yet we see strikingly similar interactions between police and minority populations in Paris and those in New York. The structure of inequality at large and the demands placed on police by powerful social classes, dominant racial groups, and the state are more important than the culture and organization of the institution. “The problem is not to know whether the police act identically everywhere within a national territory or across borders,” as Fassin points out, “but whether the type of relation they have with a certain public, the way in which political incentives influence their practice, the effects of various systems of evaluation and sanctioning on their conducts, or the justification they provide for their deviant behaviors are generalizable.”³³

Debating Riots

In the first major study of seventy-five riots occurring in the United States between 1964 and 1967, the U.S. Commission on Civil Disorders (aka the Kerner Commission) in 1968 reached the stark conclusion that “our nation is moving towards two societies, one black, one white—separate and unequal.”³⁴ While members of the commission observed that “almost invariably the incident that ignites disorder arises from police action,” they were careful to emphasize that “the disorder did not erupt as a result of a single triggering or precipitating incident. Instead it was generated out of an increasingly disturbed social atmosphere, in which typically a series of tension-heightening incidents over a period of weeks or months became linked in the minds of many in the Negro community with a reservoir of underlying grievances.”³⁵ The solution would require nothing less than “the realization of common opportunities for all…. From every American it will require new attitudes, new understandings, and above all a new will.”³⁶

The findings were simultaneously too vast and too limited. The list was too vast. While blacks and Latinos suffered immense hardships and injustice, they did not burn buildings or loot stores to protest every wrong. The commission listed the following twelve most cited grievances in descending order by intensity: 1) police practices; 2) unemployment and underemployment; 3) inadequate housing; 4) inadequate education; 5) poor recreational facilities and programs; 6) ineffectiveness of the political structure and grievance mechanisms; 7) disrespectful white attitudes; 8) discriminatory administration of justice; 9) inadequacy of federal programs; 10) inadequacy of municipal services; 11) discriminatory consumer and credit practices; and 12) inadequate welfare programs.³⁷ Of those grievances, numbers 2, 3, 4, 5, 8, 9, 10, 11, and 12 are arguably worse today, while neighborhoods are quiescent. Their most cited grievance—police practices—was given the least attention and led to the least reform, while the change in grievance mechanisms (number 6) may be the single best explanation for the lack of riots today.

It is unlikely that extreme deprivation caused the riots. Susan Olzak and her coauthors, in a massive multiple regression of 1,770 racial and ethnic collection action events and 154 race riots in the United States between 1960 and 1993, found little evidence that riots were caused by extreme deprivation. Instead they found that riots occurred where the situations for blacks were improving, where neighborhoods had become less segregated, and where poverty had begun to decline.³⁸ Riots erupt when life improves for the poor and oppressed, they argue—an equally questionable conclusion.³⁹

The conclusions reached by the Kerner Commission were too limited. The riots were triggered not simply by weeks or months of tension-producing incidents but by a long history of violent policing of racial boundaries that had only intensified in the civil rights era. In a prescient study of seventy-five riots that erupted in black neighborhoods between 1913 and 1963, Stanley Lieberson and Arnold Silverman found that in those cities where ghetto residents rioted, blacks had lower rates of unemployment, better jobs, and higher wages and the wage differential between blacks and whites was smaller.⁴⁰ Yet it was not the improvements, they insisted, that precipitated the riots. It was police violence. Moreover, in those cities where ghetto riots erupted 1) the police force was white; 2) the local government was white, elected through a system “less sensitive to the demands of the electorate”; and 3) the businesses community was white.

Local government, they pointed out, is “one of the most important institutions to consider in an analysis of race riots. Municipal policies, particularly with respect to police, can greatly influence the chances of a race riot.”⁴¹ Riots are “more likely to occur where social institutions function inadequately, or when grievances cannot be resolved, or resolved under the existing institutional arrangements … such that a disadvantaged segment is unable to obtain recognition of its interests and concerns through normal political channels.”⁴² In other words, Lieberson and Silverman’s findings coincided with number 1 and number 6 of the Kerner Commission’s list of twelve grievances.

In 1970, Harlan Hahn and Joe Feagan reported similar findings in their study of riots in New York and Detroit. With “few exceptions, every major incident of urban violence was triggered by police…. it is likely that hostility of Negroes towards white authority has been kindled by abrasive contacts with ghetto police.”⁴³ Similarly, Robert Fogelson found that “with a few exceptions … the nineteen sixties riots were all precipitated by police actions.”⁴⁴ Yet none of these authors produced a full-length book on police riots, or on the interaction between the two; and no one attempted to generalize these findings and apply them to riots outside the United States.

The Kerner Commission findings were clearly overdetermined. Lieberson and Silverman’s, Hahn and Feagan’s, and Fogelson’s findings, on the other hand, were drawn from studies of American cities that had similar characteristics. It was difficult to know if their findings were replicable in a different context. A cursory examination of riots in Great Britain, however, gives some added support to their conclusions. Before I turn to studies of riots in Great Britain, however, I consider a final study of the American riots, one that provides a clue as to why some residents and not others respond to police violence with riotous actions.

In 1968 Edward Ransford conducted a survey of residents of the Los Angeles Watts neighborhood, many of whom had participated in the 1965 riots. All residents surveyed expressed an overwhelming sense of powerlessness born of “extreme discrimination barriers … a belief that all channels for social redress are closed.”⁴⁵ Although black militants shared this sense of political powerlessness, they exhibited a much lower level of individual powerlessness. Melvin Seeman later theorized that the black militants Ransford studied were less likely to riot because activism had enhanced their sense of self-worth: “The militants [were less] likely to be low in powerlessness when this refers to their sense of personal efficacy but high in powerlessness in the sense of being aware of (and fighting against) blocked social control (discrimination and bureaucratic arrangements).”⁴⁶ While both Seeman and Ransford stress the importance of personal efficacy in dissuading black militants from participation in riots, there is another explanation: Black activists participated in organizations that both prized disciplined strategic action and offered an alternative repertoire to pursue justice. Participation in successful collective actions increased activists’ confidence in their own collective capacity and made them less susceptible to the pull of flaming streets.

British Police Studies

Not long after race-related riots declined in the United States, the first major riots erupted in Great Britain. They began innocuously enough when in March 1981, following the “Black People’s Day of Action” parade, conflicts between police and predominantly black youths burst forth into small-scale confrontations. At the beginning of April, the metropolitan police initiated Operation Swamp 81, a campaign that included a massive sweep of the black neighborhood Brixton. Over 1,000 men were stopped and frisked in a matter of days. On April 10, police chased a badly bleeding young black man who had passed by them apparently after having been stabbed by a group of boys. When a police officer tried to take the wounded youth to a waiting car (allegedly to get him to a hospital), a crowd gathered to protect the young man. The officer called for police backup. Rumors began to spread that the police had let the youth die in custody, and the crowd began to turn on them. As more backup police arrived, young people pelted them with bricks and bottles. Lootings and then fires followed. In two days 28 buildings and 120 cars were set aflame and 117 stores were damaged and/or looted. Before the riots ended, 65 civilians and 229 police officers were injured.

Then in July riots tore through Handsworth, Birmingham; Southall, London; Toxteth, Liverpool; Moss Side, Manchester; Leeds and Leicester; Halifax in Southampton; Bedford in Gloucester; Wolverhampton and Coventry; and Bristol and Edinburgh. A national commission was formed to look into the cause of the riots. The official report of the commission, known as the Scarman Report after the presiding magistrate, determined that the riots “were not planned but a spontaneous outburst of built-up resentment sparked by particular incidents, … a loss of confidence and mistrust in the police and their methods of policing, … and racial disadvantage and racial discrimination.”⁴⁷

In 1985 a new wave of violent confrontations between black youths and white police officers tore through Brixton, the Handsworth suburb of Birmingham, the West Midland areas of Coventry and Wolverhampton, and then the St. Paul’s district of Bristol after a black taxi driver was arrested by a white police officer for a parking ticket. The worst riots occurred in the mostly black Broadwater Farm housing estate in the predominantly white Tottenham district of London. Residents complained that police had occupied the area and harassed, abused, and otherwise treated violently and disrespectfully the area’s residents. Police shot rioters with plastic bullets, and the police constable Keith Blakelock was murdered with a machete. Another 20 civilians and 223 police officers were injured.⁴⁸ The 1985 riots, like the preceding 1981 riots, had come on the heels of a spectacular rise in repressive drug and street crime policing.

The two waves of riots acted as a catalyst, generating a surge of new research. They also shifted the direction of the field from a highly theoretical academic discipline to a broader policy-oriented one. Most studies, however, continued to focus on the micro-processes and institutional culture of the police, a methodological approach heavily reliant on participant observation. David Waddington, Simon Holdaway, Clive Norris, Nigel and Jane Fielding, Charles Kemp, and Robert Reiner are among the most important British scholars in the field. David Waddington developed a “Flashpoint model” listing six factors that together explained why minor altercations could spiral out of control: 1) structural—poverty, unemployment, relative deprivation, and racial discrimination; 2) political/ideological—a group’s political legitimacy, power, and influence or lack thereof; 3) cultural—the rules, norms, and self-definition a group develops and the compatibility of those norms with those of the police and society at large; 4) contextual—history of negative interactions between a minority group and the police, leading to a breakdown in communication; 5) situational—specific spatial and symbolic characteristics of the site of conflict; and 6) interactional—personal miscommunication, signaling, and misreading of particular actions.⁴⁹

Waddington’s emphasis on conflicting interpretation of events rather than the events themselves and his insistence that the precipitating event might be quite minor make specificity difficult. Waddington recognizes that structural inequality has a deleterious impact on the relationship between police and racial minorities, but he does not identify the processes or mechanisms that structure inequality along race or ethnic lines, or the ways that police actions are rooted in and perpetuate categorical inequality. Moreover studies have provided scant support for his model. In a study of police stops and seizures, Norris and his coauthors found no evidence to support the claim that black youths behave more disrespectfully toward police than do white youths.⁵⁰ Police officers were two and a half times more likely to stop blacks than their presence in the population suggested, even when blacks were significantly less likely than whites to show disrespect toward the police. Police stopped whites when they were inebriated, so those stopped behaved less respectfully. They stopped blacks, in contrast, on speculative rather than evidentiary grounds. Blacks were thus far more likely to behave respectfully.⁵¹ Police actions, in other words, had nothing to do with the behavior of blacks. Rather police were biased a priori against blacks, and blacks correctly read the situations at hand.

Racial and ethnic prejudices are embedded in police occupational culture and work, insists Holdaway. Police officers speak in racially derogatory terms even toward their own black and Asian colleagues. Black and Asian officers “must affirm rather than challenge the values of their colleagues…. Black officers find a greater measure of acceptance among colleagues when they demonstrate physical prowess when dealing with an offender.”⁵² Holdaway illustrates the ways in which police work is embedded in systems of inequality and how police work depends on categorical mapping. Yet he pays insufficient attention to broader political processes. As Robert Reiner points out, British researchers became too close to the police they studied. The result was that they ignored “the wider context of social, political and economic change … concentrating on funded, short scale projects, examining trees while failing to remark on the forest.”⁵³

French Studies of Riots and Banlieues

The 2005 riots also sparked a wave of theorizing in France. Most explanations fell into one of two categories. One set of work was based on statistical analysis of the correlates of riots. In a broad comparative study of neighborhoods in which riots broke out and those where they did not, for instance, Hugues Lagrange observed that riots erupted where a high percentage of African immigrants, large families, and youths resided. Africans, he claims, were more likely to have larger families (sometimes although not always due to polygamy), and children from large families tended to do poorly in school (lacking basic skills) and were more likely to engage in criminal and delinquent activities and riots.⁵⁴

Fabien Jobard concurs in a series of articles drawn largely from Lagrange’s data.⁵⁵ Disputing claims made by unnamed American authors, Jobard argues that the French riots were not race riots, like American or British riots, since the rioters did not attack people or stores belonging to other races or ethnic groups (although he admits there were no such stores in the banlieues) or claim to speak on behalf of a race. Instead, he argues, in his analysis of the 208 people tried in Bobigny for rioting, that while polygamous families were rare (3 percent of case files),

[t]he average size of the families to which they belonged was typically very large: the average number of brothers and sisters was actually 4.6, while as many as one-fifth of the relevant families contained seven or more children. A family size of this kind is indicative of the presence of a large number of families from sub-Saharan Africa—a feature of French life highlighted by correlations produced by Lagrange. The geographical locations of the riots, especially in the west of France, are also closely aligned to major settlements of the new sub-Saharan migrants (and, indeed, such locations are characterized by the concentration of large families).⁵⁶

Concentrating their efforts on preventing the riots from spreading, Jobard insists, police in 2005 failed to crack down on rioters, allowing youths to remain “free to continue their activities unhindered in smaller territories surrounded by police forces.”⁵⁷ In 2005, Jobard argued that there was no evidence that French police were racially biased when they stopped “a given individual or group of individuals regardless of the signs of racism that may be observed here and there.”⁵⁸ In a later study, however, conducted with René Lévy in collobaration with John Lamberth and Rachel Nield from the Open Society Foundation, they found significant racial disparities in the frequency with which police stopped pedestrians. Researchers recorded police stops at several metro stations outside shopping areas downtown. They found that police stopped Arab youths 7.8 times and black youths 6.0 times more often than they did white youths.⁵⁹ Jobard and Lévy were careful to point out, however, that style of dress was a better predictor of being stopped than ethnicity alone. Police responded to the study by insisting that young people coming in from the banlieues, dressed in the style of ghetto youth, were more likely to shoplift (although they provided no evidence to support this claim). Since there had been no studies to dispute (or support) the police officers’ claims, and Jobard and Lévy had studied only shopping centers, the widely publicized study did not change policy makers’ perspectives on police reform. More critically, the study underestimated the degree of racial profiling by failing to record stops in banlieues: neighborhoods virtually occupied by police, with few stores or shopping areas.

The second set of studies was based on ethnographic research in the banlieues. The work of Didier Lapeyronnie, for instance, demonstrates both the strengths and weaknesses of this methodological approach. On the one hand, through years of work with neighborhood youths, Lapeyronnie discovered that police violence and racial discrimination caused more anger and frustration among banlieue youths than poverty or unemployment. In a 2006 article in Déviance et société he argues that riots were a form of collective action, sparked by police violence and a perception of conflict between us and them:

This common negative experience generates an “us” which manifests itself in opposition to the police, indeed permanent conflict with them. The “we” thus negatively constituted by hostility toward the police is not devoid of significance…. In activating the “us” and forcing the change of the legitimacy of the imperative frameworks, the incident also renders the revolt both expected and legitimate…. The death of young people at the site of police interventions always elicits strong individual and collective emotion and crystallizes the feeling of “us,” victims of injustice, in opposition to a “them,” an unjust police force.⁶⁰

On the other hand, like many ethnographers, Lapeyronnie reifies space and place, projecting motives and character faults to an entire neighborhood. He exaggerates the isolation of neighborhood residents and describes their lives as if they ended at the neighborhood edge. As such, he narrows his sites on a microcosm of dysfunctional interpersonal relationships, cultures, and behaviors. Much like the works of others who have looked at the “culture of poverty,” Lapeyronnie’s study is reminiscent of the Chicago school—a view of the city as a “mosaic of little worlds which touch but do not interpenetrate,”⁶¹ where “the primary [source] of social integration and order”⁶² is the geographic community and poorly integrated neighborhoods foster an array of bad behaviors. In 2001 Lapeyronnie attributed riots to maladjustment: “Discomfort and self-doubt become a permanent condition and often end in self-destruction. Alcohol, drugs, and narcotics consumption are only the most banal forms. Riots, ‘rodeos’ and excessively risky behavior, and finally excessive TV viewing are all part of such self-destructive behavior” (emphasis mine).⁶³

Despite the fact that his 2006 article seems to be a relational analysis of riots as forms of collective action provoked by police violence, Lapeyronnie fails to trace these dynamics either in the conclusion to the article or in his 2008 book.⁶⁴ Instead he returns to the same neofunctionalist analysis that dominates his earlier work: residents of the banlieues appear to have a singular set of emotional and mental processes.⁶⁵ He says,

Through the emotion that is felt, the individual directly demonstrates his attachment to the “us” and the solidarity which connects him to those who share the same feeling, the same mindset. For a moment, he distances himself from the triviality of his own reality in order to submit himself to a “force” completely outside of himself. Because of the riot, the affirmation of “self” melts into the affirmation of “we.” The individual charges himself with an energy which allows him to move to action, a sort of electricity, says Durkheim, an “emotional energy” which is also moral for him since it is strongly linked to the attachment to life or to respect.⁶⁶

But residents of large urban areas do not all know each other or share similar characteristics. The reification of such spaces deflects attention from the real “relationship of controller and controlled,” as Robert Sack cogently notes.⁶⁷ In contrast to those who see conflicts between police and minorities as the results of dysfunctional pathologies, alternative cultures, and/or the misinterpretations of events, I argue that such conflicts are embedded in the larger structural dynamics of, and resistance to, exploitation and categorical inequality. As Elizabeth Wood so eloquently puts it, “[T]he values, norms, practices, beliefs and collective identity … evolved in response to the experience of the conflict itself.”⁶⁸

Still, many ethnographers depict the plight of banlieue and minority youths with sensitivity and empathy. David Le Poutre, Stéphanie Beaud, Michel Pialoux, Michel Kokoreff, and Christine Bachman describe a host of serious problems that minority youths in banlieues confront, including poverty, unemployment, poor housing, discrimination, and humiliating encounters with teachers, police, and other state officials. Kokoreff, for instance, writes of the negative, deleterious role played by the police in the banlieues: “The urban police no longer appear as a cause of peace; on the contrary, they arouse fear. The principal tool of the forces of order is the identity check, which systematically heightens tension instead of relieving it.”⁶⁹

French Studies of Police

Most French specialists on policing articulate an interactive model similar to that of David Waddington. Sophie Body-Gendrot, for instance, rejects studies whose subtext is that “large immigrant families concentrated in public housing estates are, themselves, the source of the problems”⁷⁰ and yet argues that police “are tired of being humiliated, sneered at and physically attacked by disenfranchised youth.”⁷¹ Similarly, Roché describes numerous incidents of police abuse but concludes, “In my view, a police action alone cannot explain a riot. Riots are only the pinnacle of a perennial conflict between particular groups of young men and police…. From one side to the other, insults fly, blows or more are exchanged, and masculine intimidation becomes part of the routine. Relationships are particularly poor with youth of foreign origin. [One study suggests that] youth of North African immigrant backgrounds see the police as a rival gang, as a racist force.”⁷²

Laurent Bonelli, Laurent Mucchielli (at times with Marwan Mohammed or Abderrahim Ait-Omar), Farhad Khosrokhavar, and Didier Fassin have been the sharpest critics of both the politics of crime control and police conduct in poor minority banlieues.⁷³ Bonelli argues that the “culture of results” encourages police to look for crimes rather than prevent them. Perpetual identity checks lead to escalating conflict: “Each victory for one side is a defeat for the other.” When a young boy dies during a police operation or is shot by the police, “the neighbourhood burns.”⁷⁴ Khosrokhavar points out that Muslim youths are far more likely to be imprisoned than white French for the same behavior or crime.⁷⁵ Marwan Mohammed and Laurent Mucchielli point out that most riots in France over the past twenty-five years have been confrontations between banlieue youths and police and that police intrusions and violence rank top place among the fears and complaints of minority youths.⁷⁶ The political elites are largely to blame. They deliberately foster panic during election campaigns and then deploy police against banlieue youths under the guise of controlling crime.⁷⁷

Didier Fassin’s engrossing and elegant ethnography of the French police stands alone.⁷⁸ Fassin challenges the Flashpoint paradigm commonly used by French scholars. It is not the actions of minority youths that provoke police violence, he contends, but rather the discourse and actions of French political elites, who have deliberately stoked racial fears and encouraged police to act like occupying armies in poor banlieues:

Among the hundreds of such incidents I witnessed, almost the only ones in which the individuals concerned displayed insolence involved youngsters from middle-class or wealthy backgrounds, particularly students who evidently had no experience of this kind of situation and seemed unaware of the potential consequences of their behavior. Yet in none of these cases—which were anyway quite infrequent, since these groups rarely face such procedures—did the officers seek to escalate the tension in order to provoke a scene that could later be set down as insulting and resisting the police.

Conversely, when checks were carried out in the projects or on the streets of the city, young people, mostly of working-class background and non-European origin, almost always kept a low profile, only speaking when they were asked a question, not reacting to the abusive or racist comments and aggressive or humiliating treatment some officers subjected them to, simply presenting their papers and submitting to the body search. Accustomed to and even blasé about these repeated irritations, knowing quite well what would happen if they protested, they appeared to be waiting until the bad moment passed, silent, expressionless, for the only way not to lose face in this confrontation was not to enter into any transaction with the police.⁷⁹

Fassin attributes the 2005 riots and nearly every major urban disturbance from Watts to London over the past fifty years to racial profiling and police aggression in “disadvantaged neighborhoods, usually leading to the death of youth belonging to a racial or ethnic minority group.”⁸⁰

Comparative Studies

There are relatively few systematic cross-national and comparative studies of riots. Three of the strongest such approaches are those of Loïc Wacquant, Janet Abu-Lughod, and Michael Katz.⁸¹

Wacquant’s scathing discussion of the link between neoliberalism and mass incarceration and his devastating critique of Chicago school analyses of urban poverty alone make his work worth reading. But his explanation for riots is less convincing. The American riots of the 1960s and the French riots in 2005 were the results of diametrically opposite processes, he insists. American riots were “propelled from outside by the crumbling of the caste system”; French riots were the result of the slow decomposition of working-class neighborhoods from within. The first was caused by “the restructuring of urban capitalism and the policy of social regression of the federal government set against the backdrop of continued ostracization of African Americans,” and the second “by the triangular relationship between the state, social classes and the city.”⁸²

Banlieues, Wacquant claims, are antighettos, places of integration and state intervention. In the United States we have “racial cloistering,” whereas in France the comparable pattern is one of “ethnic dispersion and diversity.”⁸³ “Twenty-five years after the great race riots of 1965–1968, African American neighborhoods of New York, Chicago, Philadelphia and Detroit have the look of war zones [while] degraded working-class banlieues, renamed ‘sensitive neighborhoods,’ have been the target of a concerted renovation plan under the heading of Neighborhood Social Development.”⁸⁴

Wacquant’s explanation is unsatisfying. First, he does not explain why different processes in divergent societies led to similar outcomes. If the American riots were provoked by improvements in their environment (the crumbling of the caste system), why did French riots erupt when things were getting worse (the decomposition of working-class neighborhoods)? And why did the decomposition of working-class neighborhoods provoke riots only in France and only in neighborhoods where blacks and Arabs were concentrated? The decomposition of working-class neighborhoods was certainly a more global phenomenon. Wacquant fails to offer a systematic explanation.

Second, Wacquant exaggerates the stability, homogeneity, and isolation of black ghettos in American cities, particularly during the 1960s, which is when most riots occurred. Indeed, as Wacquant has observed, American ghettos were not yet “ethnically and socially homogeneous universe[s] characterized by low organizational density.”⁸⁵ Many, such as Harlem, were characterized by high density and overcrowding. Others, such as the South Bronx and Brooklyn, were working-class areas in the process of decomposition—although admittedly decomposing far more rapidly than the French banlieues and spurred by different socioeconomic processes (specifically highway expansion, urban renewal, mass exodus of industry, and white flight). The timing of the new migrants was similarly miserable: black and Puerto Rican migrants arrived just as the last factories were closing and blue-collar jobs were exiting the city. Massive changes to the American city were blamed on the new arrivals, giving grist to politicians who promised to protect white citizens with punitive policing measures, much as French politicians did at the turn of the twenty-first century. Indeed both American ghettos and French banlieues, and not, as Wacquant claims, simply the former, were “anchored and aggravated by public politics of urban triage and neglect.”⁸⁶ And both were only “ethnic” in the “demand that the state, precisely, cease to treat them as such.”⁸⁷

Third, Wacquant’s admittedly neofunctionalist explanation is devoid of actors. If the 1960s riots in America were, as he claims, “propelled by the crumbling of the caste system,” why did blacks and Puerto Ricans abandon the streets before the caste system crumbled? If the real demands of rioting youths in France were “decent jobs, good schools, affordable or improved housing, access to basic public services, and fair treatment by the police and other agents of the state”;⁸⁸ if anger was directed at police as the last “buffer between them and a society that rejects them”;⁸⁹ and if the 1992 riots in Los Angeles were “as much about empty bellies and broken hearts as about police batons and Rodney King,”⁹⁰ why do American youths so rarely riot now? If the prison, as Wacquant claims elsewhere, was the main medium of social control, why was there a fifteen-year gap between the denouement of the riots and the explosive growth of the prison system? Finally, what explains the willingness of black and Arab youths in France to set their neighborhoods aflame if they lived in a country devoid of racial exploitation and their neighborhoods had been the “target of a concerted renovation plan”?

Janet Abu-Lughod’s book Race, Space and Riots in Chicago, New York and Los Angeles is a useful point in contrast.⁹¹ Unlike Wacquant, who treats the United States as Chicago writ large, Abu-Lughod contrasts the riots in that city to those in New York and Los Angeles. Of the three, New York had the fewest, briefest, and least damaging riots. Abu-Lughod attributes this to police training and restraint. The heterogeneity, accessibility, organization, and political power of New York’s black and Latino neighborhoods were necessary conditions for the development of better policing strategies and better police resident interactions. Where neighborhoods were more segregated, as in Chicago and Los Angeles, police forces were less accountable.

A long history of black grassroots political organizing and a less violent white political establishment explain the smaller and less frequent riots in New York. Abu-Lughod’s analysis supports a boundary activation model. Nonetheless there are some flaws in her analysis. First, segregation in large parts of the city, according to most indices, was worse in New York than in either Chicago or Los Angeles in both the 1960s and the 1990s.⁹² Integration is not a good explanation for fewer riots in New York. Second, Abu-Lughod uses different explanations for different riots, making it difficult to derive theory from her comparative method. While she explains the riots of the 1960s as unique historical events ignited by the recession and the changing course of the civil rights movement, she attributes the 1992 riot in Los Angeles to the superimposition of race, space, and poverty. Yet Los Angeles was hardly alone. Why did so few American cities follow suit, or in Katz’s words, “not burn”? Why did riots instead erupt in Paris in 2005, Manchester in 2011, and Stockholm in 2013?

Despite Abu-Lughod’s recognition that police violence sparked riots in five out of her six cases (she blames the assassination of Martin Luther King Jr. for the other), she barely explores the phenomenon. She claims that police training and restraint discouraged riots in New York, and yet police abuse spiraled during the 1990s. These gaps aside, Abu-Lughod’s keen insights and attention to the role of racial segregation and police violence in provoking riots, and the mitigating role played by social movement organizations in preventing them, provide a rich avenue for theorizing.

In “Why Aren’t U.S. Cities Burning?,” “Why Don’t American Cities Burn Very Often?,” and Why Don’t American Cities Burn?,⁹³ Michael Katz addresses the central paradox straight on. Income inequality in the United States, he notes, swung sharply upward after 1973:

The proportion of African American men out of the regular labor force soared. Among twenty-six to thirty-year-old black men, labor force non participation leaped from around 9 percent in 1940 to 30 percent in 2000…. The number incarcerated skyrocketed, jumping 82 percent during the 1990s; 49 percent of prisoners compared to 13 percent of the overall population were black. On any given day one of three twenty to twenty-nine-year-old black men was either in jail or on probation, parole or both. Nor did allegations of police violence disappear as, for instance, in reactions to the 1997 brutalizing of Abner Louima while in the custody of New York City police.⁹⁴

One study, he notes, found that “the spatial concentration of the poor rose dramatically in many U.S. metropolitan areas. The number of poor people living in high-poverty areas doubled; the chance that a poor black child resided in a high poverty neighborhood increased from roughly one-in-four to one-in-three; and the physical size of the blighted sections of many central cities increased even more dramatically.”⁹⁵ Yet riots erupted only in France.

Katz also explains this puzzle by focusing on boundary changes, which he argues, à la Tilly, “strongly affect the likelihood, intensity, scale and form of collective violence.”⁹⁶ Since Katz is principally interested in why American cities no longer burn, he focuses on boundary deactivation, specifically three far-reaching changes to the structure of race relations in the United States that have deactivated racial boundaries and made riots less likely and less frequent. Katz labels them the ecology of power, the management of marginalization, and the incorporation and control of immigrants.⁹⁷

Boundaries accentuate and cement inequality, Katz observes, by accelerating the accumulation of advantages and resources. When such boundaries are challenged, as they were when blacks migrated north to American cities, violence is often the result: “Between 1950 and 1970 the black population of many cities skyrocketed: in Newark from 17 to 54 percent, in Chicago from 14 to 34 percent, and in Detroit from 16 percent to 44 percent…. To preserve existing boundaries whites often turned to violence—a response documented with painful detail by many historians. The point for this discussion is that civil violence erupted at the height of urban boundary challenges, when huge numbers of African Americans had moved in and whites had not yet moved out.”⁹⁸

The first change identified by Katz was to the ecology of power in American cities. In the late 1960s, white flight left many cities under the control of mayors representing newly empowered black and Latino majorities. Between 1965 and 1980 the number of black elected officials jumped from 100 to 1,813.⁹⁹ Although deindustrialization and cuts in federal spending decimated the power and capacity of these mayors, black opposition movements now often faced black mayors and even black police. “The new ecology of urban power dampen[ed] the potential for civil violence by pairing class and racial segregation with the devolution of control over space to previously marginalized groups.”¹⁰⁰

Second, the management of marginalization (operating through five separate mechanisms—selective incorporation, indirect rule, mimetic reforms, consumption, repression, and surveillance) further derailed mobilization. Many blacks found employment in the public or quasi-public sector, and their selective incorporation into the middle class was fragile and dependent. Mimetic reforms further diverted class action into more manageable arenas. The decentralization of school districts replaced the struggle for community control, election campaigns replaced protest movements, and, as Katz approvingly cites Katznelson, “modest but sufficiently tantalizing distribution [substituted] for redistribution.”¹⁰¹

The combination of white abandonment, selective incorporation, and mimetic reform allowed powerful white Americans to rule black America through powerless black leaders: real power remained in the hands of the white establishment. “But indirect rule meant that civil violence or other claims on city governance would be directed towards African American elected officials, African American public bureaucracies and African American police.”¹⁰² Increasing consumption and blacks’ indebtedness further decreased motivation and arenas for political action. Mass incarceration, parole and felony disenfranchisement, and police actions designed to pit black gangs against each other weakened black resistance.

Third, the incorporation of immigrants lessened the risk that blacks might use violence to demand rights: “For the civil violence that rocked Paris and frightens other Europeans is a product of recent immigration, not the grievance and frustrations of historically marginalized citizens…. The two events—civil violence in France, peaceful protests in the United States—highlight divergent relations of immigrants to the state and to the economy. America’s immigrants sought redress through government…. They want nothing so much as the rights of American citizens.”¹⁰³

There are, however, several problems with Katz’s otherwise compelling analysis. First, new immigrants did not riot in France. Of the more than 4,000 arrested, only 120 were born outside France.¹⁰⁴ Those who rioted were French—black and Arab children and grandchildren of former colonial subjects, and, in the case of Algerians, former Frenchmen. While the French continued to refer to these youths as immigrants, their lives, in fact, paralleled those of blacks and many Latinos in the United States.

Second, the concept of indirect rule does not explain why black and Puerto Rican residents of New York City—a city with a white mayor who barely disguised his contempt for the black community, excluded blacks from his administration, and did nothing to increase minority representation in the overwhelmingly white police force—did not rise up in the 1990s. Even after the brutal slayings of Anthony Baez, Patrick Dorismand, and Amadou Diallo and the torture of Abner Louima the city remained quiet.

Third, repression is not a good explanation for quiescence, since Katz, as well as others, uses it to explain revolts. Finally, and most critically, Katz fails to question some of the most common explanations for the American hot summer or ghetto revolts. Poverty, racism, and unemployment may cause great pain and misery, but human beings are capable of tolerating great injustice in silence. Before we can understand boundary deactivation, we need to reexamine boundary activation—those rare, singular moments when all social interactions revolve around a single us/them boundary. In other words, we need to sort through the smoldering ashes of urban unrest before we can explain the quelling of the fire.

The Argument

In countries with large unaffiliated or detachable segments of the electorate, appeals to racial fears can provide the margin of victory in tight elections. The fear of being labeled soft on crime, immigration, or security encourages a rush to the right as the major political actors engage in competitive outbidding. Political campaigns capture media attention. Sensationalist and racially distorted media stories put additional pressure on political candidates in tight races. As Lewis Dexter observes, a politician “makes the world to which he thinks he is responding.”¹⁰⁵ The bellicose rhetoric of the campaign, as Fassin observes, “legitimizes not only the police’s views of the situation but also the way they work to impose order.”¹⁰⁶ The use of phrases such as “war on crime” or “war on drugs” evokes a military logic and leads to increasingly violent policing in poor minority neighborhoods.

Racially targeted police violence inflicts an ugly wound: it undermines the legitimacy of the state and sends the message that the lives of some of its citizens are not valued. In the absence of strong social movements with a standard nonviolent repertoire or state action (real or potential) on behalf of injured communities, a particularly egregious incident of police violence, such as the killing of an unarmed youth, may incite riots. Most riots begin as nonviolent gatherings and pleas for justice by families, friends, and neighbors of the victims. Police repression of such gatherings encourages others to join the fray. Network ties between residents of an affected area and those who live in places with similar conditions lead to riot diffusion, particularly to areas where affected minorities comprise the majority of residents. Where social movements or other social or revolutionary organizations channel anger and enforce discipline, however, or where alternative avenues of redress exist, riots are rare. The process, then, often proceeds as follows:

1) Voters from the dominant majority find stigmatization and control of racialized minorities politically attractive and credible, especially when political, spatial, or social-economic change threatens established racial boundaries and creates insecurity.

2) Politicians in tight races attempt to appeal to a detachable section of the electorate by playing to racial fears.

3) The political scramble to avoid being outsegued on the crime issue leads major political actors to engage in increasingly harsh law-and-order public discourse.

4) Increased political attention to crime leads to spikes in sensationalist and racially distorted media coverage.

5) Politicians favoring harsh policing measures are elected and pass harsh punitive crime and/or anti-immigrant laws.

6) Police interpret such signals to mean they are immune from prosecution when interacting with members of subjugated groups.

7) There is a dramatic increase in levels of police brutality in stigmatized minority neighborhoods, leading inexorably to a particularly egregious act of violence—usually the killing of a young unarmed minority male.

8) The state is unwilling to hold the police officers accountable or, worse, takes the side of the police officers against the victim.

9) Lacking other options (such as established institutions to address or channel grievances, strong social or community organizations, or a repertoire of successful nonviolent protest), communities explode.

10) As knowledge of or rumors about events spread, riots diffuse to surrounding neighborhoods, then to cities and towns with similar conditions, particularly those where affected minorities constitute the majority of residents. National-level incidents (such as the assassination of Martin Luther King Jr.) and officeholders provoke more rapid riot diffusion than neighborhood-level events or officeholders.

What is key is not the sum of these elements but the way in which they concatenate: the mechanisms that link an incident of police violence to riots or, alternatively, to other forms of contentious politics. This book traces the processes and sequences of interactions and events in two cities with vastly different social and political structures, racial constructions, and political economies. Police use violence against subjugated minorities in both cities, but in one city subjugated minorities, particularly poor minority youths, respond in ways that sometimes culminate in full-fledged riots. In the other, subjugated minorities and victims of police abuse more often engage in conventional forms of individual and collective action, organizing protest marches and civil disobedience; petitioning district attorneys, federal prosecutors, political officeholders, and members of the Department of Justice; and filing civil law suits.

In 1964 riots broke out in New York when a) police violence was ignored and unofficially encouraged; b) mass migration, white flight, urban renewal programs, and highway construction devastated black and Puerto Rican neighborhoods, overwhelming the capacity of black and Puerto Rican organizations to channel anger into sustained nonviolent action; and c) blacks and Latinos were largely denied access to local courts, civilian review boards, or other forms of legal redress. Riots broke out in 2005 in Paris when a) growing police violence against blacks and Arabs was officially ignored and unofficially encouraged; b) the discourse of republican equality did not allow open discussion of racial dynamics, much less the development of civil and social organizations to address these dynamics; c) revolutionary organizations such as the Front de Libération Nationale (National Liberation Front of Algeria, or FLN) had long ceased to exist; d) courts refused to hold police officers accountable for violence used against North and sub-Saharan African immigrants or black and Arab youths, and those who charged police officers with civil offenses were far more likely to be sentenced for rebellion than to win their cases.

The structure of policing shaped the geography of urban unrest. In the United States, the decentralized structure of the police, with each unit under the control of a local mayor, led to a segmented, staggered pattern of riots during the 1960s, which crisscrossed the country for half a decade. Riots erupted most violently in Los Angeles, Detroit, Chicago, Washington, D.C., and Newark, cities where residential patterns were most segregated and police abuse most profuse. By the late 1970s riots had become rare. Still in some cases—most notably Liberty City, Florida, in 1980 and Los Angeles in 1992, where racial boundaries were particularly contentious and where police operated with virtual impunity—they reoccurred. Yet even those extremely damaging riots were isolated geographically, as the federal government intervened to subdue the rioters and hold local police accountable for their actions.

In France, in contrast, the centralized structure of policing produced a different pattern of urban unrest. Since the police were national, it was the national government that determined local reactions. The centralized structure of policing ensured, as Katz observes, that “[a]ntagonism towards the police reinforce[d] the distrust of the national government…. [In 2005] their protests, neither planned nor coordinated, reflected frustration, rage, alienation, and a lack of confidence in or access to official political channels. In this they resembled African Americans in the 1960s more than immigrants to the United States late in the twentieth century.”¹⁰⁷ When then minister of the interior Nicolas Sarkozy announced on the radio that “his police had done nothing wrong” and threatened to take action against delinquents and hooligans, he provoked rapid riot contagion throughout France within days of his statement.

The logic of inquiry, as Gary Goertz and James Mahoney note, is looking for the “cause of effects” (starting with real events and moving backward).¹⁰⁸ Three elements, I argue, are present in most ghetto riots: activated categorical boundaries; violent transgressions across categorical boundaries by police or other security forces (or vigilantes protected by security forces) usually leading to the death of an unarmed youth (in interethnic riots the violence transgression is committed by members of dominant groups; in opportunistic riots, a pattern of violence against subordinate groups is suddenly disrupted when the repressive capacity of the state precipitously declines); and failure of state officials to hold police or security forces accountable. These three variables combined make riots more likely.

A wider array and combination of variables avert riots. The explanation is asymmetric, as Goertz and Mahoney observe of most explanations in qualitative research: “The causes of failure outcomes are not necessarily equivalent to the absence or negation of the causes of success outcome.”¹⁰⁹ American scholars have attributed the decline of riots to a host of factors. On the one hand, the expansion of Great Society programs, the election of blacks to political office, and the integration of police departments are said to have co-opted activists and deactivated racial boundaries. On the other hand, the decline of riots has been attributed to the mass incarceration of black and, to a somewhat lesser degree, Latino males. As Pamela Oliver notes, “The crucial thing to understand is that a repressive strategy initially triggered by massive urban unrest and other social movements was maintained and expanded long after the riots abated. It was not aimed at preventing unrest by repressing riots: it was preventing unrest by repressing potential rioters. People were not arrested and incarcerated for dissent or even for rioting: they were arrested and incarcerated for crimes.”¹¹⁰ Mass incarceration prevented riots, she notes, “by removing people from the system before they commit[ed] the undesired action.”¹¹¹

While all of these factors may have reduced the likelihood of riots, none of these explanations answers our puzzle. The first set of factors has deactivated racial boundaries in many cities but does not explain why riots did not erupt in New York during the 1990s despite escalating racial tensions and police violence, when a white mayor, a white administration, and a white police force held sway. Similarly, mass incarceration may have devastated inner-city neighborhoods but cannot account for either the sharp decline in riot frequency since the mid to late 1970s (long before the massive climb in incarceration rates) or the mammoth explosion in 1992 in Los Angeles (when California incarceration rates were substantially higher than the average and New York’s were substantially lower). Indeed, California, having one of the higher incarceration rates in the country, is among the most riot-prone states, perhaps due to the devastating impact that mass incarceration has had on activist networks.

Two factors have received far less attention but have transformed the way in which blacks and Latinos in New York, and in much of the rest of the nation, now respond to police homicides. Together they have dramatically reduced the likelihood of riots. First, social movements born in the cauldron of the great race riots of the 1960s now intentionally and unintentionally channel anger into more organized forms of collective action. Social movements provide a standard repertoire of action. Activists who had cut their teeth on the great race riots of the 1960s led black and Puerto Rican power movements in the 1970s and in the 1980s formed networks of community-based organizations on once riot-strewn streets. By the 1990s they were organizing around a host of critical local issues: struggles over control of local school and area policy boards; the creation of joint planning councils; and the availability of low-income housing and community gardens. In addition they organized against racial profiling and police brutality. Gregg Carter claims that the riots stopped because “you can only burn down your neighborhood so many times.”¹¹² Similarly a neighborhood activist on the Lower East Side told me, “If you look at areas like Watts and Newark, they are still rebuilding from the riots that took place then.” To a certain extent that is true, but other avenues had to become available. Now when police kill, activists and social movement organizations converge. They give solace to the families and friends, lead mass marches, demand indictments and federal interventions if those fail, and help families file civil suits. Some even sit on civilian review boards.

Second, the passage of significant civil rights laws has opened the courts to black and Latino plaintiffs and made the federal government a potential ally.¹¹³ “Victims of discrimination,” note Rogers Smith and Desmond King, increasingly seek “relief in the federal courts.”¹¹⁴ Courtroom battles have replaced street struggles. Demands for individual reparations have replaced demands for social justice. Not-guilty verdicts in criminal trials for homicide are not the end of the road. The Justice Department can try police officers for the violation of victims’ civil rights. Families frequently file civil suits. Between 1977 and 1998 only three New York City police officers were convicted for homicide while on duty.¹¹⁵ Yet during the 1990s New York spent approximately twenty-five million dollars a year to settle police violence cases out of court.¹¹⁶ Legal proceedings provide a clear course of action and reduce feelings of impotence, and they take years to pursue. At the end of the process (whatever the verdict) the energy of families, communities, and social movement activists has been depleted. As Vincente “Panama” Alba, the director of the Coalition to Fight Police Brutality, told me, “People burn out, get frustrated. The mobilization necessary to do justice in one case is extraordinary. It is difficult to go on and on.”

In sum, the civil rights movement and the riots that followed changed the environment for families and friends of those killed by police. On the one hand, wars on drugs and crime have led to the incarceration of two million mostly young black and Latino men and the penal supervision of another seven million. On the other hand, community organizations have developed a standard repertoire for dealing with police violence. Victims and their families now fight police violence through a combination of organized street demonstrations and legal justice. Although homicide convictions of on-duty police officers are exceptional, growing numbers of attorneys specializing in civil actions have convinced many cities to routinely settle police brutality cases out of court. Together these factors have reduced the likelihood of riots, even where white mayors, white administrations, and white police forces hold sway.

The deactivation of racial boundaries has also made riots rare in Marseille, the only city in France that did not burn in 2005. Despite the city’s endemic poverty, unemployment, and crime and its large immigrant and Muslim populations, political leaders in Marseille rejected the dominant French republican assimilationist paradigm. They created integrated spaces for deliberation and interaction and recognized and consulted regularly with leaders of ethnic organizations to improve relationships between police and minority youths (largely by letting police develop long-term relationships with neighborhoods rather than shifting them about). This and Marseille’s integrated downtown streets and beaches made racial boundaries less polarized than those of Paris, Lyon, or other major cities. But Marseille’s political system is riddled with corruption and ties to organized crime. Ironically, even mafias deactivate racial boundaries and make riots less likely, as I explain in Chapter 4, by creating a weblike political structure in place of a bifurcated boundary. They also punish those who engage in undirected violence or strike out on their own. (Organized criminal groups often play a similar role in the United States.)

Defining Riots

This is a study of modern urban riots, what have sometimes been called ghetto riots or urban uprisings. In the modular modern urban riot unruly crowds burn, loot, or otherwise assail stores, public buildings, cars, and/or symbols of state power. They also engage in confrontations with police, usually by throwing garbage or other objects. Some rioters may attack members of dominant groups, but these are usually opportunistic encounters and not central activities. Most participants attack only property. These uprisings include elements of what Charles Tilly has labeled “broken negotiations,” “scattered attacks,” and “opportunism”—that is, violent reactions to transgressions or violations of unwritten (or written) social pacts or agreements; symbolic and other attacks on property (weapons of the weak); and taking advantage of disorder either to improve one’s material standing or to wreak revenge.¹¹⁷ As in other forms of violence, riots ensue when “actors on at least one side respond by engaging in coordinated attacks on sites across the boundary while those on the other side engage in defense against those attacks.”¹¹⁸ Although property damage in such riots can run high, most deaths result from police killings. In the 1960s, as Thomas Sugrue notes, “Only a handful of cities—notably, Detroit, Newark and Los Angeles—accounted for nearly all the deaths. And, most of the casualties were the result of law enforcement actions against blacks, not black violence against the police or white bystanders.”¹¹⁹

The term “riot” is also used to describe forms of violence having little resemblance to the typical ghetto riot, other than activated categorical boundaries and violent transgressions across those boundaries. Rampages by dominant groups against stigmatized minorities, commonly called race riots, are closer to pogroms. They are usually either instigated or protected by the police, such as the white riot in 1900 in New York, which was set off after a black man stabbed a white off-duty policeman, and the anti-immigrant riot in Vitry, France, in 1977, instigated by the Communist mayor. Other examples include the smaller anti-immigrant riots in Germany in the 1990s.¹²⁰ Similarly, large-scale interethnic riots, with their extremely high death tolls, more closely fit what Charles Tilly has called coordinated destruction.¹²¹ The defining feature of coordinated destruction in addition to activated boundaries is that at least one side employs specialists in violence and includes powerful, well-connected people willing to use lethal violence to acquire or maintain control over valued resources and/or extend their jurisdiction over territory. “The communities are organized only along intra-ethnic lines and the interconnections with other communities are very weak or even nonexistent,” Ashutosh Varshney similarly observes, making “ethnic violence … quite likely.”¹²² The most deadly are those between groups with significant size and long-standing grievances, such as those between Hindus and Muslims in Asia. If only one side includes violence specialists, the conflict may become genocide. Donald Horowitz, one of the world’s experts on such conflicts, does not believe that ghetto uprisings should be called riots at all, given how dissimilar they are in form and content to these violent conflagrations.¹²³

Some ghetto riots include aspects of interethnic riots, as when minority groups attack members of other minority groups or, more uncommonly, members of the dominant majority. Some interethnic conflict between blacks and Koreans in the course of the 1992 Los Angeles riots, for instance, may have been motivated by a Korean grocer’s killing of a fifteen-year-old African American girl a year before the riot and the white judge’s decision (a week before the Rodney King verdict) to reduce the jury’s recommendation for a sixteen-year sentence for manslaughter to a five-month suspended sentence and five-hundred-dollar fine. Yet, even here only two Asians were killed and most rioters attacked property rather than individuals. In most ghetto riots the boundary activated is not interethnic but rather drawn between stigmatized minorities (both blacks and Latinos in the United States and blacks, Arabs, and Berbers in France) and the state. Police, the only specialists in violence, do the bulk of the killings.

Ghetto riots often include elements of opportunism, as disorder opens up possibilities for personal profit.¹²⁴ As Tilly observes, individuals “looted when they saw that law enforcement had disappeared and that store owners had lost control over their premises.”¹²⁵ One woman told an interviewer during the 1965 Watts riot, “Well, you could see all the stuff lying there and all those people going in and out and somebody was going to take it, so I thought I might as well take it for myself.”¹²⁶ Another, churchgoing woman put it this way: “It dawned on me as I was passing a certain store that I had been paying for my present television for five years and [therefore] the store owed me five televisions. So I got three and I believe the store still owes me two.”¹²⁷ A self-identified looter in Detroit in 1967 told Nathan Caplan that the blacks were “trying to get the goods from the white folks because the white folks own everything and they [blacks] were just trying to get something so they can own it.”¹²⁸ Even in opportunistic looting, as Tilly notes, activated racial boundaries divided “the overwhelmingly black looters and fire bombers from the overwhelmingly white fire and police departments.”¹²⁹

Opportunity alone rarely leads to riots. Where there is a long-standing pattern of police abuse of stigmatized minorities, however, a sudden decline in the state’s repressive capacity may spark riots. As with other forms of collective action, the order and magnitude of threat and opportunity lead to different patterns of mobilization. Robert Curvin and Bruce Porter, for instance, found that the pattern of participation (based on the arrest records) during the 1977 blackout riots (which erupted as the result of the decline in the state’s repressive capacity, that is, opportunity) reversed that of the 1960s riots.¹³⁰ In the 1960s those arrested during the initial days had no previous arrest record. Only toward the end did those with criminal records comprise the majority of those arrested, indicating a shift to opportunistic looting. In the 1977 riots, in contrast, those arrested during the initial period tended to have criminal records and only toward the end did those without criminal records comprise a significant proportion of those arrested.¹³¹ I will return to the 1977 riots in Chapter 1. Suffice it to say that such sudden and lengthy declines in the state’s repressive capacity are extraordinary.

Methodology

This book is based on over fifteen years of intermittent ethnographic and participant observation in New York City and greater Paris. I have tried to answer the following three questions: 1) why police behaved similarly with very different minorities in very different contexts; 2) why riots erupted in Paris and New York half a century and an ocean apart; and 3) why riots did not erupt in New York in the 1990s, when a white mayor held power and wielded it through a nearly all-white police force, or in Marseille in 2005. The methodology is threefold. First, I use a series of paired structured comparisons: 1) between New York in the 1960s and Paris in 2005, when both cities had riots; 2) between Paris in 2005 and New York in the 1990s, when only Paris had riots; and 3) between Paris and Marseille in 2005, when only Paris had riots.

Second, I engage in process tracing working backward from incidents of riots and nonriots respectively, what Goertz and Mahoney have called cause of effect.¹³² I use interviews and participant observations to trace the mechanisms that led incidents of police violence to culminate in riots or alternatively other forms of contentious action such as lawsuits, community protests, civilian review board complaints, the formation of new organizations, or political speeches and rallies. As Sidney Tarrow notes, “If we want to know why a particular outcome emerged, we need to understand how it occurred.”¹³³

Third, I used a combination of participant observation, snowball sampling, and cubist ethnography (so called because it explores a conflict from multiple angles¹³⁴) in three neighborhoods in New York, and six banlieues outside Paris. I conducted my first interviews in New York as an Aaron Diamond Fellow at the Hunter College Center on AIDS, Drugs and Community Health in the mid-1990s. The Parisian research began when I was a Columbia University Fellow at Reid Hall Paris between 2001 and 2002. In New York, I conducted research in Mott Haven, South Bronx, the section of the Lower East Side called Loisaida by Puerto Ricans and Alphabet City by whites, and South Side Williamsburg, and I worked with the following community-based groups: Musica Against Drugs; St. Ann’s Corner of Harm Reduction; Lower East Side Harm Reduction; Coalition for a District Alternative (CODA); the Institute for Latin Studies; the Puerto Rican Defense Fund; the Legal Aid Society; the Puerto Rican Committee for Human Rights; Charas, South Side Action Committee in Williamsburg; the Justice Committee (originally a subsection of the National Congress for Puerto Rican Rights); Copwatch and New York Coalition Against Police Brutality and Stolen Lives. I interviewed former radical party militants from the Real Great Society, the Young Lords, the Puerto Rican Socialist Party, and to a lesser extent the Black Panthers and Nation of Islam, most of whom were founders or activists in the community groups mentioned above. In addition to the New York City Police Department (NYPD), I interviewed members of the National Black Police Association, the Police Executive Reform Foundation, and the former police chief of New Haven. In addition, I interviewed mothers and fathers of those killed by police and activists in anti-police-brutality organizations.

In Paris, I conducted research principally in the northern districts of 95 (Val-d’Oise) and 93 (Seine-Saint-Denis). In Val-d’Oise I worked primarily in Garges-lès-Gonesse, Sarcelles, and Villiers-le-Bel. In Seine-Saint-Denis I worked principally in Aubervilliers, Clichy-sous-Bois, and the Cité des Bosquets (housing project) in Montfermeil. I participated with and conducted participant observations with Veto, Movement immigration et banlieue (MIB), and Groupe de travail banlieue and interviewed members of Indigenes de republique; Movement contre les bavures policières; Sortir du colonialism; and Association, Collectif, Liberte, Égalité, Fraternité, Ensemble et Uni (ACLEFEU). From 2001 to 2005 I worked predominantly in Sarcelles, Garges-les-Gonesse, and Aubervilliers. In 2006 I began working in Clichy-sous-Bois, Montfermeil, and Villier-le-Bel. I conducted my first interview with the French police in 2001 and continued to interview police officers until 2011. I also interviewed families and friends of young people killed by police, and I interviewed one local mayor. I have avoided using the names of those I interviewed except where they have organized or spoken out publicly on the issues. In the case of active-duty police officers, I have taken pains to disguise the neighborhoods and other identifying features as well. All of the interviews conducted after 2001 were conducted in French, although bilingual friends sometimes accompanied me and helped interpret. The translations of all interviews and French writings are my own, except where otherwise noted.

Table 1. New York City neighborhoods

Source: Christopher Hanson-Sanchez, Puerto Rican Specific Data: Institute of Puerto Rican Policy Census Through 1990 (New York: Microdata Supplies, 1995).

Source: Christopher Hanson-Sanchez, Puerto Rican Specific Data: Institute of Puerto Rican Policy Census Through 1990 (New York: Microdata Supplies, 1995).

Source: New York City Health Atlas 1994 (New York: United Hospital Fund, n.d.).

The neighborhoods where I conducted field research in New York were predominantly Puerto Rican, but I argue that the experiences of residents of these neighborhoods is generalizable to residents of other stigmatized minority neighborhoods. First, at the time I conducted field research they were among the poorest neighborhoods in New York. Second, the Puerto Rican experience in New York has been similar to that of African Americans. Both groups have lived in New York City for at least one generation; the largest wave of black and Puerto Rican migrants arrived in the 1950s and early 1960s. Third, and most important for this study, blacks and Puerto Ricans have had similar interactions with police, and most of those killed by the NYPD have been either black or Puerto Rican. As Ramiro Martinez observes,

Table 2. Parisian banlieues, 2009 statistics

Source: Institute National de la statistique et des etudes économiques (INSEE), at the following Web sites: http://www.insee.fr/fr/bases-de-donnees/esl/comparateur.asp?codgeo=com-95585#; http://www.insee.fr/fr/bases-de-donnees/esl/comparateur.asp?codgeo=com-95680; http://www.insee.fr/fr/bases-de-donnees/esl/comparateur.asp?codgeo=com-93014; http://www.insee.fr/fr/bases-de-donnees/esl/comparateur.asp?codgeo=com-93001;http://www.insee.fr/fr/bases-de-donnees/esl/comparateur.asp?codgeo=com-95268; http://www.insee.fr/fr/bases-de-donnees/esl/comparateur.asp?codgeo=cv-9315; http://sig.ville.gouv.fr/zone/3159076; http://www.journaldunet.com/management/ville/.

Scholars have noted that legal cynicism and dissatisfaction with police are both intertwined with levels of neighborhood disadvantage, an effect that trumps racial differences in attitudes towards the police even after controlling for neighborhood violent crime rates. Moreover, ecological characteristics of policing also include the use of physical and deadly force at the city level, officer misconduct in police precincts and slower response times in communities highlighting research that attitudes towards the policing may be a function of a neighborhood context and even determinants in police killings.¹³⁵

My work thus addresses an important lacuna in research on ethnicity, policing, and riots. As Martinez notes, “The scarcity of research on Latinos and policing is one of the most enduring shortcomings in the development of race/ethnicity and the criminal justice system scholarship.”¹³⁶

Moreover the categorical boundaries that New York police use to classify populations locate both blacks and Puerto Ricans on the same side of a racial divide, in diametric opposition to the city’s white populations. Police in France do the same with blacks and Arabs. As Tilly observes, “Durable inequality among categories arises because people who control access to value-producing resources solve pressing organizational problems by means of categorical distinctions. For these reasons, inequalities by race, gender, ethnicity class, age, citizenship, educational level and other apparently contradictory principles of differentiation form through similar social processes and are to an important degree organizationally interchangeable.”¹³⁷

Organization of the Book

In Chapter 1 I look at the construction of racial boundaries in the United States and the violent policing of those boundaries as large waves of black migrants from the South and Puerto Ricans arrived in New York City. I trace the processes that sparked riots in 1935; in 1943; most significantly in 1964, when conflicts in New York initiated a chain of riots throughout the country; and again in 1967, when predominantly Puerto Rican East Harlem and South Bronx burst into flames. I also introduce the three neighborhoods of Mott Haven, South Bronx; Williamsburg, Brooklyn; and the Lower East Side of Manhattan (the section popularly called Alphabet City by Anglos and Loisaida by Latinos), and I track them from their founding to the macroeconomic restructuring of the late 1950s and early 1960s, the race riots of 1964 and 1967, the radical black and Puerto Rican organizing efforts from 1969 to 1973, and the immensely destructive 1977 blackout riots. I conclude this chapter by discussing the 1989 election of David Dinkins, the first African American mayor of New York, and contrasting the dynamic that led to the 1992 Rodney King riot in Los Angeles with that which prevented a small, but similar riot in the Washington Heights section of Manhattan from exploding into a citywide conflagration the same year.

In Chapter 2 I examine the construction of racial boundaries in France and the policing of those boundaries in Paris from the occupation of Algeria and the creation of a French/Muslim racial boundary through the creation of the “North African brigades” and the violent slaughter of unarmed Algerian protesters in 1961 to violent attacks by dominant groups and police on black and Arab youths. This chapter concludes with the 2002 presidential race and the strong first-round near win of the racist National Front candidate Jean Marie Le Pen.

In Chapter 3 I pose the following question: why, when so many of the worst aspects of 1960s race relations remained in New York City between 1990 and 2001—a white mayor, impoverished and still mostly segregated black and Latino neighborhoods, racially coded law-and-order rhetoric, and violent racially biased policing—did neighborhoods not burn? I begin by analyzing the 1993 mayoral race and the way that campaign activated racial boundaries in New York. I then look at boundary activation from the perspective of the police officers whom I interviewed, including those who worked in the three neighborhoods, those who worked in the city at large, those of minority extraction, and those police officers and chiefs active in efforts to promote police reform nationally. From there I shift to neighborhood organizing efforts in the 1980s and 1990s, building on interviews I conducted with black and Puerto Rican activists and residents of Mott Haven, Williamsburg, and the Lower East Side and with leaders of anti-police-brutality organizations (many of whom were radical activists in the early 1970s and continue to live in these three neighborhoods). I conclude this chapter with the stories of four mothers and one father of young people killed by police (Nicholas Heyward, Amadou Diallo, Anthony Rosario, Malcolm Ferguson, and Timur Person) and their efforts to pursue justice for their children and to prevent others from suffering a similar fate.

In Chapter 4 I trace the processes and mechanisms that led Paris to burn for three consecutive weeks in November 2005. I use extensive interviews with French police, with the families of boys killed by police, and with residents of Sarcelles; Garges-lès-Gonesse; Villiers-le-Bel; Aubervilliers, Clichy-sous-Bois; and Cité des Bosquets, Montfermeil, to explore that dynamic. Lastly, I contrast the situation in Parisian banlieues with that in the impoverished neighborhoods of northern Marseille, where riots did not erupt.

In the Conclusion I discuss changes in policing politics in New York and Paris since 2005 and reexamine my main theoretical claims in light of the evidence of the two cases. Finally, I delve into the implications of this analysis for understanding more recent riots in Europe and elsewhere.