Instead, the perception of police behavior in a society is often the critical fact. Under Thatcher's rule, after prolonged allegations of police harassment and corruption, there was a national review of British law enforcement. But Brits still are fighting over whether their passive tradition of policing, so-called "policing by consent," is effective for modern crime, or whether they should encourage greater use of force and surveillance. Compared to the United States, the British still lag in convincing ethnic minorities to see cops as legitimate and fair. And both the right and the left have called for more consistent, effective law-enforcement services.
The lack of faith in law enforcement shaped the forms of mass unrest we saw last month in the U.K. In the first 48 hours, rioters displayed wanton disregard of police because there was no expectation of predictable response. By contrast, notice the near absence of looting in Japan after the earthquake—and historically. The Japanese may have a "culture of respect," but they also expect that law-breaking will bring about a predictable response and so they adjust their actions accordingly. Comparisons are hard to make across nations, but in general, when policing is believed to be arbitrary or capricious, those who wish to loot and riot are more likely to find like-minded souls to join them.
3. Riots are spontaneous, chaotic, and hard to influence, whereas peaceful protests are organized, planned in advance, and easily to manipulate.
Riots tend to begin as nonviolent protests and demonstrations. It is less common to see an event like Vancouver's recent violence, in which it seems that the instigators arrived even before the final Stanley Cup game was over to begin committing arson, and more common to see an event like the London riots, in which protestors started as peaceful demonstrations.
Predicting whether a large group becomes riotous is difficult. The logic of crowds is different than the logic of many individual decisions made separately. The sociologist Mark Granovetter argued years ago that crowds are uncommitted masses that require particular sparks to grow rebellious. Different individuals make different cost-benefit analyses, and some are more prone to violence than others. But a group of individuals can be influenced much like a caucus. In some groups, individuals may be deaf to religious pleas, in others, they may respond to race-based calls to action, and in others they may be sensitive to anti-capitalist screeds. You need the right spark for the right crowd. (And views of the police are just one key factor affecting public opinion.)
The idea that individuals who spread information and rumor may effectively incite action in one crowd but not another is supported by recent research on artistic influence by Yahoo! Research's Duncan Watts. By tracking the role that key "influencers" can play in mobilizing group opinion, Watts finds that "average individuals" are sometimes as powerful in shaping collective behavior as charismatic, well-known figures. (He thus concludes that companies who pay celebrities a high price to publicize their product or fashion trends may be throwing their money down the drain.) For riots, this means that the spark can come from unpredictable places, and not only from a powerful blogger, activist, or leader. Anyone could potentially be the instigator—if he or she wants to be.
The implications probably sound a bit depressing: If anyone in a crowd can start a riot, how do we make a strategic intervention? But the converse is also true: Anyone in a crowd can help stop a riot, too. A few discordant pleas from different voices, and no one will know whether to throw the rock or put it down. For this reason, to stop a riot, there are many options, big and small.
Sending in cadres of police certainly can work. But mobilizing hundreds of cops can take days and most countries aren't equipped to carry out coordinated enforcement. In this respect, the United States gets high marks; for over a century, local police forces have excelled in quelling mass unrest by strategically using force, closing off streets, and infiltrating the crowds. By contrast, European cops typically look like deer in the headlights when crowds gather. They are slow to act, they hesitate to use force, and they have poor leadership—Britain sill uses volunteer forces to respond to unrest.
In practice, riots can dissipate without excessive police actions like tear-gas spraying and the use of heavily armed battalions. In the 2005 riots, sociologist Laurent Bonnelli found what I saw in 1990s Chicago: Local religious leaders and outreach workers walked about the streets, sending youth various messages, from warnings that police were arresting their friends at great rates to updates that everyone else had left the streets and gone home. Small doses of confusion can be effective because crowds are highly susceptible to one another's actions. If a few people curtail their violence, their counterparts may feel isolated, vulnerable, and exposed.
So, then, what should we look out for in the United States?
I don't predict a wave of rioting to take over towns and cities, even if the country sinks back into recession. However, there has been a steady growth in large-crowd actions recently. This year, several dozen large "movement"-based protests were organized by unions and advocacy organizations over cutbacks in services, treatment of immigrants, and bargaining rights. Nearly all have been peaceful. But there have been spontaneous gatherings that have precipitated violence. A few weeks ago, "flash mobs" of youth in Philadelphia ransacked stores and harassed pedestrians; in January, Milwaukee suffered a similar outbreak at a shopping mall.