It's a brave act for Meares and Kahan to step forward at this instant with these arguments. As this issue of Boston Review was being finalized, major papers were giving daily front-page coverage to an immigrant to New York City named Amadou Diallo, dead from a hail of bullets because he acted furtively when confronted in his foyer by a band of plain-clothed armed men. The story quickly became a more general one about the legitimacy of New York's apparently effective crime prevention program, one based on aggressive order maintenance tactics that push the envelope of lawful police intrusion into the lives of loitering, sauntering, hustling, and "signifying" young men on the street. As protesters tried in vain to penetrate the riot barriers surrounding City Hall, Meares and Kahan were assuring us that proactive order maintenance is fine-as long as the community is ready to bear with it and police are capable of being smart about it.

They think it can work because African-Americans (the only group they really consider) have enough power in enough places to ensure that police will really work on their behalf; that communities can deliberate and decide when their crime problems are so overwhelming that they need to call friendly fire down upon themselves; that society's only other response will be even more repression; and that police on the street will be able to "fine tune" their practices enough that only the bad guys will be seriously discomforted.

Maybe. To their credit, they do not duck considering the hardest practices, those involving stops, searches, and arrests. These lie close to the core of the police function, which involves the legitimate use of violence. But I think their argument would be stronger if they folded their appeal into another policing strategy that is also on the minds of many: community policing. Community policing has vast political appeal, so much that scarcely a police chief, mayor, or city manager in the country wants to be caught without it. A thorough-going community policing program also provides structures and processes that could form a solid underpinning for both general principles of what Meares and Kahan call 1990s style rights: that a community can deliberate and decide what policies are right for its members, and that discretionary order maintenance by police can be responsive to the concerns of the good people of a neighborhood.

One feature of a real community policing program is that neighborhood residents are involved in identifying and prioritizing local crime problems, in a constructive dialogue with one another and with police officers who work in their immediate area. In Chicago this takes place at public meetings that are held monthly in each of the city's 279 small police beats. At these meetings both police and citizens also are to report on what they have accomplished since the last meeting, and what they will work on next. Residents criticize police actions, and since the police who attend work the beat every day and will be back to next month's meeting, they cannot miss the message. A large fraction of these meetings include debates over how police should respond to order-maintenance problems, and Meares and Kahan have it right that residents need and desperately want help from the police in reclaiming their communities. Because the police have been reorganized around the small beats, they develop a great deal of community-specific knowledge that they are able to apply to their problem-solving. Community policing involves decentralizing responsibility as well as authority to these local teams, so that their sergeants become the most important people in the department. That is the level where the buck has to stop if the fine-tuned exercise of discretion is to be the order of the day.

In short, community policing can provide an organizational framework for dialogue and decision-making, a forum for securing month-in, month-out accountability to residents for what beat officers are doing to help them apply local solutions to local problems, and a management structure that gives it a chance to actually work.

Would any of this have helped Amadou Diallo? He was shot by a roving tactical team that would never have recognized him as a local street peddler, and in every department finding ways to involve detectives, the drug squad, and SWATters in community policing has been a struggle. This is where the aggressive young hotshots want to go, and where a lot of problems with the community arise. In Chicago, officers committed to beat work have complained that the cowboy teams fail to keep the faith with the neighborhoods where the beat cops work every day. Washington, D.C., has probably made the most dramatic move toward reining in the SWAT teams, by forcing virtually every unit in the department to decentralize into the city's service areas and get involved in a dialogue with community residents. The jury is still out on whether Washington can make it stick, but this move reveals that police managers know that community policing has to be the entire department's program, and not just something handed off to "wave and smile" units. A fair objection is that the constructive dialogue component of community policing takes place outside the political system. Fewer people get involved than turn out to vote, and nobody elected the decision-makers. But neighborhood safety is not about taxing and spending, it is about ensuring quality of a service. In Chicago's small beats it is too local for a formal political system, and that's part of its popularity. And although recently immigrated Amedou Diallo was not eligible to vote, he could have spoken up at beat meetings; as they say in Chicago, "safe streets are everybody's business."