Let me begin with two stories from a bygone era. The first comes from the 1960s. Back when we used to build public housing, a group of progressive architects asked prospective residents what kind of apartments they would like. Much to the puzzlement of the architects, the residents wanted "railroad" apartments, with a front room where their daughters could entertain suitors. Then the architects had an idea. They built models of apartments. This time, the residents preferred the models. What happened, of course, is that when first asked, the residents reflected on how they grew up, and did not have a vision of other possibilities. The second story involves a study of unannounced caseworker visits to the homes of welfare recipients. Half the sample had telephones, and the workers would call first to make an appointment. Half did not have telephones, so the workers came unannounced. Since this was an era of benign administration, the group without the phones said that they did not mind the unannounced visits. On the other hand, the group with the phones said that they would mind very much an unannounced visit. Again, the different answers reflected different experiences.

The moral of these stories-which scholars and activists working with dependent people have known for a long time-is that it is not always easy to ascertain what people really want. One has to listen very carefully, to recognize that power also operates indirectly-through agendas and the subtle manipulations of socialization. Which brings me to Meares and Kahan. They are rightly concerned about paternalism, but paternalism takes many forms. One might consider where the respondents are coming from, and why.

Let's make the point less abstract. Meares and Kahan present an appealing case. At least as they present the facts, Chicago Housing Authority (CHA) residents actively participated in a vigorous debate about policing issues. But what happens a year from now? Crime has gone down somewhat, some residents feel more secure, others do not want to relax the rules governing sweeps. We know from housing associations and school committees that attendance at community meetings tends to decline, until a relatively few activists run the show. How would Meares and Kahan (and the courts) decide whether it is appropriate to allow "community burden-sharing" when most of the affected people are not present, debating, and voting? The standard view is that if people do not attend, they are not interested. But as Jane Mansbridge points out in Beyond Adversary Democracy there could be many reasons for non-attendance-some may feel that participation is futile, while others may think that their own interests are being adequately represented. Again, the point is that the test proposed by Meares and Kahan involves subtle and complex factual determinations that are not one-time decisions.

Consider their argument that we do not have to worry as much as we used to about inappropriate discretion because African-Americans have elected positions in local government and are now some proportion of police forces. For someone who (like me) lives in Los Angeles, this claim is truly amazing. Rodney King? Driving While Black? But it is not just this one city. Complaints of law enforcement discrimination against blacks are loud and clear throughout the country. The Meares-Kahan view of political accountability is naive. Blacks and Hispanics do not speak with one voice. And what about other minority groups? When I was growing up in the East, or working in the Midwest, I saw the world in terms of black and white; Los Angeles was an eye-opener. It reminded me that in our country we have a long history of racial hatred across and between many groups. Sadly, intra-group conflict was a major factor in the Los Angeles riots. Most of the violence occurred in transitional neighborhoods. But large public housing projects and other crime-vulnerable neighborhoods are often composed of multi-ethnic groups-a mixture of long-term residents and newly arrived immigrants. One can easily imagine minority groups within larger minority groups feeling isolated and threatened-which, in turn, would substantially complicate our judgments about the accuracy and fairness of community-based decisions regarding sweeps and curfews.

Community empowerment is a laudable goal, and we should work towards it. But it is a subtle and complex process. It's not as simple as calling a town meeting.