Eddie Johnson, Chicago’s new police superintendent, has called for a renewal of community policing. He’s held meetings with community activists and promised to make an effort to change the culture of the Chicago Police Department. That lines up well with the emerging national consensus on police reform: localist conservatives and progressives alike are eager to see police officers build relationships in their communities instead of exploiting them. But community policing has hidden costs as well—not just for law enforcement, but for the neighborhoods it’s supposed to benefit the most.

Johnson has also increased his department’s use of a computer-generated heat list that is designed to identify the people most likely to become victims or perpetrators in the city’s ongoing gun violence epidemic. The CPD claims that most people involved in violent crimes this year were on the list; the plan is for police officers to visit at-risk Chicagoans in person in order to warn them that they’re under extra surveillance and offer them social services.

Predictably, the list thrills techno-utopians and outrages civil libertarians. The latter are especially concerned by the system’s opacity; although arrest records are apparently an important data point, the algorithm also appears to incorporate information about social networks into its calculations, so that Chicagoans who have been hanging out with the wrong people may find themselves under surveillance. (Other critics note that the selective enforcement of misdemeanor prohibitions means the list will inevitably be dominated by blacks and Hispanics).

Although they seem to come from different worlds, community policing and algorithmic surveillance are actually closely linked policies, and it’s not surprising that Johnson is embracing both at once as he tries to reform the nation’s second-largest police department.

♦♦♦

The most optimistic view of the heat list conjures up an Andy Griffith vision of the police officer who recognizes that a neighbor has fallen in with the wrong crowd, gives him a stern warning, and offers him help. And by concentrating preemptive suspicion on individuals rather than entire neighborhoods or racial groups, proponents argue, the CPD will be able to abandon racial profiling once and for all.

But if surveillance programs potentially include an element of community policing, community-policing programs unquestionably include an element of surveillance. Chicago has one of the nation’s oldest community policing initiatives, CAPS. Besides encouraging police officers to develop personal relationships with the people on their beats, the program requires them to hold monthly meetings to discuss strategic priorities with locals. But the meetings (and community outreach by individual officers) are primarily intended to make residents sources of information on local crime. And although CAPS is generally regarded as a failure—the mayoral policing taskforce recommended that it be scrapped entirely and replaced with a new program—it has evidently remained a valuable source of information for law enforcement. A report by an activist group that sent observers to a series of CAPS meetings in early 2015 alleges that police officers set meeting agendas unilaterally and that residents mostly contribute to meetings by reporting petty administrative violations:

The officers moderating the event[s] encouraged residents to treat their neighbors with suspicion. Across several sites, officers told attendees they were “eyes and ears” of the CPD and urged residents to report anything that seemed suspicious, including minor crimes like loitering and public consumption of alcohol . . . “They don’t have to be doing anything,” the officer explained. “If you see someone that seems really out of place, call.” But what constitutes “out of place” is subjective, and some residents are considered “out of place” even on their own property. At one meeting, a CAPS attendee boasted that he had “told [a 14-year-old] boy next door that he’s not allowed to sit on [his] front porch anymore. He’s a target.”

The latest push for policing reform could easily become another opportunity for police departments to move on to new, subtler methods of violating civil liberties.

It’s not surprising, then, that new proposals meant to replace or supplement old programs like CAPS are vulnerable to similar criticisms. Consider Polis Station, a police station plan created by the fashionable architecture firm Studio Gang. The scheme converts the police station into a community center and uses other design features to “atomize” officers and “make them part of their communities,” as Jeanne Gang puts it. Studio Gang’s hope is to use design to induce a radical shift in police culture: the police are more likely to treat civilians like neighbors if they are, in fact, their neighbors. But you don’t have to be a hardcore libertarian to think that there’s something dystopian about making the police the catalyst for a revitalized community life.

For example, Gang suggests that a Polis Station could attract locals by offering free Wi-Fi. But if retail stores already use their free Wi-Fi networks to track the sites customers are visiting, it’s hard to imagine that police departments wouldn’t be inclined to do the same on their own networks. Suddenly, a friendly community outreach program becomes a new method of keeping an eye on passersby. (Who knows—in the future, your web history could even contribute to your ranking on the heat list.)

Beyond the surveillance problem, aggressively planting the police station at the center of neighborhood life might help address problems with police culture but hardly seems like a gift to communities that mistrust not only the police, but also a corrupt, dysfunctional city government that can’t fund its pensions or its schools and that oversees a population larger than Jamaica’s. The basic impulse to subsidiarity is worth heeding here: even a miraculously reformed police department oversteps its bounds when it makes an intentional effort to become the primary generative force in the neighborhood.

♦♦♦

“How officers define their role will set the tone for the community,” writes the Presidential Taskforce on 21st-Century Policing, adding that “As Plato wrote, ‘In a republic that honors the core of democracy—the greatest amount of power is given to those called Guardians. Only those with the most impeccable character are chosen to bear the responsibility of protecting the democracy.’” I’ve been unable to locate the quotation in the Platonic corpus, but the taskforce’s apparent misappropriation of a fiercely anti-democratic philosopher is instructive. Guardians in the Republic are drawn from among the noblest citizens and given a philosophical education so they can enforce the laws justly—analogous to the “soft” training that reformists prescribe for police departments. And, like the taskforce, Socrates asserts that the guardians must identify the common good as their own good.

What the guardians definitely don’t do, however, is community policing. They’re separated from their families at birth and live together in a commune as adults. Isolation ensures that they aren’t corrupted by common society. That kind of thinking has been influential in past American police reform movements: in New York City, for instance, police officers aren’t allowed to work in the precincts where they live, to prevent them from facing conflicts of interest.

The Platonic critique of community policing hints at another worry about Polis Station-style proposals. Chicago police stations contain holding cells and interrogation rooms, and the combined community center-police station model evokes the image of police officers frog-marching suspects past children and parents enjoying an afternoon at the community center or kids playing basketball outside. Although that kind of visual dissonance could probably be avoided through attentive design, a certain cognitive dissonance lingers around the idea of placing criminal justice and community-building under the same roof. Station design needs to respect the ways communities compartmentalize different aspects of their common life. There must be ways of securing police accountability that don’t require the essentially violent work of upholding the laws to converge with positive institutions. For Plato, just law enforcement requires a certain separation from communal passions; it might also be the case that building and maintaining robust institutions requires a certain separation from the ugly, violent work of law enforcement.

Plato’s concerns about police integrity are obviously quite different from the civil-libertarian worries I’ve brought up here, but both critiques of community policing suggest that there’s a dialectic at play between intensive engagement by the police in the community and the utter detachment of the police from the community. It’s a delicate balance, because recent events have made it clear that many American government institutions actually need closer ties to the community rather than the opposite. Consider Ferguson’s tyrannical municipal government, which bled its people dry through trumped-up fines and fees.

♦♦♦

American racism, past and present, complicates this dialectic: compared with, say, the educational segregation of Plato’s guardians, de facto racial segregation produces particularly toxic effects. While increasing minority representation in police forces patrolling minority-dominated neighborhoods is an important start, the fact that black officers also exhibit bias against blacks is perhaps the most convincing proof that, despite everything, America’s most exploited communities need some form of institutionalized community policing.

But it’s not a project we should undertake blithely, and concerns about new forms of surveillance—and, in general, about community policing’s threat to the insulation between government and the governed—need to be present alongside familiar worries about misrepresentation and domination. Even the least invasive forms of community policing will inevitably degenerate into surveillance mechanisms as long as we ignore another Platonic dictum: that the worst possible outcome for a city is to divide, according to its natural tendency, into two cities—a city of the rich and a city of the poor, or a city of one race and a city of another.

Malloy Owen is an editorial assistant at The American Conservative.