On a warm day in Baltimore last summer, with public school almost out for the year, someone—call her Jane—posts to her neighborhood Facebook page. According to her profile picture, Jane looks to be a white woman in her thirties. She adds four photos of a group of black male teenagers in various formations, mostly looking warily behind them. One carries a hover board. “Are these your kids?” the post begins. Jane says she confronted the teen carrying the hover board. When he wouldn’t tell her where it came from, she called the police. When the police didn’t show, she stayed and took pictures with her phone, following the teens around in her car. 

“Share these pictures,” Jane urges fellow residents of the Harford Road Community Collective, a coalition of racially diverse neighborhoods in Northeast Baltimore. “I’m hoping a parent/guardian sees their child’s face posted on Facebook, and at the very least lets them know that they’re being watched. Everywhere. We don’t play here in 21214.”

It’s not clear what parents would want their children to be surveilled by strangers in cars, but the message was conveyed nonetheless. Another post soon appeared from a man identifying the teen with the hover board as his son and the board as one of three the family owns. He calls the meddling racially motivated harassment. “We also live in the 21214 zip code,” he says.

In an effort to monitor the community outside their doors, people are increasingly turning to their screens.

The incident reflects the way digital neighborhood spaces amplify power dynamics in racially mixed neighborhoods. Five years ago, Jane wouldn’t have had as connected an outlet to voice her suspicions. Without a platform to post photos, she likely wouldn’t have taken them in the first place. But we live in a different era now, the era of digital neighborhoods. Virtual communities—neighborhood groups on Facebook and Nextdoor, a zip-code exclusive social network—make new kinds of connections possible among neighbors. They also embolden a new kind of surveillance. In an effort to monitor the community outside their doors, people are increasingly turning inward, to their screens.

• • •

The rise of new technologies of surveillance is buttressed by two realities. For one, white people generally don’t have meaningful relationships with people of color. Moreover, the past twenty years of gentrification have given white people—like myself—a sense of ownership over the cities we occupy. Baltimore, where I live, has a notoriously high crime rate, but murders occur overwhelmingly in a few poor black areas, while armed robberies dominate more affluent, majority-white areas. Some areas boast the rising property values and shift toward white home ownership associated with gentrification, but Baltimore is one of the few U.S. cities more people are leaving than moving to, and vacant housing remains a huge problem. The city is segregated—it was the first city to legislate residential segregation—but not along the stark geographical lines of cities such as Chicago and Boston. Instead, racial change is visible block by block, and some neighborhoods, such as my own north of Patterson Park in southeast Baltimore, have high racial diversity index scores.

Civic-minded residents often express their desire for neighbors to talk to each other. By this metric, neighborhood Facebook groups would seem to be doing a good job. The Patterson Park group has over 14,000 members and typically sees between 20 and 50 updates a day, and posts often run over 30 comments deep. Five volunteer admins frequently spend significant portions of their days admitting new members, checking with varying degrees of vigilance whether those members actually live in the neighborhood, and ousting those who don’t follow the rules pinned at the top of the page—no memes, no yard sale items or personal marketing, try to take video prior to posting about a crime.

“There is no Central Park in cyberspace.”

Like most neighborhood Facebook groups, Patterson Park’s traffics mostly in the ordinary concerns of urban life: missed trash pickup, lost pets, free park workouts. If you are new to the neighborhood, as I was two years ago, you can learn a lot about its norms: how close to the end of the curb you can park, how to tell the difference between a firework and a gunshot, when neighborhood association meetings are held, how concrete blocks in planters increase drainage and decrease theft, the varieties of pierogis available at the park’s summer Ukrainian Festival. By promoting neighborhood events, the page gets neighbors out of the house. But since it is a more reliable source of information about what’s happening outside your door than actually going outside, the page also incentivizes staying in.

The most robust neighborhood Facebook groups in Baltimore represent racially diverse neighborhoods like mine, as well as homogenous neighborhoods that skew white, like Canton to my south. The neighborhood of Sandtown-Winchester, by contrast—96 percent black and made infamous by Freddie Gray’s death in 2015—has a “like” page rather than a group.

And neighborhood groups for predominantly white populations aren’t just more robust. They also proliferate—as if the available online communities are never enough. A new group will spring up once every few months with the mission of allowing members to say what’s really on their minds. These offshoots often disdain the civility checks in place for the original neighborhood groups, and they typically frame themselves as focused on safety. The description for a group called “Community Safety Patterson Park UNCENSORED,” for example, reads

This group is for those who have been CENSORED or removed by the Patterson Park groups for speaking your mind about crime. If you are sick and tired of all the excuses for the criminals and want to express your feelings and suggestions on how to eradicate the criminal element in our neighborhood, this is the place for you.

Like much of reddit, such groups delight in testing the boundaries of language regulation. They chafe at civility checks, such as repeated urgings not to refer to black teens as “thugs” or “hood rats,” or to abbreviate African American to “AA,” as it’s unspecific (and presumably overly clinical, though I haven’t seen that argument made explicitly). Members bristle at what they see as a culture of PC censorship or a violation of a central imperative not to “blame the victim,” as Gregory Krumm, a commenter who frequently complains about language policing, told me. The Harford Road Community Collective recently felt compelled to add the following to its guidelines: “When posting about an incident, please do not profile. Descriptions focusing exclusively on race, ethnicity, sexual orientation, or religion will not be tolerated.” But pages like yet another for Patterson Park proudly declare, “This group is open to any and all opinions, regardless how ridiculous or offensive.”

Like much of reddit, neighborhood Facebook groups delight in testing the boundaries of language regulation.

In truth, safety-related posts make up a large percentage of those allegedly censorious groups. Rarely does a day go by without a resident asking for tips on how to keep safe (“Do thugs generally avoid targeting people with dogs?”), reporting a robbery they were witness to or victim of, posting pictures of discarded Amazon packages taken off stoops, or lamenting the decline of safety in the neighborhood (“Believe it or not…there was a day when it was safe to walk through the park in the middle of the night”). And security infrastructures are increasingly capitalizing on this focus on crime. Specially designated moderators relay information from meetings with district officials. Nextdoor has official partnerships with many city police departments, including Baltimore’s. Several Facebook pages purporting to focus on neighborhood security in Baltimore are simply fronts for selling home alarm systems. Some of these pages, such as the “Hampden, Medfield, Woodberry, Wyman Park Crime Watch,” operate out of other states entirely, transparently monetizing Baltimore residents’ fear of crime. Many home security cameras even enable seamless sharing of received images to social media.

• • •

Zeke Cohen, my district’s councilman, understands how the language used in online neighborhood groups affects perceptions of the community. “A crime will be committed, and some people will say, ‘An AA male broke into my house.’ All of a sudden you’ve got what are largely segregated neighborhoods looking for AA males,” Cohen told me.

Cohen’s political engagement often involves the neighborhood Facebook pages in his district. He regularly publicizes neighborhood initiatives and posts inspiring stories of community members’ success, such as businesses started by immigrant refugees. “It is a great venue for events,” Cohen says of Facebook. “It can be a powerful tool for advocacy on a given issue.” It’s hard not to see Cohen’s relentlessly positive posts as correctives to the many “suspicious man” posts and denunciations of crime in the neighborhood. Cohen’s unwillingness to engage with them makes him a frequent target of the neighborhood commentariat. Residents—mostly older white men—often refer to him derisively in the wake of a recent crime. “Zeke Cohen and his cookies and milk solutions are part of the problem in this city. These punks, thugs, animals, etc. etc. etc. need to be dealt with,” goes one comment.

Cohen is well aware of the less savory aspects of the neighborhood Facebook groups he frequents. “It can amplify peoples’ racism,” he says. “It can create a relatively safe space for people to be racist, sexist, misogynistic, homophobic, and everything else. People will say things to me when they’re upset that they probably wouldn’t say in real life, or at least they will temper. It becomes this kind of id.”

As a community advocate who has served my neighborhood in various capacities for almost a decade, Kim Wiggins regularly engages with the id. She used to moderate the Patterson Park Facebook group; when I contacted her, it turned out that she had recently quit “for a better quality of life,” but she agreed to debrief with me. We met in the parking lot of our local Chipotle. Wiggins, who pulled up in an open-top convertible overflowing with shrubs, said that her civic activism started with another greening project: getting a tree for the front of her property. She would show up at neighborhood association meetings trying to get this tree, and before she knew it she was vice president of one of the neighborhood sectors. Then she started moderating the neighborhood Facebook group.

In the late-capitalist sharing economy, social network companies profit from decreasing the gap between thought and (online) action.

It was here that Wiggins, who is black, discovered the fixation of her white neighbors on crime. She began to notice a disparity in the way crimes were reported: posters used a great deal of detail to describe white people suspected of crimes, but hardly any detail to describe black people. Wiggins imitated someone posting about a power drill taken by a white suspect: “‘White guy with long blond hair, green floppy hat, white shirt, denim jeans, and tennis shoes going door to door trying to sell a power drill.’” For black people, Wiggins said, postings are much less specific. “Black description: ‘two black kids on bikes.’” 

What went assumed in Wiggin’s account was the race of the person doing the posting. While my (admittedly statistically unrigorous) analysis puts the Patterson Park Neighbors group at about a third black and half white, a sampling of the most recent fifty crime-related posts showed only five black posters and forty-two white ones. Members talking about crime in any context also tend to assume, by default, a white victim. If white residents are the ones talking about crime, and if criticizing people for talking about crime is “blaming the victim,” as many in the marginal groups would have it, then by a trick of metonymy all white residents start to look like victims.

Wiggins thinks that part of the reason white residents don’t tend to be good at identifying black people is that, even in racially diverse neighborhoods, white people don’t often relate meaningfully to people of color. “They don’t think there’s lots of black kids [living in the neighborhood],” she says, “because they don’t know any black kids because they’re not doing anything to know this population.”

Posters also often assume that criminals come from outside the community. “Honestly, the problem is not with the people who live in the area, it is the incursion of thugs on bikes,” writes a commenter in typical post-crime feedback. In some ways, this assumption reflects faith in the moral goodness of the community; in other ways, it allows residents to write off people committing crimes as “not one of us,” and thereby less worthy of attention (through rehabilitation, focus on underlying social problems, or policy solutions).

Wiggins believes that there are negative consequences to the fabric of the community when black residents are seen as indistinguishable threats and outsiders, a perception reinforced when black community members go on neighborhood Facebook pages: “There’s a psychological thing that happens when you see your neighborhood page, and everyone you think is white thinks everyone is black is bad.” Wiggins thinks that more black residents would either join or become active on the neighborhood Facebook page if white residents changed their rhetoric.

Online communities abet incivility by minimizing the barrier between instinct and sharing. In the late-capitalist sharing economy, social network companies profit from decreasing the gap between thought and (online) action. But that gap serves an important purpose in communities on the ground. It encourages you to be courteous to your neighbors and accountable in your interactions.

Neighborhood Facebook group admins try to replicate these civility checks, but there’s only so much they can do to combat the platform’s frictionless ease. For its part, Nextdoor has moved in the direction of implementing more such tools. Responding to accusations that the Oakland, California, group fostered a racist environment, the company introduced a questionnaire that asked users to say whether they witnessed an actual crime or just saw something suspicious. Nextdoor keeps track of started-but-unsubmitted posts; after the questionnaire, the number of such posts increased by 50 percent, suggesting users had not witnessed an actual crime.

Kim Wiggins often attempts to get neighbors to talk face to face. When hostile neighborhood group members message her saying she’s racist, she invites them to brunch. “And the racist people say, ‘I’m not going to come and hang out with your racist friends,’” Wiggins recounts. “And I’m like, ‘There’ll be bacon and eggs!’” But despite her best efforts to encourage face-to-face friction in the neighborhood id, Wiggins says the brunches end up comprising mostly self-selecting, well-intentioned white progressives who want to learn more about how to better their communities. The trolls stay behind their screens.

• • •

Public space in the city isn’t what it used to be. As Antero Pietila, author of a book on the history of Baltimore housing called Not in My Neighborhood (2010), told me in an email, the old infrastructure of public space—“churches and synagogues, ‘improvement associations’ and fraternal organizations”—is gone or faded past recognition. Pietila recounts mid-twentieth-century Baltimore as having “a totally different synergy between residents and merchants because they knew one another,” where people could be found “strolling around [the] neighborhood or sitting on porches, easy to reach for a chat.”

In some ways, neighborhood Facebook groups have the potential to reinstate a different version of this utopian commons. If you have Internet—not to be taken for granted in many poor urban areas—you have the tools to enter a community. You don’t have to make time to attend a neighborhood association meeting. Instead, there’s a fully formed neighborhood awaiting your engagement online. 

But there was a dark side to the mid-century urban commonwealth, and, as public space has begun to migrate online, so has its unsavory parts. Those white neighbors shooting the breeze on their marble stoops? They were bonding over their shared discomfort with the black neighborhood a quarter of a mile west. After the Supreme Court deemed residential segregation illegal in 1917, Pietila recounts in his book, Baltimore’s white residents turned to restrictive neighborhood covenants to ensure African Americans remained in prescribed areas of the city. In other words, there was no true utopian New Urbanist ideal, or at least not one that didn’t constitute itself through the exclusion of others.

We must ask of online communities the same questions we ask of brick-and-mortar ones.

Neighborhood Facebook groups occupy a middle ground between private and public. On the one hand, they are closed groups that are supposed to be exclusive to the residents of a single geographic area. On the other hand, they are supposed to be open to all the residents of that area, not just ones who share the background or interests of the majority. Users expect that online neighborhood groups mirror the demographics of real-life neighborhoods; that’s why, for example, some residents are concerned that the primarily white admins on the Patterson Park Facebook page don’t reflect the racial makeup of the actual neighborhood. We must ask of online communities the same questions that we ask of brick-and-mortar ones: Who feels comfortable in these neighborhoods? Who dominates, and who gets sidelined? Who has access, and who is excluded?

For Margaret Kohn, a scholar who works at the intersection of political theory and urban space, the ideal public space must meet three criteria. It should be publicly owned, accessible to everyone, and allow for intersubjectivity (rather than a spectator-spectacle relationship). Kohn acknowledges that few spaces conceived as public perfectly fulfill these criteria. Still, they are useful metrics for assessing how public—and by extension, how democratic—communities truly are.

Since both Facebook and Nextdoor are privately owned, they already fail to meet the first criterion. As Kohn notes, “There is no Central Park in cyberspace.” Users must sign up for a Facebook account or verify their zip code through Nextdoor, an authentication process that excludes those with temporary or unstable housing. And access to online neighborhood groups is, of course, contingent on access to the Internet. In 2013, around 30 percent of Baltimore households lacked Internet access. For many of those who do have access, the Internet is still not a meaningful part of their lives. This includes the elderly, primarily Spanish speaking residents (bilingual Zumba announcements notwithstanding), and young people who aren’t eligible for a Facebook account. 

Digital neighborhoods fail on accessibility, as well. To Kohn, accessible means more than the mere ability to participate. As she concludes from an analysis of the American mall, “apparent inclusivity can be based on subtle or invisible forms of exclusion” such as aesthetic status markers, implicit norms, or physical space barriers like seats that keep you from lingering. In online neighborhood groups, too, there are ways the dominate ethos is expressed while maintaining the illusion of free access for all. These expressions can be as innocuous as complaints about loud music or reminders that the historical statues in the park are not meant to be played on by children. Or they can be as explicit as saying that all undocumented immigrants in the community should be deported.

What about intersubjectivity? People are using these online spaces to talk to each other, even if many arguments are not productive. In most cases, users are interacting with their neighbors more substantively than they would without Facebook or Nextdoor. Viewed this way, online neighborhood groups represent an unprecedented opportunity. But anyone who has spent time in these online communities becomes quickly disillusioned with the potential of the public sphere. Geographic proximity betrays a startling sociocultural remoteness. Like most Facebook “debates,” they serve only to reinforce already calcified positions or cultural divisions between groups. 

In the afterword to Brave New Neighborhoods: The Privatization of Public Space (2004)Kohn hazards a few conjectures on the under-charted terrain of public space in cyberspace. Drawing on Adam Smith’s idea of the imagined “impartial spectator” who helps us see our biases, Kohn suggests that diverse online communities can help us productively shape and regulate our self-presentations. But Kohn cautions that this value is contingent on the diversity of the online community: “This moderating influence disappears when the community from which I conjure my impartial spectator is made up of others who are similarly obsessed with a particular experience.”

As more and more aspects of our neighborhood lives go digital, we must continue to scrutinize this politics of spectatorship—who watches, and who is watched. One summer Friday night last year I stood outside of my rented row home as a group of black teenagers walked by. “What are you, neighborhood watch?” one of them said. Maybe I was.