Stop and Frisk Online: Theorizing Everyday Racism in Digital Policing in the Use of Social Media for Identification of Criminal Conduct and Associations

Abstract

Police are increasingly monitoring social media to build evidence for criminal indictments. In 2014, 103 alleged gang members residing in public housing in Harlem, New York, were arrested in what has been called “the largest gang bust in history.” The arrests came after the New York Police Department (NYPD) spent 4 years monitoring the social media communication of these suspected gang members. In this article, we explore the implications of using social media for the identification of criminal activity. We describe everyday racism in digital policing as a burgeoning conceptual framework for understanding racialized social media surveillance by law enforcement. We discuss implications for law enforcement agencies utilizing social media data for intelligence and evidence in criminal cases.

In April 2015, Rose Hackman of the Guardian penned an article that asked, “Is online surveillance the new stop and frisk?” We have seen this played out in the intense scrutiny over the New York Police Department’s (NYPD) stop-and-frisk policies where pedestrians, most often Black and Latino, are stopped, questioned, and frisked for weapons and other forms of contraband. Hackman examined new online practices where police departments create exhaustive gang or “clique” lists, and the creation of online identities are used to monitor large swaths of communities of color expected to engage in violent and criminal behavior. Hackman’s argument is situated in historical and controversial debates of racial bias in policing practices in the United States.

In 2013, the United States District Court for the Southern District of New York ruled that stop-and-frisk practices were unconstitutional and directed the police to adopt new policies, which required more justification and documentation if a pedestrian is stopped. Concomitantly, however, the “datafication” of the criminal justice system in New York City occurred as the NYPD increased its resources and interest in social media surveillance of potential violent and criminal perpetrators. In an era of smartphones, status updates, and the sharing and “liking” of the most mundane and the most significant of our everyday experiences, social media platforms shape networking. Social media facilitates a process by which our public and private lives become integrated, a concept now known as “context collapse” (boyd, 2002; Marwick, 2011). Information that was once unique to one sphere can now be accessed in another. The converging of life domains is further complicated when police scrutiny moves into one’s domestic or private life (Trottier, 2012 a).

In this article, we advance implicit bias research within the context of social media policing. We propose everyday racism in social media policing as an emerging conceptual framework to theorize how, in an era of big data in criminal justice practice, social media policing may adversely affect communities of color. In the following sections, we examine how urban policing has moved online, the racial implications of digital policing, we consider how race affects social media policing through the juxtaposition of social media use during the 2014 arrest of Black youth in New York City public housing. We then pose an emerging conceptual framework for examining implicit bias in digital policing, utilizing the post hoc social media monitoring in the case of Dylan Ruff, the perpetrator of the Charleston church shooting in 2015 that killed nine people including State Senator Clementa C. Pinckney. We conclude with a discussion of how implicit bias in digital policing as a conceptual framework may support new empirical studies and challenge current social media policing policies and practices that lack a firm consideration of how race impacts how online data are interpreted and used in criminal cases.

Social Media Use in Identifying and Investigating Criminal Conduct

Proactive policing strategies and multiple forms of police misconduct and violence have disproportionately affected people of color who live in urban, disadvantaged communities (Brunson & Miller, 2006). Justification of this behavior is rooted in historical narratives and a belief structure, often perpetuated in the criminal justice system that views Black men as “symbolic assailants” (Anderson, 1990; Bridges & Steen, 1998; Brunson & Miller, 2006; Holmes, 2000; Jackson, 1997; Jacobs & O’Brien, 1998; Kennedy, 1997; Smith & Holmes, 2003; Quillian & Pager, 2001). Feagin (1991) warns that these covert and routine strategies have widespread, cumulative effects on both individuals and the collective consciousness of Black communities. Writing through the recollections of his interview subjects, he cites the experiences of his interlocutors to make the point that

the cumulative impact of racial discrimination accounts for the special way that Blacks have of looking at and evaluating interracial incidents [ . . . ] What many whites see as black “paranoia” is simply a realistic sensitivity to White–Black interaction created and constantly reinforced by the two types of cumulative discrimination. (Feagin, 1991, p. 115)

Given this frame then, the use of social media to predict criminal activity is problematic. Are police culturally equipped to understand the nuanced linguistic styles of youth communicating on social media? Sociologists suggest that many youth and young adults who live in violent neighborhoods may project a tough image or follow a “code of the street” in their community to stay safe and protected (Anderson, 1999). Could the same be true online? Does an individual who talks about drugs and violence on social media actually engage in these activities? According to Jeff Lane (2016), online is the new digital street on which the code of the street meets the concept of networked publics, or the space and community that is “ . . . restructured. by networked technologies” (boyd, 2008). The questions posed above are urgent as “Police and prosecutors see the street online and offline. [ . . . ] It is on the terms of the digital street that justice turns” (Lane, 2016).

We start from a standpoint that “ethnic prejudices and ideologies are not innate, and do not develop spontaneously [ . . . ] they are acquired and learned [ . . . ] through text and talk” (Van Dijk, 1993, p. 146) and that embedded in the interpretation of online discourse central to social media policing is who’s talking to whom and about what. Traditional criminal profiling relies on the correlation between behavioral factors and the past experience of law enforcement in discovering criminal behavior associated with those factors; thus, profiling rests on the perceived accuracy of the profile as a predictor of criminality (Carter, 2014). Proponents argue that if crime is greater among certain races, racial profiling makes sense (Carter, 2014). However, opponents show that there does not exist disproportionate rates of criminal activity in minority communities (and certainly not rates proportionate with the amount/intensity of aggressive policing of said communities) and point to societal costs of the practice (Carter, 2014). Carter argues that excessive and undue policing of Blacks is a contemporary analog of slavery, an extension of the long-held myth of Black criminality, which served, and ostensibly still serves, to justify and reify the inhumane treatment of Blacks (Carter, 2014). African Americans are disproportionately subjected to law enforcement attention, and thus, disproportionately prosecuted, convicted, sentenced to jail, disenfranchised from voting, and increasingly removed from the mainstream world of jobs, families, and community involvement (Carter, 2014).

Police agencies use social media for a variety of reasons. Data from International Association of Chiefs of Police (IACP) suggests that 96.4% of police agencies use social media in some capacity (IACP, 2015). The most common use of social media is for criminal investigations (e.g., a police officer finds evidence or additional information about a missing or wanted person, gang participation, or web-based crimes such as cyberstalking or cyberbullying). The police may perform a manual search on a social media platform or request specific information from social media platforms. In some instances, the police may also install software on a targeted individual’s computer and perform analysis on social media data that have been captured (Trottier & Fuch, 2014). Social media is also used for community outreach and information sharing. Online policing practices are formed from a multiplicity of technologies available to various agencies. These technologies provide a wide spectrum of access to social media content with differing relation between the police, social media users, and those who manage the platforms. One’s visibility on social media and communication of everyday experiences, practices, and activities provides the perfect platform for covert criminal surveillance by the police. To that end, social media policing reduces transparency and increases the power imbalance in police work as it further underscores the unbalanced “ . . . asymmetrical relations of visibility . . . ” between police and the public (Trottier, 2012b, p. 234).

Daniel Trottier argues that the use of these strategies—what he terms “social media policing”—may represent a new paradigm for profiling and preemptive policing (Trottier, 2012b). With this approach, one predicated on the resistant readings demonstrated in our model, social media and its products are recontextualized in terms of its utility for crime reduction, investigatory purposes, and production of evidence—all uses that are well outside the intended purview (Trottier, 2012b). Moreover, while room for these uses is regularly included in Terms of Service agreements, investigations by law enforcement officials on social media violate users reported privacy expectations (Trottier, 2012a). Social policing requires that social media users monitor their own output and that of their network, as close network connections and relationships may become liabilities if output is deemed problematic by law enforcement (Trottier, 2012a). All told, as stressed by Trottier, the arrangement is an asymmetrical one—the very platforms that amplify user visibility, allow police to act invisibly—and a counter to the rhetoric of “ . . . mutual transparency . . . ” (Trottier, 2012b).

The IACP maintains a Center for Social Media in partnership with the Bureau of Justice Assistance, Office of Justice Programs, and the US Department of Justice, in an effort “to build the capacity of law enforcement to use social media to prevent and solve crimes, strengthen police-community relations, and enhance services” (Orin, 2011). The site also provides access to a survey on law enforcement’s use of social media and a primer for such agencies’ establishing a social media presence.

The IACP’s stated social media mission is echoed on the websites of the technology vendors whose products and services they employ. The website of Geofeedia, a location-based social media monitoring service, speaks of the importance of geo-location in the monitoring of social media to pinpoint hotspots and to anticipate trouble. It also features testimonials from law enforcement agencies across the country. These testimonials, however, must be read within the context the discursive relationship between broader society and law enforcement. This relationship, as we have demonstrated, excludes communities of color and by so doing turns the technological gaze on them. It can be reasonably argued then that technology such as Geofeedia that affords its clients the opportunity to “Discover trends and patterns within the world’s largest set of location-based social data to inform better decision-making” (geofeedia.com) inherently allow for racist practices as the parameters they employ are user defined and not response driven. If communities of color are socially constructed as problematic sites, then this is where the technological gaze goes, in anticipation of a problem—the social controls morphing into punitive cognitive controls.

David Lyon emphasizes that the role of computers in electronic surveillance is an intermediary for human-on-human observation—as in storing and processing massive troves of personal data (Lyon, 1994, 2003). Increasingly, material derived from social media is being used in police investigations (Trottier, 2012b). Traditional surveillance—such as a guard tower or closed circuit television—involves a comprehensive view from above (Trottier, 2012b), whereas surveillance via social media offers a synthesis of views from both above (Trottier, 2012b). Thus, social media offers more ways of seeing more people, due in large part to the saturation of social media and its associated practices in people’s everyday lives around the world (Trottier, 2012a). This enables greater and more robust social surveillance, which is defined by Alice Marwick as the “ . . . ongoing eavesdropping, investigation, gossip, and inquiry that constitutes information gathering by people about their peers.” This process, amplified by ubiquitous social media exposure, allows users to collect “ . . . information about their friends and acquaintances . . . ” (Marwick, 2012). For Marwick, social surveillance is distinct from other surveillance types in three different areas—power, hierarchy, and reciprocity—assuming distributed, equitable, and reciprocal relations in each (Marwick, 2012).

Racial Profiling in Crime Control Technologies

Current concern with the federal surveillance of social protests on social media, particularly that of Twitter’s #blacklivesmatter (Joseph, 2015), has coalesced the topics of the Fourth Amendment, civil liberties in a broad sense, and technology into conversation around the scale, scope, and cost of civilian surveillance. Documents made available by the Department of Homeland Security (DHS) in response to Freedom of Information Act (FOIA) requests underline the nature of this surveillance. Using events surrounding the fatal police shooting of Michael Brown in Ferguson, Missouri, in 2014, as a trigger, the DHS has been collecting information from social media sites in what it terms as an effort to provide “situational awareness” to law enforcement agencies. When paired with the problematic constructions of race in the United States, these actions and legislations erode the personal security of people of color; on the other hand, various lenses of surveillance can also have an empowering effect on their existence.

Undeniably, however, the use of technology to highlight the questionable and often violent actions of law enforcement has brought national prominence to the issue.

There is no doubt that situational awareness through information gathering is a vital process to safeguard against issues of national security and manage specific threats. However, focusing information gathering on mundane, non-violent social protests has made “situational awareness” a euphemism for what Baher Azmy [xix] terms as a way to both watch and intimidate individuals, as well as “an effective way to chill protest movements” (Joseph, 2015). Citing the United States Supreme Court’s 2012 decision in United States v. Jones, in which GPS surveillance was found to be warrantless and a violation of Fourth Amendment protections, Bankston and Soltani (2014) have argued that “if the new tracking technique is an order of magnitude less extensive than the previous technique, the technique violates expectations of privacy and runs afoul of the Fourth Amendment.” In response to shifting technological and social mores, this equilibrium adjustment, essentially, prescribes the entrenchment of Fourth Amendment rights (Ohm, 2012). Rather than forcing individuals to sacrifice their liberties as technology expands the breadth of law enforcement’s capabilities, new limits should be placed on the scope and accountability of such surveillance. For example, the NYPD’s current practice of social media monitoring has been described as a virtual stop-and-frisk (Hackman, 2015). Although the NYPD is casting a wider net, these probes are largely informed by prejudices of color, religion, and socioeconomic status (Hackman, 2015). In fact, 48% of all evidence found in New York City indictment documents was attributed to social media activity or communication (Lane & Ramirez, 2016).

Illustrative Cases

First, the racial nature of social media policing is implicit in Commissioner Raymond Kelly’s comment on the efficacy of Operation Crew Cut, the codename of the NYPD’s initiative to monitor gang members’ social media presence. Commissioner Raymond Kelly claims that “strategic enforcement and proactive policing combined with strong prosecutorial partnerships, including attention to the new battleground of social media, have resulted in lives being saved in New York City, mostly young minority men” (New York Police Department, 2013). On the surface, this is a commendable achievement, but the NYPD’s rhetoric omits an important factor: this success is both the result of surveilling communities of color and using social media links and interactions as the basis for drawing conclusion on the existence of real world relationships between members of these communities. This latter point is particularly troubling, as it has led to the indiscriminate criminalization of many young men of color who simply communicate or sympathize with young people on these watch lists. Additionally, Operation Crew Cut, which has resulted in the NYPD quadrupling its gang division (from 150 officers to 600 officers), came into effect when shootings and homicides were “lower than at any time in the four decades since crime statistics have been maintained” (Howell, 2015).

Commissioner Raymond Kelly also stated that, at Operation Crew Cut inception, issues in the City were not due to “established gangs such as the Bloods and Crips”; rather, it was due to “looser associations” of individuals who were organized based on neighborhoods and blocks (Parascandola, 2012). The influx of police resources did nothing to stunt the crime in the City (Howell, 2015), as the NYPD was already actively collaborating with the District Attorneys and federal prosecutors to learn more about these crews (Parascandola, 2012). Thus, the notion that increased surveillance was necessary operates merely as a façade for police enforcement to erode the rights of swaths of minorities. Such programmatic implementations are often described as creating a sense of “moral panic,” wherein police officials commercialize supposed rises in gang-related violence to both obtain new resources and, more importantly, to pass legislation that would otherwise be impassable (Howell, 2015). In examining the statistics of the ethnicities categorized in the NYPD’s Gang Database, as of August 2013, a mere 1% of the 20,000 individuals were Caucasian or White (Goldstein & Goodman, 2013). These numbers do not adequately reflect the commonly accepted theory in social science scholarship that it is common for Caucasians or Whites to account for a statistically large portion of gang members in the United States. Despite a turnover in leadership, Mayor Bill de Blasio has maintained Operation Crew Cut’s prominence and force (Goldstein & Goodman, 2013). Thus, the real question concerns what, if any, limitations are placed on the NYPD and other enforcement personnel in the United States? How far is too far in an age where “national security” is heralded as a license to, in a sense, broadly trample the rights of minorities?

Second, in Raza v. City of New York, a June 2013 action filed in the US District Court for the Eastern District of New York challenged the constitutionality of the NYPD’s discriminatory surveillance of Muslim individuals and communities in the City (American Civil Liberties Union [ACLU], 2016). The plaintiff claimed that the surveillance of Muslims and Muslim institutions by the NYPD has placed an unwarranted suspicion on all people of the faith: one that other religious groups have yet to face. The scope of this surveillance, as outlined in the suit, seems to highlight an inherent cultural distrust that it then reinforces. More specifically, the American Civil Liberties Union (ACLU) claims that such actions are violations of “the 14th Amendment’s Equal Protection Clause, the First Amendment’s right to the free exercise of religion and guarantee of neutrality toward all religions, and the New York State Constitution” (ACLU, 2016). Further troubling is that any validation of the NYPD’s conduct provides a “dangerous justification” for other enforcement techniques that operate under such narrow-minded, unjust constraints (Chowdhury, 2015).

The NYPD’s actions are part of a broader issue of unaccountable surveillance, as their actions were largely supportive of the Central Intelligence Agency (CIA)—an organization which faces minimal oversight restrictions (Apuzzo & Goldman, 2011). While the United States Supreme Court has recognized various exceptions to searches in the absence of a warrant, such exceptions are granted if there exists reasonable suspicion (e.g., Terry stops). However, merely exercising searches based on an individual’s race or religion does not fall-in-line with precedent. Such tactics provide a blurring of the lines “between ethnic and religious profiling and the Fourth Amendment’s ‘reasonability’ requirement” (Chowdhury, 2015). The act of categorizing centers of worship and focusing searches on solely large sects of Muslims “initiates a debate on ethnic and religious profiling, which Courts have found closely parallels the complex discussion on racial profiling” (Chowdhury, 2015). In light of precedent, such as Korematsu v. United States, where the United States Supreme Court upheld that “public necessity” validated curtailing the rights of the Japanese (which, in conjunction with Plessy v. Ferguson, another race-related case, are viewed as some of the Court’s most heinous precedent), law enforcement and other personnel must be careful to ensure that, as a society, we do not permit such injustices to occur. Sadly, there does not appear to be a reduction in law enforcement’s use of such questionable tactics. The possibility of regressing to a Korematsu-era mentality is shockingly evident in the September 2015 arrest in Texas of 14-year-old Ahmed Mohammed for bringing a homemade clock to school. Such an arrest is but the most recent national example of the bias that Raza addresses (Fernandez & Hauser, 2015). Thus, despite decades of advancement, is the American government embracing and mirroring a mentality propagated during World War II? Even if the NYPD’s program is found not to be on the same-level of such an extreme, the actions of law enforcement undeniably perpetuate an “extremely intrusive and comprehensive system of documenting and watching an entire minority community based on nothing but their religious affiliation” (Chowdhury, 2015).

Third, the story of Jelani Henry is another demonstration of racial profiling in digital policing (Popper, 2014). Henry, a young Black man from Harlem, found himself incarcerated in Rikers Island prison for 19 months—including 9 months in solitary confinement—due to social media associations that, in the eyes of New York City law enforcement, made him a criminal affiliate. Henry became a person of interest due to his Facebook activity and his brother’s gang involvement (Broussard, 2015). Jelani was arrested in 2012 at his girlfriend’s apartment in the Bronx for two counts of attempted murder. Just weeks before his arrest, a young woman was spat on by members of a local Harlem crew and the following day, individuals from the opposing crew shot the spitter and others to retaliate. This entire scene took place just two blocks from Jelani’s home. At the police station, Jelani was questioned about his connections to the Goodfella’s gang. He stated that individuals in that gang were his friends, but he did not participate in their activities. Although Jelani had never been convicted of a crime, the District Attorney labeled him a known member of a violent gang and pointed toward his pictures and “likes” of Facebook posts from Goodfella’s gang members. Charged with two counts of attempted murder along with other lesser charges, Henry faced up to 40 years in prison in a trial that never came. When offered a plea deal, Henry refused and the Manhattan District Attorney’s office later dismissed the charges without explanation. While the charges were ultimately dismissed, Henry and his family failed to receive any compensation for the wrongful incarceration (Gendreau, 2016). In examining Henry’s case in light of the way gang-involved youth in Chicago use social media, it is evident that law enforcement struggles to understand the various contexts of online communication, and instead assume deviant behavior—an example of “conxt collapse” (Patton et al., 2016). Rather than relying on the binary logic of databases (Broussard, 2015), police efforts would result in less discrimination if databases were used “to find a clearly identifiable suspect” (Behrman, 2015). Before proceeding based on a database “hit,” officers should be required to produce information that is more specific, determine a motive, or “look for other inculpating information” (Behrman, 2015). Current efforts, such as Henry’s case, are representative of law enforcement’s too frequent use of generalizations and over-intrusive tactics to surveil minorities.

Researchers have argued that the threat of being surveilled plays as significant a role as actual surveillance in affecting the way we behave (Bentham, 1995; Foucault, 1995).Flashing a gang sign on Facebook may be a young man’s way of joking with friends, proclaiming solidarity or neighborhood affiliation, or harmless posturing. An inherent danger in social media interactions, however, is the issue of the aforementioned context collapse where, in the absence of the cues that frame face-to-face interactions (e.g., physical expression, tone of voice, or conversational context), messages may be misconstrued. In its use of the social media activities of young men of color as the basis for conspiracy charges (O’Connor, 2013), the NYPD has been shown to be treading heavily on their civil liberties (Rakia, 2015). Whether such individuals are convicted or merely arrested in suspicion of criminal conduct, it is highly difficult to facilitate a process of rehabilitation. The often unwarranted and unredeemable broad label of “criminal” is a stigma that is hard, if not nearly impossible, to shed. This has a particularly profound effect on certain minorities who are already faced with difficult socioeconomic stressors and racial stratification. Thus, victims of unwarranted police probing are faced with devastating short- and long-term issues in a system that provides little support. These factors coalesce, propagating a troubling mechanism for suppressing the rights, and arguably future, for many minorities.

Theorizing Everyday Racism in Social Media Policing

The encoding of social media communication happens when a post is written and the interpretations of said post are to be understood by members of the intended audience.

The decoding of a social media post is how an audience member is able to understand and interpret it. Race discourse theory presents a mechanism that can elucidate interpretations of the unique and complex forms of social media communication of young people of color, unearthing manifestations of everyday racism that inform how surveillance is experienced online and by whom. Our models work through a conceptual framework that highlights the impact of resistant and reductive readings (Hall, 1973/1980), of the social media messages of communities of color. The models show how race discourse theory is deployed to the detriment of these communities. The solid black lines illustrate how the information flow facilitated by social media networks and encoded for a specific audience are at the root of the detrimental discursive relationship between broader society and law enforcement. In most instances, this relationship triggers the decoding of these messages as criminal and result in punitive action being taken on the communities of color. In rare occasions, the messages are interpreted as intended, leading to the social acceptance for the community by broader society and access to social capital. In most instances, the resistant readings lead to a self-fulfilling forecast in which members of the community, frustrated by the punitive measures levied against them as a result of the receiver’s biases toward individuals and communities of color, reclaim their agency through deviant performances of self.

Jelani Henry

In the case of Jelani Henry, the police used Facebook photos, YouTube videos, and Myspace threats as evidence against Jelani and his brother. Figure 1 below illustrates that theorizes the pathways through which the encoding process is interrupted through the text and talk of broader online public discourse. Central to understanding the model’s application is cognizance of the discursive relationship between broader society and law enforcement. Broader society here refers to non-hyphenated members of society whose ethnic descriptions match the ideal of belonging. This discursive relationship and the tropes and stereotypes it promulgates are fundamental to appreciating how Henry’s performance of self, encoded for his social network, would have met some type of cognitive resistance predicated on context collapse at the point of encoding (the social media site, in this instance Facebook) before entering into the broader public discourse.

Figure 1.

Everyday Racism in Digital Policing Conceptual Model 1.

The model shows the cognitive flows, the production, and comprehension of discourse/interaction, controlled by context, and based on knowledge and ideologies indicated by the solid black lines, between Henry’s post on the social media site, and it being entered into the broader public discourse through the surveillance practices of the NYPD. Both context collapse and the resistant reading of Henry’s post, implicit because of the discursive relationship between broader society and law enforcement, lead to a decoding of his behavior as criminal. The impact of the social system of racism, acknowledged by the very existence of the 1964 Civil Rights Act and operationalized through a discrimination defined by Feagin and Eckberg as “actions or practices carried out by members of dominant racial or ethnic groups that have a differential and negative impact on members of subordinate racial and ethnic groups” (Feagin, 1991, p. 102), on the online public discourse, constructed as theorized by the model through the interactions between broader society and law enforcement, left Henry with no mechanism to refute the resistant reading of his social media performance of self by law enforcement—which in turn, led to his arrest and incarceration. The model also illustrates how the enforcement of cognitive controls (racial attitudes and ideologies; Van Dijk, 1993, 2001) on the digital street is underpinned by the social control exerted by broader society on the everyday lived experience of persons of color in the United States. Factors such as access to work, housing, education, social security benefits, healthcare, and education are represented by the dotted line which flows from online public discourse to acceptance and then into communities of color, becoming the lenses through which this acceptance is refracted and participation allowed. Taking up Rios’ (2011) deployment of sociologist Ann Swidler’s concept of “culture as repertoire,” “individuals deploy different, often contradicting actions in the social world based on the needs demanded by specific social situations” (p. 109). We elucidate that in response to lack of access to these social indices, some members of communities of color push back through apparently deviant performances of self that challenge authority figures who threaten their safety, autonomy, and self-determination.

In direct counterpoint to Henry, Figure 2 illustrates, in the case of Dylann Roof, how the power of prejudice affects who is surveilled online. In June 2015, Roof, a young White man, walked into the Emanuel AME church in Charleston, South Carolina, a historically Black church, and shot nine people to death. Until this act, Roof had not been on law enforcement’s radar in spite of a public social media presence including a manifesto of racial hatred and murderous intent. Roof is but the most recent domestic terrorist whose public performance of self on social media was conveniently neutralized by his race.

Figure 2.

Everyday Racism in Digital Policing Conceptual Model 2.

We argue Roof’s privilege as a White male allowed him to enter into the broader public discourse without either the benefit or hindrance of having his social media communications encoded and decoded by the police. Therefore, he begins his interface with the discursive relationship between society and law enforcement from a position of acceptance and curiosity, as represented by the solid lines. The impact of Roof’s privileged position as a young White man, indicated by the absence of dotted lines between him and deviant politics, the “political actions—the resistance—that youth labeled by society as deviant use to respond to punishment that they ubiquitously encounter” (Rios, 2011, p. 118) clearly demonstrates the inimical effects of this type of policing. Roof’s privilege, as we have seen, rendered his online public performance of deviance virtually invisible. In contrast, the ways in which the social media behaviors of people of color are internalized and interpreted transform groups into gangs, young men into criminals, and common gestures into threats. These associations are often driven by one’s negative implicit attitudes and stereotypes toward people and communities of color.

Discussion and Conclusion

In this article, we argue that the use of online surveillance as part of policing strategies raises critical policy, legal, and ethical questions. In a 21st century age where much of social interactions between individuals have migrated, at least in part, online, it seems only obvious that these interactions be treated as legitimate parts of law enforcement investigations. For law enforcement agencies, developing policing practices adapted to the new challenges posed by people narrating their lives online appears to have created tactical challenges and possibilities. Among the challenges is an overwhelming quantity of available data and questions over how this data should be used, sorted, and stored, adapting to online talk, and the pitfalls of taking online talk out of context. Because posts online are highly visible and permanent, young people often leave a firm paper trail that can play a key part in building cases for investigations, charges, and convictions.

Are there racial implications of social media policing? First, we know that vast majority of prison inmates are men of color. Second, this self-validating argument has been used by law enforcement agencies to increase surveillance not just on people of color but also on communities of color—a surveillance through which just being Black becomes shorthand for engaging in criminal activity and is enough to prompt often fatal police interest. Third, a Lexis Nexis (2012) survey of 1,200 federal, state, and local law enforcement professionals done in 2012 found that approximately 80% used social media platforms as intelligence gathering tools. Conversely, perhaps, current policing practices like necessary regulation and transparency to counter the potential for discrimination in their social media strategies?

The ubiquitous use of social media policing, particularly within the criminal justice system begs the question, are we a safer society? Does surveillance reduce violence or does it permit targeted violence against Black and Brown individuals? Who is being surveilled and why?

Social media policing as a criminal justice strategy raise questions about the unquestioned and un-scrutinized social media tactics being developed that one could argue have little regulation. These tactics appear to be used disproportionately in cases relating to urban American communities of color, in “gang” related cases, and in “anti-terrorism” monitoring, which has affected Muslim communities and communities thought to be Muslim-affiliated. Lane and Ramirez (2016) examined seven formal conspiracy charges against “gangs” in Harlem and found that 46% of all the evidence referred to communication or activity on social media. The online targeting of distinct communities, and individuals within them, made clear in indictments and other court documents, policy announcements, and strategy outlines, point very convincingly to possible racial and religious bias and profiling when it comes to the way in which online spaces are policed. There are, however, opportunities for law enforcement to create a stronger culture of transparency in a digital era. In an effort to maintain effective policing strategies and avoid racially discriminatory practices, law enforcement should consider (1) clearly articulating the technology rights of citizens, (2) invest in resources and collaborations with academics and communities to support efforts in contextualizing social media posts, (3) sharing practices and strategies with violence prevention organizations to encourage safe technology use.

We offer everyday racism in social media policing as an emerging conceptual framework for conceptualizing how various forms of racism affect social media policing strategies. Well known cases such as Jelani Henry and the massive gang raid associated with Operation Crew Cut in New York City elucidated years of social media monitoring, in some cases before any crime was committed, driven by labels such as “gang” or “gang-affiliated” that are highly racialized (Lane & Ramirez, 2016). Indeed, until it bumps into them, like it did in the NYPD shocking mass arrest, young people appear to be unaware of the breadth of online social media monitoring. This may be due to a combination of factors, including a false sense of protectiveness and privacy online (Marwick 2012), a complete lack of public debate around the use of online social media surveillance for law enforcement activities, and the fact that, by its very nature, law enforcement surveillance of social media communication is imperceptible (Trottier, 2012b). Given that law enforcement continues to train its technological gaze on young men of color in anticipation of their anticipated future deviance, performances such as those of Dylann Roof serve to highlight the grim reality that though social media surveillance is inherently imperceptible by nature, it is equally inherently racist by practice.

If social media policing by law enforcement raises the same questions of preemptive, dragnet policing, harassment, and racial and religious profiling that stop-and-frisk as a practice does, it is vastly different in its perceived, immediate nefariousness. As our models demonstrate, the same cognitive and social controls that dominate the everyday lived experiences of people of color in the United States are those which dictate who is watched and what is seen online. Being on the digital street has become no different to people of color to being in the world—the same cognitive and social controls apply. Unlike stop-and-frisk as a practice, which is physically invasive and threatening to the subject questioned and searched, the effects of social media policing on a watched individual are not immediately apparent. However, the lack of direct physical intrusiveness may hinder any kind of mobilization, grassroots or otherwise, around the use of online, preemptive policing. In this way, the imperceptible nature of social media policing may help the ongoing survival of preemptive, suspicion-less policing and racial profiling as policing tactics without it are replaced (as they were in 2013, when a federal court ruled stop-and-frisk unconstitutional). In this way, the asymmetry favoring police forces over community members, documented earlier in this article, is arguably exacerbated in unregulated online spaces.

The future research should explore whether the wholesale online surveillance of urban communities of color amounts to gross police misconduct and racial profiling. Are we creating an “entrapment” mechanism, waiting for certain members of the community to commit a crime (or in gang-related cases, be affiliated to a committed crime), and letting other unwatched members of the community off scot-free? Does “paper trail” collecting possibilities offered through online surveillance make law enforcement effective at preventing crime, or is it a legitimizing mechanism for racial profiling and social control? Answering these questions would most likely involve identifying best practices and cultivating trust and transparency within agencies, where policies regarding the ongoing development and adaptation of social media surveillance should be disclosed and challenged.