Abstract
Contemporary American policing has routinized involuntary police contacts with young people through frequent, sometimes intrusive investigative stops. Personal experience with the police has the potential to corrode adolescents’ relationships with law and skew law-related behaviors. We use the Fragile Families and Child Wellbeing Study to estimate how adolescents’ experiences with the police shape their legal socialization. We find that both personal and vicarious police contact are associated with increased legal cynicism. Associations are present across racial groups and are not explained by teens’ behaviors, school settings, or family backgrounds. Legal cynicism is amplified in teens reporting intrusive contact but diminished among teens reporting experiences characterized by procedural justice. Our findings suggest that aggressive policing risks weakening teens’ deference to law and legal authorities.
Recent high-profile incidents of police violence toward citizens have underscored the everyday presence of police in the lives of young people. Contemporary policing, including “proactive policing” models, has routinized police contacts between citizens and police (Kubrin et al. 2010; Tyler, Fagan, and Geller 2014). These regimes expose teens to police in their everyday routines, translating into regular and involuntary police-citizen interactions through frequent and sometimes intrusive investigative stops and frisks, often on threadbare suspicion of criminal behavior (Fagan et al. 2010; Fagan and Geller 2015; White and Fradella 2016). In both large and small cities, these contacts can lead to official sanctions in the form of noncriminal summons for violations of municipal codes or arrests for minor misdemeanors (Fagan and Ash 2017). Studies show that the burden of these police contacts and arrests conditional on police encounters fall on young minority males (on burden, U.S. Department of Justice 2015, 2016, 2017; on arrests, Kochel, Wilson, and Mastrofski 2011).
Police are also regularly present in urban and suburban schools, and often have the authority to make arrests and engage in other enforcement activity, often for minor incidents that could be handled informally by school officials (Kupchik 2010). As is true of aggressive street policing, the burden of police contact in schools falls predominantly on black and Latino youth (on policing, Fagan et al. 2010; Weitzer, Tuch, and Skogan 2008; on schools, Nance 2016; Rocque and Paternoster 2011; on youth, White 2015).
Personal experience with the police and other forms of interpersonal racial discrimination are critical factors in the legal socialization of adolescents (Berg et al. 2016; Brunson 2007; Burt, Lei, and Simons 2017; Fagan and Tyler 2005; Fagan and Piquero 2007). By legal socialization, we refer to the interaction of natural maturation with a broad set of situational experiences. Interactions with legal authorities are a key feature of those experiences, because for most adolescents, police stand alongside school authorities as the face of the state (Shedd 2015). Through those interactions, children and adolescents develop values and attitudes about law and the legal actors that enforce it; these legal interactions frame their cognitive schema of the socio-legal landscape around them (Burt, Lei, and Simons 2017).
The frequency of police-youth contacts in poor neighborhoods skews the locus of adolescent socialization in those places toward their interactions with police. Carla Shedd finds that Chicago youths stopped by the police show high rates of distress and perceptions of injustice (2015). Rod Brunson and Ronald Weitzer identify feelings of “hopelessness” and being “dehumanized” (2009). Benjamin Justice and Tracey Meares contend that people gain information about their position in society from interactions with the legal system throughout adolescence (2014). This forms the basis of their relationship with legal authorities and their sense of democratic belonging and obligation to the law (Epp, Maynard-Moody, and Haider-Markel 2014; Bell 2016; Soss and Weaver 2017).
Others show that legal cynicism runs deeper among youths and adults in neighborhoods that are more heavily policed (Kirk and Matsuda 2011; Desmond, Papachristos, and Kirk 2016). Particularly if the “dosage” of police contact is strong, citizens who feel they have been treated harshly or unfairly by the police, experienced procedural injustice (Bell 2016), or were stopped due to racial discrimination are at risk of diminished perceptions of police legitimacy (Tyler, Fagan, and Geller 2014) and the development of legal cynicism (Brunson 2007; Fagan and Tyler 2005; Kirk and Matsuda 2011). Because policing is woven into the social fabric of urban neighborhoods, teens’ legal socialization might also be influenced by police activity that they witness in their neighborhoods, even if they are not personally involved (Stuart 2016). Both Dennis Rosenbaum and his colleagues and Brunson and Weitzer identify a “vicarious” experience of policing, in which perceptions of the police are influenced not only by one’s own experiences, but also by the experiences of others (Rosenbaum et al. 2005; Brunson and Weitzer 2009; compare Fagan and Piquero 2007). Each additional direct or vicarious interaction provides new information and experiences that can add to their evaluations of legal authorities (Fagan, Tyler, and Meares 2016). These interactions and socialization experiences influence crime over time, especially in the distinct contexts of adolescent development for African American youths (Burt, Lei, and Simons 2017).
In this article, we examine the intersection of aggressive policing and legal socialization of teenagers and young adults, with a focus on one dimension of legal socialization: legal cynicism (Sampson and Bartusch 1998; Bell 2016). Following Robert Sampson and Dawn Bartusch, we define legal cynicism as “anomie about law” (1998, 778). “Anomie” was a state of disconnection of individuals from both community and the social and legal norms of the state. More recent expressions of legal cynicism emphasize the rejection of the law and its agents as “illegitimate” and “unresponsive” to concerns about safety and justice (Kirk and Papachristos 2011, 1191). These perspectives view legal cynicism as disrupting willing deference to legal actors and as unraveling social cohesion and the social bonds that connect people to each other and the state (on legal actors, Tyler and Huo 2002; Carr, Napalitano, and Keating 2007; on bonds, Sampson and Bartusch 1998). Monica Bell expands the concept of legal cynicism to include an animating process resulting from experiences of procedural injustice in a situational context of social exclusion and marginalization (2016). Exposure to young adults to these policing tactics is woven into the developmental landscape of children and adolescence, potentially skewing their socialization to law, legal actors, and underlying social norms.
From this framework, we assess how police-youth interactions shape the legal socialization of adolescents under social conditions of intense police surveillance and contact. We focus on adolescents with high exposure to the criminal justice system and estimate the extent to which their personal and vicarious contacts with the police are associated with a reduction in their respect for the police and an increase in legal cynicism. Using regression and matching models, we find that adolescents who have been stopped by the police, witnessed police stops, or know people who were stopped report greater levels of legal cynicism than their counterparts without police contact. Moreover, the conduct of these police encounters matters: legal cynicism is amplified in teens reporting more intrusive contact but diminished among teens who report that the police behaved with consideration for procedural justice. These associations are present for black, white, and Hispanic teens, robust across multiple model specifications, and not explained by the teens’ behavior, school settings, or family backgrounds.
PROACTIVE POLICING
Policing in the United States has changed substantially over the past four decades (Skogan and Frydl 2004; Braga and Weisburd 2010; Weisburd and Majmundar 2018). Many urban police departments have shifted from a reactive posture to aggressive tactics such as “proactive policing” (Kubrin et al. 2010), “order maintenance policing” (Livingston 1997), and “broken windows” policing (Kelling and Coles 1996; Kelling and Wilson 1982). These models emphasize the active engagement of citizens at low levels of suspicion and aggressive enforcement of minor crimes and civil violations. Debra Livingston and Philip Heymann each describe this as the “new policing,” featuring the integration of advanced statistical metrics, new forms of organizational accountability, and aggressive enforcement of minor crimes (Livingston 1997; Heymann 2000). Police also apply this model to use field interrogations or investigative stops as prophylactics to scrub from local areas the social conditions thought to contribute to crime (Skogan 1990; Harcourt 1998; Taylor 2001). The model has been adopted in large and small cities, and institutionalized in everyday police-citizen interactions, especially among residents of poorer, minority, and higher crime areas (Fagan et al. 2016; Livingston 1997; Skogan and Frydl 2004; Kohler-Hausmann 2014; Soss and Weaver 2017; Weisburd and Majmundar 2018).
Applying these tactics, police have saturated many communities with surveillance and proactive contacts. In 2011, more than 62.9 million U.S. residents, 26 percent of the population age sixteen or older, reported contact with the police over the previous twelve months (Langton and Durose 2013). About half of those experienced police-initiated, or involuntary, contact, such as an investigative stop while driving or as a pedestrian (Langton and Durose 2013). The rich data on police stops in New York City provide a basis for estimating the prevalence of police stops (White and Fradella 2016). Between 2004 and 2012, the New York City Police Department recorded more than two hundred thousand police-initiated stops of youth between the ages of thirteen and fifteen (NYPD 2016). Jeffrey Fagan and colleagues estimated that up to 80 percent of African American males between sixteen and twenty-four may have been stopped once or more by the NYPD in 2008, versus 38 percent of Latino males and 10 percent of white males (2010). Precision in these estimates is difficult given variations in police reporting, the presence of nonresidents in the population of those stopped, and the possibility that individuals are stopped multiple times. Tom Tyler, Jeffrey Fagan, and Amanda Geller estimate, based on a 2012 stratified random sample of eighteen to twenty-six year old males living in New York City, that 43.2 percent of respondents were stopped by police in the year leading up to the survey (2014).
A survey of Chicago public school students found that approximately half had been stopped, questioned, and “told off or told to move on” by ninth or tenth grade (Shedd 2015). Police officer presence has also become prevalent in schools, and police often have the authority to make arrests and engage in other enforcement activity (B.H. v. City of New York. Amended Complaint 10 CV 0210 (RRM)(ALC) (2010); Fowler et al. 2010; Kupchik 2010; Na and Gottfredson 2011; Owens 2017). Driven in part by this police contact, evidence from the National Longitudinal Survey of Youth shows that by age eighteen, cumulative arrest prevalence rates range from 30.2 to 41.4 percent (Brame et al. 2012).
Police exposure and resulting sanctions are racially skewed. The cumulative arrest rate by age twenty-three is 49 percent for black males and 38 percent for their white counterparts (Brame et al. 2012). With greater exposure to police who are applying these aggressive patrol tactics comes a greater risk of police contact and violence during those contacts (Fagan 2017; Eckhouse 2018). Even in a period of declining police stops in New York, racial and neighborhood disparities in intrusive policing persist, with the distribution by race no different in 2015 than in the peak year of police stop activity in 2011 (Zimroth 2017).
Police Contact and Legal Socialization
Legal socialization is a developmental process of forming a relationship to the law and legal authority (Trinkner, Jackson, and Tyler, in press). It begins in early adolescence and continues into young adulthood (Fagan and Tyler 2005; Piquero et al. 2005; Fagan and Piquero 2007; Stewart et al. 2009; Berg et al. 2016). During this period, from their everyday exposure to policing, young people develop views about the social and moral norms that legal actors enforce and about the norms and rules that those authorities represent. These views develop from adolescents’ interactions in social institutions and social settings where authority can exert control and has the capacity to punish and to confer status about a person’s social value and societal role (Justice and Meares 2014). Adolescents also expand the empirical basis for their judgments by observing the interactions of family, peers, and neighbors with legal authorities to broaden their views of the moral authority and fairness of law.
Three features of legal socialization or experience with law inform this project. First, police matter more than other authorities. In an era when school discipline overlaps with policing, and when policing is integrated into the school environment and embedded in many neighborhoods, police figure prominently in how adolescents view legal authority and legal rules (Weitzer and Tuch 2006). The totality of adolescents’ contacts with police in schools and on the street in the new policing models places the locus of legal socialization in their contacts with police, particularly during early to mid-adolescence.
Second, legal socialization is inherently a learning process. Through interactions with legal actors, adolescents and young adults experience “teachable moments” that signal the underlying rules and norms of legal regimes and actors (Stewart et al. 2009; Tyler, Fagan, and Geller 2014; Justice and Meares 2014; Berg et al. 2016). Experiences move from teachable moments to socialization processes through not just the content, but also the emotional and cognitive weight of the sum of their experiences. Positive experiences with legal actors can reinforce law; negative experiences can teach the opposite lesson through anger and fear reactions to the unfair or abusive exercise of legal power. These competing and reinforcing processes create a tension between viewing legal authorities as fair and respectful or as abusive and illegitimate (Fagan and Piquero 2007). The elements of procedural justice can be thought of as powerful emotional engines that can bind or distance adolescents from the police or other legal actors (Kirk and Papachristos 2011).1 When interactions with police are harsh or intrusive, the psychological fallout—stress, stigma, anger—can skew the meaning of legal actors and the laws they stand for (Geller et al. 2014). Moreover, the effects of these experiences are cumulative, so the emotional weight of one experience can shape the cognitive frame through which subsequent experiences are evaluated and internalized.
Third, legal reasoning and decision-making are influenced by these experiences. Both Monica Bell and Mark Berg and his colleagues, using quite different methods, show how perceived injustices can produce legal cynicism and alienation from—or even opposition to—the law or its agents (Bell 2016; Berg et al. 2016). Legal socialization can influence deference to the law by conferring or withholding the law’s legitimacy, and the emotional aftermath of accumulated negative experiences can produce cynicism that changes legal reasoning (Tyler and Fagan 2008). Both at the individual and neighborhood levels, high rates of legal cynicism can lead to higher offending rates (individual, Tyler and Fagan 2008; Fagan and Piquero 2007; neighborhood, Kirk and Papachristos 2011; Tyler, Fagan, and Geller 2014). Legal cynicism also can reduce incentives to cooperate with police in solving crimes, leading to a spiral of crime, intensive policing, and legal cynicism in the most heavily policed communities (see Kirk and Matsuda 2011; Tyler, Fagan, and Geller 2014; Gau and Brunson 2010; Desmond, Papachristos, and Kirk 2016).
Race Differences in the Effects of Police Contact
There are reasons to think that the effects of police contact on legal socialization may vary by race; however, the nature of race moderation is theoretically ambiguous. If police officers use racial invective, or subjects believe they were targeted because of their race, the stress and stigma of an encounter may be compounded and have consequences for legal socialization (Anderson 2013; Hatzenbuehler et al. 2010; Krieger 1999; Phelan and Link 2015; Sawyer et al. 2012). The effects of police contact on legal cynicism may also be amplified for minority youth if they perceive racial targeting and that their encounter was therefore unjust (Stewart et al. 2009; Gau and Brunson 2010; Berg et al. 2016). On the other hand, the increased exposure of minority youth to the police has the potential to foster resilience, depending on the neighborhood, family, and other social contexts of interactions with police, and may attenuate any adverse effects of a given encounter on their legal socialization (Burt, Lei, and Simons 2017; Geller, Fagan, and Tyler 2017).
METHODS
The current project extends our understanding of these implications of adolescents’ contacts with the police for legal socialization, using new data from a large multiwave population-based sample of urban teens across multiple cities and social contexts.
Data
Data are drawn from the Fragile Families and Child Wellbeing Study (FFCWS), a birth cohort survey of children born in large cities that has become a leading source of data on urban families and the social environment. The study follows a cohort of nearly five thousand couples with children born between 1998 and 2000 in twenty large U.S. cities (Reichman et al. 2001). The study systematically oversamples unmarried parents, providing a sample that contains mostly racial and ethnic minorities and faces significant social disadvantage, but when weighted or regression-adjusted is nationally representative of urban births. Parents are surveyed at the time of their child’s birth, and follow-up surveys are conducted when the children are one, three, five, nine, and fifteen years old (Y1, Y3, Y5, Y9, and Y15 follow-up waves). The study’s “focal children” were interviewed at the Y9 and Y15 follow-ups; at Y15 more than three thousand were asked about their experiences with the police, police contact among their peers and others they know, and their perceptions of the law and police-community relations. These data build on five previous waves of interviews with parents and other caregivers, assessments of child development and behavior, and various measurements of the children’s social environments.
Key Measures
Key measures in this study include legal cynicism and legal socialization; adolescent-police contact; and demographic, socioeconomic, and behavioral characteristics.
Legal Cynicism and Legal Socialization
Legal cynicism is measured using a series of six questions related to the focal teens’ perceptions of the police and the law (Sampson and Bartusch 1998; Kirk and Papachristos 2011; Berg et al. 2016; Geller et al. 2014). Subjects report their level of agreement (on a 5-point Likert scale) with the following six statements: “I have a great deal of respect for the police.” “It’s okay to do anything you want.” “There are no right or wrong ways to make money.” “Laws were made to be broken.” “If I fight with somebody it’s nobody else’s business.” “The police create more problems than they solve.” Responses are combined in an additive scale (α = 0.66), each coded so that higher values indicate a greater legal cynicism.
Adolescent-Police Contact
Adolescent experiences with the police are measured using self-reports of personal contact (in which the teen reports having been stopped by the police), and vicarious contact (in which the teen reports having witnessed a police stop of someone else, or personally knowing someone who has been stopped by the police). Although teens have opportunities to report both personal and vicarious contact, and many (approximately 25 percent) report having experienced both, our analyses examine differences between mutually exclusive groups: teens reporting personal contact, teens reporting vicarious but not personal contact, and teens reporting no contact.
Teens reporting personal or vicarious police contact report on several domains of the contact that they have experienced, witnessed, or heard about. (Adolescents with personal and vicarious experience are asked specifically about their encounters rather than encounters they witnessed or heard about). Asked about “the incident that stands out most in [their] mind” (their critical stop or most memorable stop), teens report whether the stop involved the officer frisking them (or, for vicarious contact, the person stopped), searching their bags or pockets, using harsh language, using racial slurs, threatening physical force, and using physical force. Binary indicators of these force domains are totaled to form an index of police intrusion in the critical stop (α = 0.75). Teens with stop experience (personal or vicarious) also complete a three-item omnibus measure of procedural justice that measures, in the incidents the youth experienced, witnessed, or heard about, whether the police “explained why they stopped [the person stopped] in a way that was clear to them,” “treat[ed them] with dignity and courtesy,” and “respected [their] rights” (on procedural justice, see Tyler 2003). Questions were answered “often,” “sometimes,” or “never,” and were totaled with higher values on the scale (α = 0.71) indicating greater procedural justice.
Demographic, Socioeconomic, and Behavioral Characteristics
Adolescent experiences with the police, and their potential consequences for legal socialization, were evaluated in the context of teens’ demographic, socioeconomic, and behavioral characteristics. Respondent race is self-reported, and supplemented by parents’ self-reported race when teens’ responses cannot be coded. Analyses also consider adolescent age, their mothers’ educational attainment, and their parents’ relationship status at the time of their birth. Finally, we consider adolescents’ likely exposure to the police and criminal justice system, measured individually (based on self-reported measures of their early externalizing and delinquent behavior, both reported at Y9), and as an aspect of their family background (such as whether either of their parents is known to have ever been incarcerated) and school environment (specifically, whether a police officer is regularly stationed at their school).
Analysis
Our analysis sample includes the 3,001 teens interviewed who provided information on their experiences with the police (whether they had ever been stopped or experienced vicarious police contact), and their attitudes about the law. Table A1 presents a model, based on the 4,897 families interviewed at baseline, predicting inclusion in our analysis sample, and suggests that our sample differs from the broader FFCWS sample in several ways. Teens in the analysis sample are significantly (p < .05) less likely to be born to Hispanic or “other race” mothers, reflecting greater attrition over fifteen years in these minority families, but marginally (p < .10) more likely to be born to black mothers than white mothers. Children born to mothers in deep poverty (below 50 percent of the federal poverty line) are also significantly less likely to be included in our analysis sample (p < .01). Controlling for mothers’ race and poverty status, teens in our analysis sample do not differ from their counterparts in terms of their parents’ baseline relationship status. Missing covariate values are imputed in fifty datasets. Most covariates are missing 1 percent of observations or fewer; exceptions include whether the teens report a police officer stationed at their school (3 percent), peer delinquency (6 percent), teens’ self-reported delinquency (7 percent) and Y9 externalizing behavior (7 percent), and their fathers’ incarceration histories (13 percent).
Table 1 provides descriptive statistics for the analysis sample. Teens reporting either personal or vicarious police contact also report significantly more cynicism than their counterparts who have no experience with the police. However, those reporting police contact also report more delinquency, externalizing behavior, and aspects of socioeconomic disadvantage that may contribute to their elevated levels of both police contact and legal cynicism. We estimate the association between legal cynicism and stop experience net of these additional factors.
Sample Description: Means, Standard Errors, and Percentages
Our analysis proceeds in three stages. In the first, we use regression and propensity score analyses to assess differences in legal cynicism between teens with and without police contact. We next assess outcome differences by the nature of contact that teens report: whether the teens report that they were personally stopped or that they witnessed involuntary police contact or knew someone stopped, as well as their reports of officer behavior during these encounters. Finally, we assess the moderating effects of respondent race and ethnicity on our estimates of the relationship between police contact and legal socialization.
Socialization Differences by Police Contact
Our first models examine how legal cynicism differs between teens with and without contact with the police, whether personal or vicarious. Model 1 is an ordinary least squares regression estimating differences with controls for race, age, sex, and a series of behavioral and socioeconomic characteristics likely to be associated with adolescents’ police exposure and their subsequent legal cynicism: mothers’ baseline educational attainment, parents’ relationship status at baseline, their own externalizing behavior, and their exposure to the police through self-reported delinquency (measured at Y9), fathers’ criminal justice history, and their school environment.
Model 2 uses propensity score matching to estimate the effects of police contact experience, controlling for the distributions of the covariates of police contact. Within each of the fifty imputed datasets, we use a probit specification to generate a propensity score for each individual, and use nearest-neighbor matching with replacement to identify outcome differences in each dataset and combine estimates across imputations (Dehejia and Wahba 2002). We next estimate two parallel models (models 3 and 4) focusing on differences in legal cynicism between teens who report personal experience with the police and those who have not. In these models, teens reporting only vicarious experience are modeled as having been “untreated,” whereas in models 1 and 2 they were considered to be part of the treatment group.
Socialization Differences by the Nature of Contact
In the second stage, we move from examining binary indicators of police contact to indicators that provide additional detail on the nature of police contact that teens report. Model 5 parallels models 1 and 3 to examine associations between police contact and legal cynicism, with controls for the complete set of covariates laid out in table 1 but identifies differences in legal socialization between teens with personal, vicarious, and no experience with the police. Model 6 controls not only for whether respondents had personal or vicarious experience with the police, but also for the level of intrusion they reported in their critical stop. This model includes an interaction term to distinguish whether the respondents’ critical stop was personally or vicariously experienced. In these models, teens with no stop experience (personal or vicarious) are coded as having experienced zero police intrusion. Finally, model 7 controls not only for the intrusion in respondents’ critical stops, but also for respondents’ perceptions of procedural justice in their experiences with the police. We again use an interaction term to distinguish between procedural justice in personal and vicarious stops. Teens reporting no police contact are assumed to perceive the maximum level of procedural justice.
Racial Differences in the Legal Socialization Relationship
The final stage of our analysis re-estimates models 5 through 7 separately, in turn, for respondents who are white, black, and Hispanic. The FFCWS has too few respondents of other race or multiple races for race-specific models to be meaningful.
RESULTS
Our analyses indicate that the criminal justice system is deeply embedded in the lives of urban adolescents. As shown in table 1, more than 25 percent report having personally been stopped by police once or more, and nearly 80 percent report vicarious police contact.
Exposure to Criminal Justice
Table 2 provides details of these contacts and criminal justice exposure generally. Rates of police contact are racially skewed: black and Hispanic teens are significantly more likely than others to have personal police contact. Adolescent exposure to the criminal justice system extends beyond police stops: more than half of teens in the analysis sample have fathers who have been incarcerated, and more than 80 percent report a police officer regularly stationed at their school.
Intrusion Reported by Teen Respondents in Most Memorable Police Contact
Respondents report considerable intrusion in their critical police encounters. More than one-third of teens with personal experience and approximately half of those reporting vicarious contact report that they (or the person stopped in the vicarious contact) were frisked or searched during their most memorable stops. More than 20 percent of teens personally stopped and approximately 15 percent of those reporting vicarious contact report that the officers used harsh language (a smaller proportion noted that the officer used racial slurs), and more than 10 percent reported that the officer threatened or used physical force.
Police Contact and Legal Cynicism
Table 3 presents the estimated associations between the binary indicators of police contact and respondent legal cynicism. Respondents with police contact (personal or vicarious) report significantly more legal cynicism than those with no contact, a difference that is slightly amplified when focusing on personal contact. Notably, estimates obtained through propensity score matching are of comparable magnitude to those obtained through regression analysis with our full set of controls, increasing our confidence that regression analysis is a suitable approach for our subsequent analyses, which focus on the nature of the stops that teens report.
Associations Between Adolescent Stop Experience and Legal Cynicism
In addition to being linked to personal experience with the police, legal cynicism is also significantly associated with other indicators of social disadvantage. Minority (specifically, black, Hispanic, and multiracial) teens report significantly more legal cynicism than their white counterparts, net of racial differences in their reported personal and vicarious police experience. Boys report greater cynicism than girls, and reported legal cynicism increases with respondent age. Finally, legal cynicism is greater among teens whose fathers have incarceration histories, though the difference is only marginally significant in the model focusing on personal contact.
The Nature of Police Stops
Table 4 shows results from analyses examining the nature of police stops and their implications for adolescent legal cynicism. Model 5, which separately examines personal and vicarious contact, indicates that although personal and vicarious police contact are associated with increases in legal cynicism, teens with personal contact report significantly greater legal cynicism than those reporting only vicarious contact. Notably, legal cynicism is significantly associated not only with whether teens report (personal or vicarious) contact with the police, but also with the adolescents’ reports of what happened during the stop. Model 6 finds that controlling for the indicators of police contact, teens reporting more intrusive encounters with the police also report significantly greater levels of subsequent legal cynicism. The significant negative interaction between stop intrusion and vicarious stops indicates that intrusion in a stop the teen witnessed or heard about is a weaker predictor of legal cynicism than intrusion in a stop they experienced. However, a significance test of the sum of the main effect and interaction term indicate that stop intrusion is associated with increases in legal cynicism for teens reporting vicarious as well as personal police contact.
Associations Between Aspects of Adolescent Stop Experience and Legal Cynicism
Model 7 shows that teens perceiving greater procedural justice in their police encounters report less legal cynicism than teens reporting lower levels of procedural justice. As in model 6, the interaction between perceived procedural justice and having only vicarious, rather than personal, contact with the police is in the opposite direction than the estimated main effect, suggesting that effects of procedural justice are stronger for teens with personal, rather than vicarious contact. However, the interaction is not statistically significant, and the combined procedural justice estimate indicates significantly less legal cynicism for teens reporting greater levels of procedural justice in the encounters they have witnessed or heard about.
Our binary indicator of personal experience with the police is independently associated with teens’ legal cynicism, as is stop intrusion for teens with both personal and vicarious contact. However, model 7 shows that controlling for stop intrusion and perceived procedural justice in the encounters they witnessed or heard about, adolescents reporting only vicarious contact report less legal cynicism than those reporting no contact. Accordingly, the association between vicarious police contact and adolescent legal cynicism is inextricably linked to the interaction quality in the stops that teens see and hear about. Teens with vicarious exposure to stops with minimal intrusion, and stops with high levels of procedural justice, report little cynicism; those with intrusive stops, and stops with low levels of procedural justice, report significantly more.
Race Differences in Police Contact and Legal Cynicism
Notable in tables 3 and 4 are the increased levels of legal cynicism reported by black, Hispanic, and multiracial teens relative to their white counterparts, controlling for their personal and vicarious contact with the police and multiple alternate sources of criminal justice exposure (such as their fathers’ incarceration histories and the presence of police officers at their schools). Selected coefficients from a race-specific estimation of models 5 through 7 are presented in table 5. Although estimated associations between legal cynicism and personal experience with the police are slightly greater in magnitude for racial and ethnic minority teens than for white teens, the most notable finding in table 5 is the relative consistency of the estimated relationships between personal experience with the police and legal cynicism. Regardless of race, teens stopped by the police report significantly more legal cynicism than their counterparts with no contact, and this relationship increases significantly with critical stop intrusion and declines significantly with their perceptions of procedural justice.
Race-Specific Estimations of Models Predicting Legal Cynicism
For all three racial groups, the relationship between police contact and legal cynicism is less pronounced for teens experiencing vicarious contact only, particularly in model 6 and model 7, which consider stop intrusion and perceptions of procedural justice. For white teens, model 7 indicates that vicarious contact is not consistently associated with increased legal cynicism in these models, nor is stop intrusion. For minority teens, the opposite is true: stop intrusion and legal cynicism are significantly associated for black and Hispanic teens with vicarious contact.
Sensitivity Analyses
Because our analyses are based on data from fifty datasets created by multiple imputation, we test the robustness of our findings to an alternate mode of dealing with missing data. Table A2 presents estimates of models 5 through 7 based on the 2,155 teens with complete data on all included measures. Our estimated relationships of primary interest are almost identical in magnitude and direction to those in table 4, differing by magnitudes of tenths or hundredths of scale score points. Statistical significance declines somewhat in our complete case sample, presumably partly because our complete case sample is 28 percent smaller than the imputation datasets. However, most relationships that are statistically significant in our full sample are also significant in our complete case sample. It is thus highly likely that our estimated associations are the result of a substantive relationship between police contact and legal socialization, rather than a statistical artifact of the process used to account for missing data.
We also examined the sensitivity of findings to our choice of outcome measure. As noted, our measure of legal cynicism is an additive scale consisting of six items, each coded as four-point Likert measures with higher values indicating greater legal cynicism. We re-estimated models 1, 3, and 5 using ordered logit models to predict each item individually, as a function of, respectively, any police contact, personal experience with the police, and separately, personal and vicarious experience with the police. Results, presented in table A3, indicate that the estimated relationship between police contact and legal cynicism is largely robust to our choice of outcome. Five of the six survey items are significantly and positively associated with all measures of police contact in all three models examined; the sixth (“It’s okay to do anything you want”) is significantly and positively associated with both any contact and personal contact, but its association with vicarious contact is not statistically significant. We also re-estimate models 5, 6, and 7 predicting a two-item outcome combining only the items specifically measuring attitudes toward the police (“I have a great deal of respect for the police” and “The police create more problems than they solve”) and including the other items as control variables. Results from these models are largely substantively consistent with those in table 4. Taken together, these results suggest that the associations in tables 3 through 5 are not the result of a particularly influential survey item, but are instead robust to our measure of legal cynicism.
Finally, we examined the sensitivity of our findings to a measure of whether, in addition to contact with the police, the adolescent respondents report having been arrested. Our results, presented in table A4, indicate that arrests are indeed significantly associated with legal cynicism: teens who have been arrested score more than a unit higher on their legal cynicism scales. However, the estimated associations between stop experience (personal and vicarious), stop intrusion, reported procedural justice and legal cynicism remain strong and statistically significant when controlling for arrest experience. This finding suggests that the association between stop experience and legal cynicism is not simply the result of an adverse outcome (such as arrest), but is associated with the stops themselves.
DISCUSSION
We identify a significant and robust relationship between adolescent exposure to the police and legal cynicism. In nearly all models, teens reporting personal or vicarious police contact report more legal cynicism than their counterparts with no contact, and teens with personal contact report significantly more cynicism than teens with vicarious contact. Our main findings are consistent across racial and ethnic groups; however, we identify significant moderation in these relationships by the nature of teens’ reported contact with the police. Teens reporting intrusive stops report significantly more cynicism, while teens reporting encounters with greater procedural justice report less. Notably, model 7 indicates that intrusion and procedural justice are both significant predictors of legal cynicism, and one relationship is not wholly accounted for by the other. The simultaneous significance of both relationships underscores that intrusion and procedural justice are, at least in part, independent descriptors of teens’ police contact.
Our method relies on reports of stop intrusion that reflect specific actions by police, which teens recall from a particular stop that stands out in their mind. Although these reports are open to interpretation (one teen may report an officer’s language as harsh but another may not) and are at some risk of misinterpretation (the teen may not understand the difference between a frisk and a search), they are likely to be relatively well measured and consistently interpreted. The procedural justice scale measures—in part—the emotional salience of experience with the police. Rather than recalling a single, specific stop, respondents are asked to draw on all encounters they experienced or witnessed or heard about. Some teens draw on a single incident, others a diverse array of personal or vicarious experiences. The procedural justice measure is also based on more subjective aspects of the stop—whether the police officer explained the reason for the stop “in a way that was clear” to the person stopped, whether they treated the person stopped “with dignity and courtesy,” and whether they “respected [their] rights”—and more open-ended measures of quantity, such as whether the officer conduct occurred often, sometimes, or never. This linkage of emotion to a rejection of institutional authority and social norms provides a processual picture of the development of legal cynicism (see Bell 2016; Sampson and Bartusch 1998; Kirk and Papachristos 2011; Berg et al. 2016).
Policing, Legal Cynicism, and Social Inequality
The importance of legal cynicism for law-related behavior, and the increased levels of legal cynicism reported by teens with police contact, is of particular concern given the well-documented racial disparities in police-public interactions (Fagan et al. 2010; Nance 2016; Rocque and Paternoster 2011; Weitzer, Tuch, and Skogan 2008; White 2015). Strong and significant associations between personal police contact and legal socialization are observed among black, white, and Hispanic adolescents. The concentration of police contact among minority teens, particularly when coupled with concentrated racial residential segregation, suggests that legal cynicism may also be ecologically concentrated in minority communities (on segregation, Massey and Denton 1989; on minority communities, Kirk and Papachristos 2011; Kirk and Matsuda 2011; Sampson and Bartusch 1998). To the extent that legal cynicism is associated with subsequent offending behavior, police activity may undermine public safety in these communities (Fagan and Piquero 2007; Kirk and Papachristos 2011; Tyler and Fagan 2008).
When police routinely intervene in the everyday lives of teens, they impose psychological and social interaction costs that inevitably deter young people from moving freely (Fagan and Ash 2017). And when these police actions have legal and economic consequences for those already in disadvantaged social positions, those consequences effectively lock such individuals in by constraining choices of neighborhood selection. Because police deployments and actions are racialized and focused in poor, segregated places, police in effect reproduce inequality, racial stratification, and segregation through their criminal legal enforcement actions and in turn constrain social and economic mobility. More policing in poor neighborhoods leads to more arrests in those places, deepening the ecological concentrations of criminal stigma and social exclusion in places sometimes characterized as poverty traps (Sampson and Morenoff 2006; Fagan and Ash 2017).
Racial segregation and intrusive contact with the police seem to be inextricably linked (Desmond, Papachristos, and Kirk 2016; see also Brunson and Weitzer 2009). The aggressive policing of minority communities and neighborhoods place black and Hispanic youth at increased risk of arrest and subsequent criminal justice involvement (Kochel, Wilson, and Mastrofski 2011). The adjudication process, even for low-level arrests, involves considerable burdens, including financial impositions, exacerbating economic inequality, and impeding the ability of minority residents to move out of high-crime and heavily policed neighborhoods (Feeley 1979; Geller 2016; Kohler-Hausmann 2014; Harris 2016). The conflation of racial segregation and economic mobility means that, typically, a black adolescent or young adult male in a U.S. city lives in very different economic and social circumstances than his white counterpart: different types of schools, different social networks, different levels of access to social capital leading to crime, and different exposure to the police and to violence (Sharkey 2013). The burdens of police contact combined with the blocking effects of segregation mean that these teens are far less likely to better their economic circumstances in adulthood.
These disadvantages extend to health and mental health. The adverse health outcomes associated with police contact also threaten to exacerbate racial disparities in health (on contact, Geller et al. 2017, 2014; Sewell 2017; Sewell and Jefferson 2016; Sewell, Jefferson, and Lee 2016; on health, Harris et al. 2006; Hill 2016). To the extent that the link between policing and legal cynicism undermines public safety in minority neighborhoods, and in turn, increases the perceived need for police surveillance, these disparities may be exacerbated further still.
Limitations and Future Research
Although our analyses identify robust relationships between adolescent reports of police contact and their self-reported legal cynicism, we caution against causal inferences. Our analyses are limited by a dataset that, though it provides a rich description of family circumstances over the teens’ first fifteen years, includes only periodic interviews with family members and has interviewed the study’s teen respondents only twice, about the time of their ninth and their fifteenth birthdays. We have a single measure of legal cynicism and are therefore unable to measure whether the teens’ (personal or vicarious) experience with the police caused a change in their attitudes toward the law, or whether their reports at age fifteen reflect long-standing attitudes unaffected by police contact.
It is also possible that long-standing attitudes about the law (or other personal characteristics) might cause teens to engage in illegal or other risky activities that increase their exposure to the police, escalate the level of intrusion in an encounter they are exposed to, or influence their perception of procedural justice in their reported encounters. To guard against this risk, we control for the teens’ self-reports of early (Y9) delinquency and externalizing behavior, which precede nearly all reported police contact and would likely be affected by long-standing legal cynicism. However, without a pretreatment measure of legal cynicism, our observed associations may still reflect aspects of a reverse causal relationship, as well as any direct effects of police encounters and their conduct on subsequent attitudes.
Our conclusions are also limited by the risk of shared method variance—that our treatment of police contact and our legal socialization outcome are measured by the same teen reporters (on shared method variance, see Bank et al. 1990). Specifically, unmeasured characteristics of the teen respondents may be drivers of behavior that increases their exposure to the police, their perceptions of any police contact they experience or hear about, and their attitudes toward the law. For example, teens who are pessimistic by nature may both perceive a reported encounter as more intrusive or involving more procedural injustice than their peers would and report greater legal cynicism. Such unobserved characteristics of our adolescent respondents may drive a spurious relationship between reports of their personal and vicarious experiences and their legal socialization that are conflated with any causal effects of police contact. A contextual analysis incorporating measures of police activity, including arrests and use of force, can begin to address this limitation, at least in part. Our analyses are also limited by sample attrition. Our analysis sample represents approximately 60 percent of the initial Fragile Families sample, and as noted is less disadvantaged and has a racial composition that differs from the sample as a whole. The extent to which our sample can generalize to a broader population is therefore limited. Nonetheless, the robust associations between police experiences and adolescent attitudes observed, particularly given the high prevalence of contact reported by teens in the sample, suggest that exposure to the police—both positive and negative experiences—have the potential to shape legal cynicism at a turning point of their social development. Future research would advance the field by unpacking the emotional content of legal cynicism and legal socialization more broadly.
One final potential implication remains unstudied for now: city differences in the aggregate behaviors of adolescents exposed to city-specific differences in policing. Policing regimes matter in this framework because they determine the extent and nature of police contact for adolescents. Integrating city indicators of crime and policing is another critical next step.
Appendices
Odds Ratios from Model Predicting Analysis Sample Retention from Baseline Family Characteristics
Complete Case Estimation of Legal Cynicism Models, Selected OLS Coefficients
Selected Regression Coefficients from OLR models
Sensitivity Analysis Examining Arrest as a Predictor of Legal Cynicism, Selected OLS Coefficients
FOOTNOTES
- © 2019 Russell Sage Foundation. Geller, Amanda, and Jeffrey Fagan. 2019. “Police Contact and the Legal Socialization of Urban Teens.” RSF: The Russell Sage Foundation Journal of the Social Sciences 5(1): 26–49. DOI: 10.7758/RSF.2019.5.1.02. The Fragile Families and Child Wellbeing Study was funded by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) of the National Institutes of Health under award number R01HD36916, R01HD39135, and R01HD40421, as well as a consortium of private foundations. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The authors are grateful for feedback from participants at the Russell Sage Foundation authors’ conference and from anonymous reviewers. Direct correspondence to: Amanda Geller at amanda.geller{at}nyu.edu, Department of Sociology, New York University, 295 Lafayette St., New York, NY 10012.
Open Access Policy: RSF: The Russell Sage Foundation Journal of the Social Sciences is an open access journal. This article is published under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.
REFERENCES
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.