Individual-level intervention: How perceptions of neighborhood matter for social organization

ABSTRACT

One of the most robust areas of research in the communities and crime literature investigates the relationship between neighborhood social control and levels of neighborhood crime. While macro-level research provides much insight into the necessary conditions for neighborhoods to exert social control, and as a result experience less crime, far less attention is paid to the determinants of individual-level decisions to intervene in neighborhoods. This study takes a micro-level approach to studying neighborhood social organization by investigating how individual-level perceptions shape resident behaviors within neighborhoods. This paper argues that more attention should be placed on what motivates people to intervene in their neighborhoods, thus contributing to collective efficacy through individual actions. Using survey data from the Study of Race, Crime, and Social Policy in Oakland, California (Street, 2000), analyses show that residents within the same neighborhoods vary in how they perceive objective neighborhood characteristics, and that this variation partly accounts for the likelihood of intervention. Specifically, results from a series of logistic regressions show that perceptions of neighborhood disorder and perceptions of the legal system influenced whether neighborhood residents intervened when presented with problems in their communities.

Acknowledgments

The author would like to thank Drs. María Vélez, Christopher Lyons, Brian Soller, Lisa Broidy, Sharon Nepstad, and Owen Whooley for their comments on previous drafts of this manuscript. The author also thanks the editors and anonymous reviewers at the Journal of Urban Affairs for the insightful feedback they provided.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. A similar concept to formal social control is public social control, which refers to a neighborhood’s ability to solicit resources from the mayor’s office, public resources such as funds for various cleanup efforts (Medoff & Sklar, 1994), and to relate with city officials who tend to the needs of a community (Carr, 2003; Vélez, 2001). Although formal and public control are similar, nuances emphasize that agents of formal social control, such as police, are state-operated, whereas public social control focuses more on the resources provided by lending institutions, and on a community’s relationship with the mayor’s office or other public offices (Vélez, 2001). For the purpose of this paper, the term formal social control will represent both public and formal types of social control. The reason being, this study combines “calling the police,” formal social control, and “calling the mayors office,” traditionally considered public social control, into one variable. Had the option for “calling the police” not been available for respondents to answer they used this method of intervention, then the term “public” or “parochial” social control would be used in this study.Parochial social control refers to the within community supervision that takes place to protect against neighborhood problems, but as Carr (2003) argues, both parochial and public interventions are nearly indistinguishable, as both at times rely on institutions within or outside the community for intervention resources. This concept is also referred to as “public informal social control” (Wickes et al., 2017), and just “informal social control” when looking at any intervention that is not direct police action (Carr, 2003). Since a number of conceptions of informal social control consider all actions that community members themselves take as informal social control (e.g., Carr, 2003; Warner, 2007), to abate confusion between those who intervene using resources and ties within the neighborhood (e.g., attending a neighborhood meeting, calling a neighbor) versus those who use city-resources, including the police, from outside the neighborhood, the terms “informal social control” will refer to citizen actions utilizing within neighborhood resources, and “formal social control” will refer to citizens using city-resources to intervene. Further, an aspect of this paper is to better understand what prompts residents to look within or outside their communities for help, as variation in perspectives may lead to one or the other.

2. Nevertheless, the macro-level focus on collective efficacy illuminates important neighborhood characteristics that contribute to safe communities. Regardless of what this literature does not accomplish, what it does contribute has been invaluable to our understanding of how neighborhood dynamics shape resident livelihoods and has uncovered countless neighborhood inequalities that need attention.

3. For a full articulation of Routine Activities Theory, see Cohen and Felson (1979)

4. Although the presented evidence focuses on perceptions of the police, the police belong to the broader institution of the legal system and are just one arm of the criminal justice system. Yet, negative and vicarious experiences with the police can influence perceptions of the broader legal system to which the police belong. Hurwitz and Peffley (2005) examined the political consequences of negative perceptions of the criminal justice system and found that when people viewed the broader criminal justice system as unfair, they were more likely to view police interactions with civilians as suspicious. Paraschiv (2012) explores the relationships between procedural justice, emotions, and how people interact with the police, and demonstrates that experiences with the broader legal system shape how people interact with police officers. These studies show how the police and the broader legal system are inextricably linked, and experiences with one can shape perceptions of the other. Therefore, the following analyses include measures not just of perceptions of the police, but also perceptions of the broader institution of the legal system.

5. Neighborhood dummies are used in place of a multi-level design for a number of reasons. First and foremost, there are only eight neighborhoods in which to nest residents; and while there is some debate in the methodological literature regarding necessary sample sizes at each level of multilevel designs (e.g., see Bell et al., 2010), the following analyses follow the “rule of thumb” requiring at least 30 units at each level of analysis for robust results. Accordingly, eight clusters (neighborhoods) are inadequate for a multilevel design (Hox, 1998; Maas & Hox, 2004, 2005) Second, 1,129 of respondents were connected to neighborhoods in the study, and 670 of respondents were not connected to a particular neighborhood; thus, the analyses cannot control for neighborhood-level characteristics for those not connected to a neighborhood. Third, the survey does not collect data on actual neighborhood characteristics; the only way to derive neighborhood-level characteristics from these data would be to aggregate respondents’ interpretations of their neighborhood to the neighborhood-level. In doing so, what would be captured in a neighborhood-level variable is simply the aggregated interpretations of neighborhood by respondents, not objective neighborhood characteristics. Finally, the primary research question in this study is how interpretations of neighborhood relate to intervention, not how actual neighborhood characteristics or aggregated perceptions shape intervention practices.

6. Testing for a quadratic effect also somewhat addresses the potential for endogeneity between perceptions of disorder and intervention. It is possible that respondents who perceived high levels of disorder felt compelled to then answer that they “did something” about it. But as results will show, those who perceived neighborhood disorder as an extreme problem facing their neighborhoods were slightly less likely to report having done anything to address the problem.

7. The scale consists of the following questions: Level of crime in the neighborhood; Chances of being burglarized; Chances of being robbed; Chances of having car stolen; Chances of having car vandalized; Chances of being defrauded; Chances of being assaulted; Chances of being raped; Chances of money stolen at work; Chances of a hit & run accident; Chances of being injured by drunk driver; Chances of theft from respondent’s car.

8. Income categories as follows (1 = No income; 2 = Up to $5,000; 3 = Up to $8,000; 4 = Up to $12,000; 5 = Up to $18,000; 6 = Up to $25,000; 7 = Up to $35,000; 8 = Up to $45,000; 9 = Up to $45,000; 10 = Over $45,000). Education categories as follows: (1 = Less than high school; 2 = High school or equivalent; 3 = Some college; 4 = BA/BS; 5 = MA; PhD; Law or Medical school.

9. At the time of the survey, Hispanic was considered an exclusive race category. Categories were mutually exclusive without the option of choosing Hispanic as ethnicity along with an additional race category. Respondents who chose “other” as a race category were dropped from the analysis due to their small sample size (N = 27).

Additional information

Notes on contributors

Erin R. Coleman

Erin R. Coleman is a PhD candidate in the Department of Sociology and Criminology at the University of New Mexico. Her interests are in social movements, criminology, political sociology, and sociology of law. Her current project is a longitudinal examination of the relationships between political mobilization, crime, and police behavior.