Neighborhood Effects and Social Networks Research Paper

This sample Neighborhood Effects and Social Networks Research Paper is published for educational and informational purposes only. If you need help writing your assignment, please use our research paper writing service and buy a paper on any topic at affordable price. Also check our tips on how to write a research paper, see the lists of criminal justice research paper topics, and browse research paper examples.

Features of neighborhood social networks are central to many of the most prominent theories of crime. Empirical tests of a number of criminological theories underscore the importance of neighbor networks in both facilitating and deterring crime, both at the individual and community levels. However, most studies that focus on the link between neighborhoods and crime are limited by the use of “perceptual” and other proxy measures that aim to capture neighbor cohesion, social capital, and other social processes thought to mediate the association between neighborhood structural characteristics (e.g., concentrated disadvantage, residential instability) and offending. Exactly which features of neighborhood social organization are the most consequential for crime remains unknown due to the paucity of relational data that capture theoretically relevant features of actual neighborhood network structures. Furthermore, the mechanisms through which social networks influence neighborhood crime remain underexplored.

Despite the significant challenges associated with collecting neighborhood-based social network data, great strides have been made in the study of neighborhood networks in recent years. This research paper provides a brief overview of past research on neighborhood networks and crime and highlights notable recent advances in the conceptualization and measurement of neighborhood-based social networks in empirical research.

Whether explicit or implied, several of the most prominent criminological theories focus on the link between social networks and individual and aggregate offending patterns. For instance, theoretical constructs such as bonding, cohesion, social control, and peer influence are all central correlates of crime in criminological research (Papachristos 2010). These constructs in turn are commonly thought to represent emergent properties of the structure of affiliations among social groupings at various levels of aggregation (e.g., peer groups, neighborhoods, cities).

A number of macro-and multilevel theories also specifically focus on the direct and indirect effects of neighborhood and community networks on aggregate and individual rates of offending. Unfortunately, despite the centrality of social networks in criminological theory and empirical research, criminologists rarely employ social network analysis in research, especially when compared to researchers from other disciplines (e.g., public health and sociology; see Papachristos 2010). The use of social network analysis in community and neighborhood studies of crime occurs even less frequently. This neglect is quite noteworthy considering social processes related to neighborhood-based affiliations are central in explanations of neighborhood “effects” on crime (Sampson et al. 2002).

The disjuncture between the conceptual significance of social networks and the relatively infrequent use of network data and analysis in research on crime highlights the need for more effective integration of social network analysis into the criminological toolkit. Initial efforts toward this end have resulted in significant advances in the articulation and testing of classical and contemporary theories of crime. For instance, researchers focusing on “peer effects” (Haynie 2001) and gangs (Papachristos 2009; Radil et al. 2010) are increasingly employing network designs to test and expand upon criminological theories of diffusion, group conflict, and influence. Research on the implications of neighborhood social networks has also integrated more conceptually sophisticated social network insights into their explanations of criminal behavior. Despite the inherent difficulty of collecting data on neighborhood social networks, neighborhood scholars are developing innovative methods of data collection and analysis that may be used to study the association between neighborhood network structures and crime.

This research paper presents an overview of recent advancements in the study of neighborhood social networks and crime focusing on neighborhood networks as they pertain to the dominant macro-and multilevel theories of crime and community.

This research paper proceeds as follows. First, a brief overview of classical and contemporary theories of neighborhoods and crime is presented, highlighting the theoretical associations between the structure of social relations and crime.

Attending to these theoretical associations will help identify which properties of social networks are most consequential to crime. This research paper then highlights the unique properties of network data while focusing on the potential application of social network analysis to the study of neighborhoods and crime. Finally, this research paper describes recent studies that have employed innovative social network approaches to understanding the association between neighborhoods and crime.

Neighborhood Networks In Theory: From Yesterday To Today

Durkheim (1951) provides one of the first attempts to empirically examine the association between the structure of social relations and aggregate social problems. Durkheim proposed high levels of communal social integration promote social control over individual members, constraining individuals’ natural tendency to obviate social rules in the pursuit of self-interested ends. Thus, strong affiliations among members control crime and maintain social order through social control processes.

Early Chicago School scholars such as Wirth (1938) suggested increasing urban development and division of labor has deleterious effects on the community structure and solidarity characteristic of nonindustrial societies. The linear model of development has been characterized as a prototypical example of the “community lost” paradigm, which suggests pervasive division of labor in society disrupts solidary communities, leading to the formation of fragmented networks consisting of “impersonal, transitory, and segmental ties” (Wirth 1938, p. 12). Such patterns of community organization are thought to hinder neighborhood residents’ capacity to collectively organize to deal with social problems (e.g., crime and delinquency) within local communities.

Around the same time, Shaw and McKay (1942) proposed their theoretical model of community social organizational influences on crime. In this view, neighborhood structural characteristics, such as concentrated poverty, racial and ethnic heterogeneity, and residential stability, determine the extent of social interaction among neighbors and their corresponding capacity to maintain informal social control over local youth, with consequences for the prevalence of crime in the community. Building on the work of Shaw and McKay, Kornhauser (1978) developed a community control model of neighborhood social disorganization, which more precisely articulates how concentrated disadvantage, population turnover, and racial/ethnic heterogeneity shape consequential neighborhood social networks. For instance, residential instability, or the frequency with which residents move in and out in a neighborhood, is thought to limit opportunities to form trusting relations among neighbors. Racial/ethnic heterogeneity is hypothesized to limit interaction among members of different racial groups, which hinders the formation of “weak ties” (Granovetter 1973) within neighborhood settings. One potential outcome of social and cultural heterogeneity is a fragmented neighborhood network structure that hinders the collective capacity to maintain informal social control. This systemic model of crime thus proposes exogenous neighborhood structural characteristics indirectly lead to crime through shaping neighbor networks, which determine more proximate causes of crime, such as informal social control.

While Kornhauser extracted a pure control model from Shaw and McKay’s theory of social disorganization, other theorists (Cloward and Ohlin 1960) propose that the criminological consequences of community social organization are dependent upon the types of individuals and cultural models that are present in the community. From this perspective, deviant “value systems” represent emergent properties of neighborhood social organization. Accordingly, structures of social ties among neighborhood residents take on additional criminological consequences by fostering the formation and maintenance of delinquent subcultures conducive to crime.

From an individual standpoint, neighborhood contexts are important in shaping peer associations, especially among youth and adolescents. Peer delinquency in turn is one of the most robust predictors of offending (Haynie 2001). Different features of neighborhood social organization also more directly influence individual offending. For instance, routine activity perspectives suggest adolescent offending is in large part a function of the time adolescents spend engaging in unstructured socializing with peers in the absence of adult authority figures (Osgood et al. 1996). Other prominent theories, such as collective efficacy theory (Sampson et al. 1997), also point to the importance of intergenerational closure (i.e., adult embeddedness in neighborhood social networks) in promoting pro-social behavioral norms, which in turn fosters adolescent development and desistence from crime.

Finally, studies on the criminological consequences of residential segregation point to the importance of intra-neighborhood social ties in determining neighborhoodand city-level crime rates. As Krivo et al. (2009, p. 1792) suggest, “city segregation contributes to neighborhood violent crime indirectly, through the way it produces isolation and structural disadvantage in predominantly minority areas, and directly, by making it difficult for separate and unequal groups to work together to foster common goals and solve shared problems.” From a network perspective, racial segregation hinders social integration across neighborhoods, which in turn obstructs the formation of ties between political institutions and community investors and members of disadvantaged and isolated communities. Low levels of city-level social integration limit intergroup contact, which has deleterious effects on efforts at collective organizing to realize both the formal and informal controls over crime (Hunter 1985).

Empirical Tests Of Neighborhood Theories Of Crime

The aforementioned theoretical approaches present different, and at times, competing mechanisms with regard to the role neighborhood social networks play in shaping individual and aggregate rates of offending. While empirical support mounts for a number of these criminological theories, to date, few studies have tested these theoretical models with social network data.

The vast majority of studies focusing on neighborhood networks and crime instead utilize “perceptual measures” to capture aspects of neighborhood social networks. Most typically, researchers ask neighborhood residents a series of questions regarding perceptions of neighborhood social organization (e.g., intergenerational closure, friendship ties) or the extent of neighbor interactions (e.g., see Bellair and Browning 2010). Unfortunately, such measures are unable to capture features of the structure of neighborhood social networks (e.g., cohesion, network closure) that are most consequential for crime in the aforementioned theories. As few studies have attempted to directly measure the structure of social ties at the neighborhood level, “the structural features of locally based social networks have been inferred from the reports of individual respondents, measured indirectly, or simply presumed to operate” (Entwisle et al. 2007, pp. 1500–1501). As demonstrated below, social network data (i.e., actual relational data linking nodes in a network) better capture the social phenomena that are at the heart of many theories of community and crime.

The dearth of research that employs social network designs in the study of neighborhoods and crime is in no small part due to the significant (but not insurmountable) challenges associated with collecting data on neighborhood networks. In the following section, a brief overview of social network analysis is provided, highlighting the unique features of social network data. At the same time, the inherent difficulties associated with collecting network data across several neighborhoods that is necessary for the statistical analysis of “neighborhood effects” are discussed. However, this section also presents examples of certain forms of social network data (e.g., affiliation or two-mode network data) and network analytical techniques (e.g., network-based simulations) that may be used to enrich the understanding of the association between neighborhood networks and crime.

Some Unique Properties Of Social Network Data

From a theoretical standpoint, social network analysts stress the importance of interdependence among actors (e.g., individuals, institutions, nations) to explain both individual and collective outcomes. From this standpoint, individual and collective behaviors represent emergent properties of identifiable patterns of ties between actors and institutions and other units of interaction (Papachristos 2010). At the same time, the processes that explain variation in the structural characteristics of social networks (e.g., density, clustering) are also of interest to sociologists (Entwisle et al. 2007) and criminologists (Young 2011). In this latter case, social networks themselves serve as dependent variables. As the effects of neighborhood networks on crime itself are of primary interest in this research paper, networks as dependent variables are not discussed in this research paper.

While an extensive review of the various types of social network data is beyond the scope of this research paper, this section provides a brief overview of the constitutive elements of network data that are most relevant to research on neighborhood networks.

All forms of social network data consist of actors, which may be conceptualized as individuals, organizations, or any other unit of analysis that are thought be related to one another, and ties, which refer to specific types of relationships that link actors within the population of interest. Just as actors may be defined in any number of ways, there are several different types of ties that link actors. One of the most common types of tie that is studied among criminologists is friendship. Friendship networks are relevant to a number of criminological questions, such as, “Is adolescent delinquency related to the delinquency of one’s friends?” and “Do dense neighborhood-based friendship networks protect against crime?”

Ties need not be limited to those capturing friendship relations. Indeed, a number of different types of ties have also been considered in criminological research. For example, Papachristos (2009) used official police records to reconstruct a network consisting of Chicago gangs. In this network, two gangs were linked through a directed tie if a member of one gang killed a member of another gang. Schaefer (2012) used official police data to construct a neighborhood-based co-offending network in Maricopa County, AZ. In this network, two census tracts were linked if juvenile delinquents from two different tracts were arrested during the commission of the same crime. An emerging approach to studying neighborhood effects on delinquency and adolescent development focuses on overlap in “routine activity spaces” among neighbors to measure network structural characteristics of neighborhoods. Integrating insights from activity space approaches (which is given more attention near the end of this research paper) and social network analysis, an emerging approach to measuring neighbor networks focuses on residents’ overlap in routine activities. Demonstrating the utility of this approach, Browning and colleagues (2011) recently used data from Los Angeles Family and Neighborhood study (LAFANS) to construct 65 “actor-location” neighborhood networks consisting of individuals tied to block groups in which residents’ routine activities took place. Findings indicate various structural features of actor-setting networks (e.g., density, clustering) vary across the census tracts in the study. The exact types of tie that can link actors in a network that are considered are determined by particular research questions.

Ties may be further distinguished by their directionality. In network terminology, directed ties are referred to as arcs, while ties that are undirected are referred to as edges. Such a distinction is important, as some relationships are inherently directional, as in the case of resource exchange between actors. Other ties may not entail directionality. One prime example of non-directed ties is ties between spouses.

Social network data are most often collected through the use of two primary research designs. Ego network data most typically focus on the direct ties and associates of an individual “focal” node. Ego network data are typically collected through random samples by asking subjects to identify all, or a subset, of the actors to whom they have some form of connection (e.g., friendship, economic exchanges). Respondents then report any existing ties among the identified “alters” or actors to whom the focal respondent is tied. However, ego network data cannot typically be used to measure aggregate features of social networks, which are of primary interest to neighborhood theories of crime. However, through the use of snowball sampling techniques (Sampson and Graif 2009) and respondent-driven sampling (Papachristos et al. 2012), researchers have constructed relatively complete network data from ego network data.

Whereas ego network designs focus on the structure of ties that surround single actors, full or complete network designs collect information regarding every actor’s ties within a population or multiple populations of actors. For instance, the groundbreaking National Longitudinal Study of Adolescent Health or Add Health attempted to collect information on every adolescents’ school-based friendship ties within several middle and high schools in the United States. Complete network data allow for the examination of a number of topics of criminological interest. For instance, as with ego network data, complete network data allow for the examination of the association between peer delinquency and individual offending. However, complete network data also allow researchers to examine how features of the larger network (e.g., cohesion) and individuals’ positions within larger networks (e.g., centrality) influence offending. For instance, Haynie (2001) found that peer delinquency is more strongly associated with individual offending among adolescents who are tied to others who are also well connected in school network.

Network data are also comprised of radically different data structures. For example, bipartite networks (also known as affiliation or two-mode networks) consist of two different sets of actors that can only be linked to members of the opposite set. Browning and colleagues’ (2011) actorlocation network represents a bipartite network, given neighborhood residents (the first node set) are indirectly linked through sharing common activity spaces (the second node set). The data structure of bipartite networks differs from that of one-mode networks, which consist of single sets of actors and the direct ties between them. While one-mode networks are more commonplace in network research, as this research paper will illustrate, certain features of bipartite network data open up new and exciting avenues for future research on neighborhood networks.

Network data capture simplified versions of the social relations between actors. However, in reality, actors are tied to one another in any number of different ways. For example, students within a school network may be tied to one another through friendships, romantic relationships, overlap in their courses, participation in extracurricular activities, neighborhood residence, or bullying relations. Multiplex networks, consisting of actors who are connected by two or more different types of relations, explicitly take into account the different ways in which actors are connected to one another. Importantly, multiplex networks allow for the simultaneous examination of how different types of ties overlap and create specific relational structures within networks. While most techniques of network analysis are designed for networks consisting of one type of tie, network analysts are increasingly focusing on designing methods for multiplex networks.

In sum, social network data provide unique opportunities to examine the influence of social relations on individual and collective outcomes. This property of network data is particularly attractive to researchers focused on the association between neighborhoods and crime, as the structure of social relationships among neighbors is fundamental to all of the prominent theories of neighborhoods crime.

Unfortunately, there are a number of methodological barriers to gathering network data among neighbors. For example, constructing a one-mode network consisting of friendships among neighbors within just one neighborhood would require every resident to identify each of their friends from a neighborhood roster. The likely nontrivial amount of survey nonresponse in such a study would impede reliable estimation of structural properties of the neighborhood network.

Another problem with measuring neighborhood networks with survey data stems from the “boundary specification” problem that afflicts both social network research and conventional (i.e., nonnetworked) research on neighborhoods and crime. With regard to the latter, a number of different levels of community aggregation have been considered in research on neighborhoods and crime. For example, studies have (in order of their typical size in area) examined crime and disorder at the census block level, the block group level (which in the United States typically consists of 39 blocks), the census tract level, neighborhood cluster level (i.e., two or three contiguous census tracts), and zip codes. There are two significant limitations related to all of these forms of aggregation. First, certain neighborhood boundaries are based on administrative definitions, as in the cases of census tracts and zip codes. Despite the plethora of research that identifies associations between the characteristics of neighborhoods, as defined by administrated boundaries, and individual and aggregate outcomes (see Sampson et al. 2002 for a review), there is no guarantee that administrative boundaries encompass the actual “neighborhoods” that neighborhood researchers aim to study. Second, the association between neighborhood characteristics and crime has been shown to depend on the level of aggregation. For example, Hipp (2007) demonstrates that economic resources are most strongly associated with neighborhood crime and disorder among neighborhoods that are defined at the block level, while racial heterogeneity is most strongly associated with crime and disorder when measured at the census tract level.

A similar problem of boundary specification exists in social network research. In social network analysis, the issue of boundary specification pertains in large part to both the inclusion of certain actors and types of affiliations. With regard to the latter, a researcher must decide upon which types of ties are to be measured. Decisions regarding the measurement of ties are most often driven by research questions prior to execution of the study. Conversely, identifying which actors comprise a network in question takes considerably more thought in certain situations. As Laumann et al. (1983, p.18) explain:

because individual behavior is viewed as at least partially contingent on the nature of an actor’s social relationships to certain key others.. .care must be given to specifying rules of inclusion. Such rules pertain both to the selection of actors or notes for the network and to the choice of types of relationships among those actors to be studied.

Accordingly, much thought must be put into identifying both the actors as types of ties that are to be included in the empirical network, as noninclusion of relevant actors may have severe consequences for the measurement of the structural properties of networks.

While researchers employing network methods and concepts continue to make novel contributions to the understanding of crime, difficulties associated with collecting network data among neighborhood samples have contributed to the relative paucity of network-based studies of neighborhood effects. In the following section, four recent alternatives to studying neighborhood networks are discussed. Each takes a distinct approach to the conceptualization and measurement of neighborhood networks. It is our contention that these approaches hold promise for advancing the understanding of how neighborhood-based social network processes shape individual and aggregate offending patterns.

New Directions

Despite the challenges facing investigation of network effects on crime, emerging directions in the conceptualization and measurement of neighborhood networks offer the potential to significantly advance the understanding neighborhood effects on crime. The remainder of this research paper focuses on recent studies that have attempted to capture the structure of social relations among residents and other important actors (e.g., community organizers) in neighborhood networks. While to date, no known published studies stemming from these projects have applied social network data to examine the association between neighborhood or community networks and crime itself, studies have focused on processes that are thought to mediate the association between neighborhood networks and crime, such as trust, cohesion, and informal social control. The described approaches attempt to capture consequential neighborhood networks through innovative efforts to reconstruct the pattern of actual ties among actors (Sampson and Graif 2009) or by integrating spatial and social network information to arrive at alternative conceptualizations and measures of neighborhood networks (Butts et al. 2012; Radil et al. 2010; Browning et al. 2011).

A large share of recent research on neighborhood effects on crime has utilized data from the Project on Human Development in Chicago Neighborhoods (PHDCN). A lesser-known component of the PHDCN, the key informant (KI) study, utilized a snowball sampling design to collect information on dimensions of relationships among community leaders across religious, educational, business, law enforcement, political, and community organization domains within several of Chicago’s community areas. As part of initial interviews, key informants were asked to identify up to five people they went to in order to “get things done” in the community. From these nominations, Sampson and Graif (2009) constructed community leadership networks for a subsample of the community areas, measured a number of their structural properties (e.g., density, leadership inequality, and path distance). The authors also found that network centralization in the leadership network captures the extent to which nominations from all community leaders are sent to a few key leaders to be positively associated with community residents’ trust in law enforcement and trusting relations among community leaders. The authors suggest that the presence of highly central leaders may foster working trust among leaders and residents by making leaders more visible and accountable to the community. Thus, leaders and community residents may be better able to exercise social control over central leaders, creating a multilayered “enforceable trust” (Sampson and Graif 2009, p. 207) within the neighborhood network.

A number of recent studies have integrated spatial and social network approaches to arrive at theoretically relevant conceptualizations and measures of neighborhood networks. For instance, Radil et al. (2010) build on insights from geography and network perspectives to highlight the roles that the geographic distribution of gangs and the network structure of gang rivalries play in shaping patterns of gang violence within a larger community. The authors suggest that space is socially constructed in large part through processes related to geographic and network embeddedness of actors. One important feature of network embeddedness is spatialized network structural equivalence, which captures the extent to which two or more geographic areas are similarly situated in terms of positions within “spatialized” social networks.

Relying on information from gang informants and the Los Angeles Police Department, the authors reconstructed a spatialized community network for which actors represent collections of census block groups that comprise gang territories. Gang territories are in turn tied to one another through rivalries with other local gangs. Through a network analytic technique called convergence of iterated correlations (CONCOR), the authors identify geographic subsets that are comprised of gang territories that are similarly situated within the rivalry network. Upon mapping the territorial subsets, the authors demonstrate that the geographic distribution of the gang territories has a distinct spatial pattern and that the spatialized patterning is meaningfully related to the distribution of gang violence across the larger community. Results from Radil et al. (2010) study illustrate the importance of the structure of social relations among actors across different geographic spaces in shaping crime rates across different communities.

Butts and colleagues (2012) present an alternative, simulation-based approach to measuring community networks that relies on assumptions about the spatial organization of social ties. The authors ground their analysis on the well-established notion that the probability of a social tie between individuals increases with their geographic proximity to one another (McPherson et al. 2001). Using data on geographic distributions of individuals from the US Census, Butts and colleagues constructed simulated networks for several metropolitan (i.e., an area that includes at least one city with a population greater than 50,000 individuals) and micropolitan (areas that contain at least one city with a population between 10,000 and 50,000) areas in the USA. Through specifying and applying a spatial interaction function (SIF), which captures the “marginal probability of a tie between two randomly selected individuals at some given distance” (Butts et al. 2012, p. 83), to data on geographic location of actors, the authors constructed simulated metropolitan and micropolitan networks comprised of friendship relations and face-to-face interactions. Butts and colleagues also identified associations between the spatial distribution of actors across the units of aggregation and various structural properties of the networks. This simulation process can also incorporate several other nongeographic processes that are known to influence tie formation, such as selective mixing on race, age, and socioeconomic status (McPherson et al. 2001) to potentially simulate networks that more accurately reflect actual networks of distinct areas. While this innovative method holds much promise for understanding neighborhood network processes and crime, future research that compares the simulated networks with actual networks constructed from network-based studies will help to assess how well-simulated networks reflect neighborhood social networks.

One of the most attractive qualities of the network simulation approach of Butts and colleagues is that it enables researchers to construct social networks with readily available data across several cities and neighborhoods. This is particularly useful, as collecting data on every individual’s ties across a single, let alone, several neighborhoods through conventional survey approaches is a daunting, if not impossible, task.

However, individuals may be connected to other persons through any number of ways. For instance, neighborhood residents may be indirectly tied to one another through shared participation in activity spaces. Activity spaces may be understood to encompass all of the locations individuals come into contact with as a result of their routine activities (Golledge and Stimson 1997). Incorporating activity space data into a social network analytic framework may help address the challenges facing research on neighborhood networks and crime in two primary ways. First, activity space data enable the measurement of individual exposure to specific places and settings that vary in their criminogenic properties (i.e., extent of disorder and crime). Activity space data can also capture exposure to certain types of individuals and allow for the measurement of the extent of overlap in individuals’ routine activities. Finally, as illustrated below, network structural characteristics describing patterns of overlapping routine activities among neighborhood residents are likely important in the functioning of neighborhood social processes that are relevant to crime (e.g., collective efficacy, individuals’ access to social capital, participation in neighborhood organizations).

Apart from capturing individuals’ place-based exposures, sampling activity spaces among geographically contained populations may allow investigators to measure the extent to which neighborhood residents share activity locations within given boundaries or larger areas (e.g., entire metropolitan areas). Consistent with Jacobs (1961), the structural patterns of activity overlap may capture important features of the social organization of a neighborhood. For instance, “co-location networks” may impact a number of neighborhood processes such as social network formation as well as trust and shared expectations for informal social control. While these factors have been linked to individual offending and neighborhood crime rates, the social and ecological dynamics that are associated with such features of neighborhood organization remain elusive. Moreover, by measuring the co-location network more precisely, researchers may apply a continuously developing catalogue of network analytic techniques to characterize routine activity patterns using sophisticated global network measures (e.g., density) and characterize individuals’ locations in the network (e.g., centrality). Additionally, through the use of panel and other longitudinal study designs, one may examine how co-location networks evolve over time with stochastic actor models (Ripley et al. 2012).

Browning and colleagues (2011) recently examined the association between structural features of actor-location networks among neighborhoods in Los Angeles and a number central processes in neighborhood-based studies of crime and delinquency. The authors find that triadic closure, or “clustering” (Opsahl Forthcoming), in the actor-setting network is positively associated with access to social capital as well as perceptions of network exchange (i.e., frequency of exchange of favors among neighbors), intergenerational closure among neighborhood residents, and collective efficacy. Future research that builds upon the approach of Browning and colleagues may help further identify the association between structural features of neighborhood actor-setting networks and offending patterns.

Conclusion

The primary objective of this research paper was to provide an overview of network approaches to studying the association between neighborhood social organization and crime. Importantly, the approaches highlighted in this research paper were situated within prominent theoretical approaches to neighborhood networks and crime. After outlining central network processes of neighborhood-based theories of crime, this research paper highlighted the limitations of “perceptual” measures of community organization that are most commonly used to assess structural features of neighborhood networks. The limitations of perpetual measures were further illustrated by highlighting how relational properties of social network data allow for a more complete understanding of neighborhood social organization as it relates to individual and aggregate offending patterns.

Most importantly, this research paper detailed emerging approaches to measuring structural properties of neighborhood social networks. Through the use of network simulations, snowball sampling, spatialized gang networks, and two-mode actorlocation affiliation network data, recent studies have made great strides in measuring structural properties of neighborhood and community networks. Perhaps as importantly, these techniques circumvent a number of the inherent difficulties and cost restrictions of collecting neighborhood-based social network data. Future research that builds on these and similar network analytic techniques may advance the understanding of the association between neighborhood social organization and offending rates by identifying which structural properties of neighborhood networks (e.g., network density, closure) are most strongly associated with crime. In addition, considering different types of ties (e.g., friendship tie versus gang rivalry) may further demonstrate the criminological consequences of differential social organization and advance the understanding of the neighborhood context of crime.

Bibliography:

Bellair PE, Browning CR (2010) Contemporary disorganization research: an assessment and further test of the systemic model of neighborhood crime. J Res Crime Del 47:496–521
Browning CR, Jackson AL, Soller B, Krivo L, Peterson RD (2011) Ecological community and neighborhood social organization. Paper presented at the annual meeting of the American Society of Criminology, Washington, DC, 16–19 Nov 2011
Butts CT, Acton RM, Hipp JR, Nagle NN (2012) Geographical variability and network structure. Soc Networks 34:82–100
Cloward RA, Ohlin LE (1960) Delinquency and opportunity: a theory of delinquent gangs. Free Press, Glencoe
Durkheim E (1951) Suicide: a study in sociology. Free Press, New York
Entwisle B, Faust K, Rindfuss RR, Kaneda T (2007) Networks and contexts: variation in the structure of social ties. Am J Sociol 112:1495–1533
Golledge RG, Stimson RJ (1997) Spatial behavior: a geographic perspective. The Guilford Press, New York
Granovetter M (1973) The strength of weak ties. Am J Sociol 78:1360–1380
Haynie DL (2001) Delinquent peers revisited: does network structure matter? Am J Sociol 106:1013–1057
Hipp JR (2007) Block, tract, and levels of aggregation: neighborhood structure and crime and disorder as a case in point. Am Sociol Rev 72:659–680
Hunter A (1985) Private, parochial and public social orders: the problem of crime and incivility in urban communities. In: Zald M, Suttles (eds) The challenge of social control: citizenship and institution building in modern society. Ablex Publishing, Norwood, pp 230–242
Jacobs J (1961) The death and life of great American cities. Random House, New York
Kornhauser R (1978) Social sources of delinquency. University of Chicago Press, Chicago
Kossinets G (2006) Effects of missing data in social net- works. Soc Networks 28:247–268
Krivo LJ, Peterson RD, Kuhl DC (2009) Segregation, racial structure, and neighborhood violent crime. Am J Sociol 114:1765–1802
Laumann EO, Marsden PV, Prensky D (1983) The boundary specification problem in network analysis. In: Burt RS, Minor MJ (eds) Applied network analysis: a methodological introduction. Sage Publications, London, pp 18–34
McPherson M, Smith-Lovin L, Cook JM (2001) Birds of a feather: homophily in social networks. Annu Rev Sociol 27:415–444
Opsahl T (Forthcoming) Triadic closure in two-mode networks: redefining the global and local clustering coefficients. Soc Networks
Osgood DW, Wilson JK, O’Malley PM, Bachman JG, Johnston LD (1996) Routine activities and individual deviant behavior. Am Sociol Rev 61:635–655
Papachristos AV (2009) Murder by structure: dominance relations and the social structure of gang homicide. Am J Sociol 115:74–128
Papachristos AV (2010) The coming of a networked criminology. In: MacDonald J (ed) Measuring crime and criminality: advances in criminological theory, vol 17. Transaction Publishers, New Brunswick, pp 101–140
Papachristos AV, Braga AA, Hureau DM (2012) Social networks and the risk of gunshot injury. J Urban Health 89:992–1003
Radil SM, Flint C, Tita GE (2010) Spatializing social networks: using social network analysis to investigate geographies of gang rivalry, territoriality, and violence in Los Angeles. Ann Assoc Am Geogr 100:307–326
Ripley, Ruth M, Snijders, Tom AB, Preciado P (2012) Manual for SIENA version 4.0. University of Oxford, Department of Statistics; Nuffield College, Oxford, http://www.stats.ox.ac.uk/ snijders/siena/
Sampson RJ, Graif C (2009) Neighborhood networks and processes of trust. In: Levi M, Hardin R, Cook KS (eds) Whom can we trust? How groups, networks, and institutions make trust possible. Russell Sage Foundation, New York, pp 182–215
Sampson RJ, Raudenbush SW, Earls F (1997) Neighborhoods and violent crime: a multilevel study of collective efficacy. Science 2778:918–924
Sampson RJ, Morenoff JD, Gannon-Rowley T (2002) Assessing ‘neighborhood effects’: social processes and new directions in research. Annu Rev Sociol 28:443–478
Schaefer DR (2012) Youth co-offending networks: an investigation of social and spatial effects. Soc Networks 34:141–149
Shaw C, McKay HD (1942) Juvenile delinquency and urban areas. University of Chicago Press, Chicago
Wirth L (1938) Urbanism as a way of life. Am J Sociol 44:1–24
Young J (2011) How do they ‘end up together’? A social network analysis of self-control, homophily, and adolescent relationships. J Quant Criminol 27: 251–273