Psychology by BogdanovaThe Encyclopedia of Psychology for the Artificial Era

RESEARCH DOSSIER

Collective Efficacy and Informal Social Control

Abstract. Collective efficacy describes residents' shared expectations that neighbors can coordinate to maintain valued conditions, while informal social control refers to everyday responses to local problems through social relationships, communication, monitoring, mutual support, and intervention. This fundamental research monograph examines the conceptual formation of these constructs, their roots in social-disorganization theory, the 1995 Chicago neighborhood investigations, multilevel ecometric measurement, and comparative studies in Australia and the Netherlands. It reviews mechanisms connecting cohesion, structural inequality, informal regulation, violence, fear, public disorder, spatial dependence, and neighborhood stigma. A critical appraisal distinguishes statistical neighborhood associations from causal effects, separates generalized trust from task-specific capacities, and evaluates measurement validity, spatial interference, reciprocal causation, institutional conditions, and distributional fairness. The synthesis finds a substantial empirical tradition linking collective efficacy to violence-related outcomes while identifying limits to causal identification and uniform cross-context generalization. It concludes with methods for stronger longitudinal and multimethod tests and specifies novel hypotheses concerning digital neighborhood coordination and artificial systems. This work is a critical narrative research synthesis of published research, not an original empirical investigation, preregistered systematic review, or meta-analysis.

Keywords: collective efficacy; informal social control; neighborhood cohesion; social disorganization; community psychology; neighborhood effects; community organization; ecometrics; neighborhood trust; social networks; violent crime; public disorder; social capital; task-specific efficacy; spatial dependence; residential instability; neighborhood stigma; causal inference; social inequality; community intervention; digital coordination; Artificial Era

Scientific Object, Definition, and Research Questions

Collective efficacy is a neighborhood-level capacity constituted by residents' shared expectations that neighbors can and will act together to maintain conditions valued by the community. In its influential urban-neighborhood formulation, the construct combines social cohesion and trust with expectations for informal social control. Informal social control concerns residents' willingness and capacity to respond to locally consequential conduct through ordinary social relationships, communication, mutual support, monitoring, and intervention outside formal law-enforcement authority. The relevant unit is a relational and territorial setting; an individual's favorable judgment of neighbors supplies evidence about that setting only after the researcher demonstrates that reports can be aggregated into a credible contextual measure Sampson et al. (1997) Raudenbush & Sampson (1999).

The research problem is to explain why equally situated individuals experience different levels of locally coordinated protection, supervision, and public-order maintenance, and how these capacities are produced or depleted by residential histories, social networks, institutional access, inequality, and violence. Crucial questions concern which social resources make action feasible; whether informal-control expectations predict conduct; how cohesion differs from shared intervention norms; whether collective efficacy affects violent victimization; and how disorder, crime, reputation, and community intervention influence one another over time Sampson & Groves (1989) Lanfear et al. (2020).

This monograph concentrates on neighborhood and community processes that connect efficacy beliefs to informal regulation, with particular attention to public-space safety, young people's everyday social environments, and collective responses to local problems. It considers adjacent work on social capital, civic mobilization, collective agency, and political efficacy when these distinctions clarify its subject. The synthesis is a critical narrative review of foundational books, primary observational studies, community surveys, ecometric methods, comparative field research, and causal-methodology assessments. It contains no newly collected observations, preregistered systematic search, pooled meta-analysis, or claim of experimental proof Wickes et al. (2013) Sampson (2012).

Conceptual Anatomy of Collective Efficacy

The term joins two components that must be specified separately. Cohesion denotes accessible, reliable relationships, shared trust, and a readiness to recognize other residents as legitimate partners in ordinary community life. Expectations of informal control denote beliefs that neighbors will respond to defined events, such as children skipping school, disturbances outside a home, or threats to a shared public space. A neighborhood may contain many acquaintances without a widely shared expectation of intervention; conversely, residents may coordinate around a specific problem despite limited personal intimacy. The construct is thus anchored in anticipated joint capacity as well as social relations Sampson et al. (1997) Wickes et al. (2013).

Efficacy statements must specify an object and a task: efficacy for what, by whom, on whose behalf, and under which circumstances? A survey question about whether neighbors will stop interpersonal violence measures a different intervention expectation from a question about petitioning for city services or collectively caring for children. The Australian community study of Wickes, Hipp, Sargeant, and Homel argues empirically for task-specific measures distinct from generic social ties and cohesion. Their analysis of 4,093 respondents in 148 communities supports treating locally perceived capacities for violence, delinquency, and civic problems as analytically separable Wickes et al. (2013).

Collective efficacy also depends on the quality of the proposed intervention. An observer can code a reported willingness to confront teenagers, but interpretation requires attention to proportionality, consent, prejudice, and the availability of safe alternatives. Informal regulation can operate through conversation, care, accompaniment, public reporting, or exclusionary pressure. Consequently, the same reported willingness to intervene may have different consequences across communities. Scientific use of the concept requires specifying beneficiaries, behavioral mechanisms, and the normative criteria by which an intervention is judged Sampson & Raudenbush (2004) Lanfear et al. (2020).

Distinct Constructs: Social Capital, Belonging, Networks, and Agency

Social capital names resources embedded in social connections, such as information, reciprocal obligations, organizational access, and opportunities for coordinated exchange. Collective efficacy concerns the activation of those resources in relation to locally recognized goals. Neighborhood connections may supply channels for communicating an incident, but shared expectations and credible commitment determine whether available channels support a response. This is one reason social-network density, social cohesion, and efficacy must be modeled separately rather than assumed to rise together Sampson et al. (1999) Browning et al. (2004).

A sense of community or belonging concerns a resident's perceived membership, connection, and shared emotional significance. It may facilitate involvement, yet belonging can be strong in communities that resist intervention or distribute protection unevenly. Similarly, personal self-efficacy refers to judgments about one's own capacity to act, while collective efficacy concerns a judgment about coordinated capacity. Bandura's wider social-cognitive treatment of collective efficacy provides a useful conceptual antecedent, although neighborhood collective efficacy became a distinct empirical tradition with specifically territorial measurement and social-control hypotheses Bandura (1997) Sampson et al. (1997).

Collective agency describes coordinated action actually undertaken. Efficacy beliefs may precede action, result from successful action, or be revised following failed interventions. Treating intentions, expectations, and performance as different variables helps prevent a conceptual shortcut from residents' survey answers to assertions about realized collective action. The governing distinction is between available resources, shared judgments of capacity, observable practices, and consequences. A complete study ideally measures all four and their change over time Raudenbush & Sampson (1999) Lanfear et al. (2020).

Origins in Urban Ecology and Social Disorganization

The historical background lies in Chicago School urban research and the ecological investigation of patterned delinquency. Clifford Shaw and Henry McKay's 1942 study, Juvenile Delinquency and Urban Areas, compared delinquency rates across local areas and linked durable geographic distributions to socioeconomic and institutional environments. Their argument directed attention to community organization and mechanisms connecting structural disadvantage to children's social conditions. The work remains historically pivotal, although its original classifications, administrative crime measures, and causal claims require interpretation in the context of subsequent methodological developments Shaw & McKay (1942).

Sampson and Groves reformulated social-disorganization theory as a testable model of intervening community structure. Their 1989 research analyzed local friendship ties, supervision of street-corner teenage groups, and organizational participation, using independent British survey materials from the early 1980s. The investigation connected socioeconomic status, instability, and family disruption to community-level mechanisms and then to crime and victimization. It supplied an essential bridge from ecological correlations to explicit, survey-measured regulatory processes, without resolving every question of selection, reverse causation, and differential reporting Sampson & Groves (1989).

The next step was to identify why some networks produce public-regarding intervention and others do not. Collective efficacy theory shifted attention toward mutual expectations, grounded in social trust but oriented toward action. Its major formulation in the 1997 Chicago study made the community process operational and subject to multilevel measurement. The change matters historically because it treats neighborhood organization as a measured mechanism linking place-based conditions to outcomes, rather than as an unspecified moral characterization of a neighborhood Sampson et al. (1997) Sampson (2012).

From Personal Efficacy to Neighborhood Expectations

Bandura's 1997 treatment of collective efficacy argued that beliefs about shared capability influence coordination, persistence, and responses to collective obstacles. Neighborhood research borrowed the insight that perceived capacity can be socially consequential, while defining a different empirical referent: residents assessing the likely conduct of other residents in a particular place. A general belief that people can cooperate does not necessarily measure expected intervention by identifiable neighbors facing a specified local challenge Bandura (1997) Wickes et al. (2013).

The psychologically important mechanism involves second-order expectations. An individual may believe an intervention is justified but anticipate that others will remain silent, creating a barrier to acting alone. Another individual may doubt the value of intervention but expect neighbors to provide assistance, changing the perceived feasibility and risk of participation. These mechanisms are plausible interpretations of shared efficacy expectations; the classic neighborhood studies primarily measured reported judgments and outcomes, rather than experimentally manipulating such beliefs at the individual level Sampson et al. (1997) Lanfear et al. (2020).

Investigators must distinguish endorsement of a public value from a prediction about neighbors' behavior. Asking whether children should be supervised measures a normative attitude; asking whether residents would intervene in a concrete situation measures an expectation of local regulatory action. Neither item alone captures institutional support, freedom from retaliation, trust in authorities, or the resident's own opportunity to act. A credible empirical theory therefore connects the belief measure to context, available pathways for intervention, and observable outcomes Raudenbush & Sampson (1999) Wickes et al. (2013).

The Chicago Collective Efficacy Study and Its Research Design

The landmark investigation by Sampson, Raudenbush, and Earls used the 1995 Project on Human Development in Chicago Neighborhoods community survey. Its analytical sample included 8,782 residents distributed across 343 neighborhood clusters. Respondents evaluated local cohesion and whether neighbors would intervene in circumstances threatening the common good. The authors modeled residential reports using a multilevel framework and examined neighborhood variation in several forms of violence, with adjustments for measured individual attributes, historical neighborhood crime, and selected structural conditions. The data established that a neighborhood-level efficacy measure could be constructed with substantial between-neighborhood reliability Sampson et al. (1997).

The reported association ran in the predicted direction: neighborhoods characterized by stronger collective efficacy tended to have lower levels of violence. The investigators also found that collective efficacy statistically accounted for an important share of observed associations between concentrated disadvantage, residential instability, and neighborhood violence. These findings supported a mediating-process interpretation. Because the focal neighborhood efficacy measures and major outcomes were observed rather than randomized, the results did not by themselves eliminate unmeasured community differences or establish temporal ordering across every pathway Sampson et al. (1997) Lanfear et al. (2020).

The study's historical importance lies in how its conceptual model, survey instrument, and ecometric statistical strategy fit together. Individual responses were treated as imperfect reports of shared contextual conditions; neighborhood associations were estimated after accounting for different respondent characteristics; and the explanatory claim concerned collective processes rather than moral traits of particular residents. Subsequent researchers could directly test transportability, measurement invariance, and the distinction between generic cohesion and task-specific intervention. The same foundation also made criticism possible where an alleged causal mechanism rested on perceptions rather than observed collective conduct Raudenbush & Sampson (1999) Wickes et al. (2013).

Ecometrics and the Measurement of Social Context

Ecometrics is the measurement of ecological settings by combining multiple items and multiple informants while estimating sources of sampling and response error. Raudenbush and Sampson's 1999 methodological account sets out why a neighborhood survey should not be treated as if its aggregated mean were an error-free property of place. Residents differ in experience and reporting tendencies; items differ in difficulty and content; neighborhoods differ in how many respondents are sampled. A multilevel model can separate those uncertainties and estimate a latent setting characteristic Raudenbush & Sampson (1999).

Collective-efficacy measurement typically asks about trust, mutual help, closeness, and expected intervention under specified scenarios. A common analytical pathway evaluates whether items form coherent components, whether within-neighborhood agreement justifies aggregation, and whether sufficient between-neighborhood variation remains after adjusting for the composition of respondents. Reliability of the resulting contextual score depends partly on respondents per neighborhood and the size of true differences among settings. A high internal-consistency coefficient for individual ratings alone cannot resolve whether a reliable neighborhood-level variable has been measured Sampson et al. (1997) Raudenbush & Sampson (1999).

An adequate contemporary design also tests whether item meanings travel across demographic groups, languages, housing conditions, and institutional contexts. For example, residents living on different blocks may use the same phrase about intervention while imagining different forms of conduct or different risks. Surveys of perceived control should be supplemented with measured opportunities to act, independent behavioral observations, administrative records, and where possible repeated assessments. The resulting construct validation is an ongoing task rather than a property conferred permanently by use of a well-known scale Wickes et al. (2013) Sampson & Raudenbush (2004).

Intergenerational Closure, Reciprocal Exchange, and Child Environments

Sampson, Morenoff, and Earls examined collective efficacy for children through intergenerational closure, reciprocal local exchange, and shared expectations for informal social control. Intergenerational closure describes the extent to which parents or other adults know each other and can communicate about children; reciprocal exchange concerns local practices of mutual assistance; intervention expectations concern readiness to respond to behavior affecting children's welfare. Their 1999 study analyzed Chicago survey data and modeled variation across 342 neighborhoods, considering spatial dependence between neighboring areas Sampson et al. (1999).

The investigation found that residential stability and concentrated affluence were especially relevant to intergenerational closure and exchange; structural circumstances shaped dimensions of social organization differently. This finding discourages collapsing every local social resource into one undifferentiated quantity. A resident's familiarity with nearby families can facilitate monitoring, but actual intervention also depends on perceived legitimacy, safety, ability to communicate, and prospects of reciprocal support. Parents, youth, and other adults may therefore describe the same neighborhood differently Sampson et al. (1999) Wickes et al. (2013).

Children's everyday environmental exposures connect community processes to developmental outcomes without establishing that any single intervention norm determines development. Adult oversight, school contact, street conditions, mobility, and opportunities for interaction overlap. Child-centered collective efficacy should be assessed against children's own experiences and outcomes, including the possibility that excessive surveillance or exclusion targets some young people more than others. Multi-informant designs involving youth, caregivers, and external observers can clarify whose protection is being measured Sampson et al. (1999) Sampson & Raudenbush (2004).

Theoretical Mechanisms of Informal Regulation

Informal regulation can proceed through information, opportunity, normative signaling, and practical intervention. Local ties make events more visible, while reciprocal expectations allow residents to interpret a request for help as credible and legitimate. Informal sanctions, persuasion, adult supervision, assistance to vulnerable residents, and coordinated petitions represent different behavioral channels. Each has a different opportunity structure, cost, and potential consequence. The shared underlying proposition is that residents anticipate some willingness to act together when locally consequential conduct occurs Sampson & Groves (1989) Sampson et al. (1997).

These channels should be connected to explicit behavioral predictions. Communication networks may reduce the delay between recognizing a local problem and mobilizing a response. Repeated successful cooperation may strengthen beliefs that future action will succeed. Threats of retaliation, interpersonal distrust, or inconsistent institutional support may inhibit participation even among residents who agree about a problem. These are mechanisms to investigate by measuring events and decisions in sequence, not established consequences inferred automatically from one cross-sectional correlation Bandura (1997) Lanfear et al. (2020).

An important unresolved question concerns how formal and informal systems interact. Informal response can involve helping an affected resident, contacting service agencies, speaking to an offender, or coordinating with local authorities. Effective neighborhood organization need not imply hostility toward institutions. Conversely, informal intervention can impose costs on marginalized groups when the operational definition of disorder reflects unequal enforcement or neighborhood stigma. Researchers must distinguish safety-enhancing cooperation from discriminatory regulation by examining targets, procedures, and independently measured outcomes Sampson & Raudenbush (2004) Sampson (2012).

Networks, Cohesion, and the Paradox of Organized High-Crime Neighborhoods

The density of relationships is an insufficient index of public-regarding neighborhood control. Browning, Dietz, and Feinberg's 2004 analysis of urban social organization asked why some areas contain extensive social networks while experiencing considerable violent crime. The study's conceptual contribution was to challenge a simple assumption that dense local relations always translate into collective efficacy or reduced violence. Connections may link residents to competing obligations, informal economies, and organizations that support particular interests without establishing widely shared expectations for public protection Browning et al. (2004).

Network ties can facilitate cooperation among residents pursuing different ends. Strong ties can protect some groups while limiting the ability of others to report threats. A neighborhood may be cohesive within subgroups and fractured across their boundaries. Survey indicators of generalized neighborly warmth may fail to detect these structures. A research design capable of investigating the paradox would map network composition and bridge ties, separately measure task-specific efficacy, and observe outcomes by social position, location, and type of event Browning et al. (2004) Wickes et al. (2013).

The empirical implication is that correlations between cohesion and crime should be disaggregated by the social function of ties. Researchers ought to estimate relationships among network density, mutual recognition, cooperation across boundaries, anticipated informal intervention, and harmful conduct. Such decomposition also helps interpret the difference between social integration and effective community regulation. The strongest inference available from the literature is conditional: cohesion is a potential enabling resource whose consequences depend on the goals, structure, and situational deployment of collective action Sampson et al. (1999) Browning et al. (2004).

Spatial Spillovers and the Geography of Violent Crime

Neighborhood exposure extends beyond an administrative boundary. Morenoff, Sampson, and Raudenbush (2001) examined spatial dynamics in Chicago and modeled how violent crime and neighborhood characteristics co-varied across local areas. Their framework treated spatial dependence as a substantive part of urban social organization: the conditions immediately surrounding an area may shape its opportunities for informal control and its risks of violence. Findings concerning collective efficacy were therefore assessed alongside disadvantage, residential stability, and neighboring contexts Morenoff et al. (2001).

Spatial autocorrelation presents both an empirical clue and a statistical danger. A simple regression assumes that observations are conditionally independent; bordering communities may share resources, transport corridors, policing patterns, housing markets, and violence exposure. If these dependencies remain unmodeled, uncertainty can be understated and a local contextual effect can be confused with an unmeasured regional process. Spatial lag and spatial error models, multilevel specifications, and flexible neighborhood definitions provide different ways to analyze these dependencies. The appropriate choice follows the hypothesized causal pathway rather than an automatic preference for one technique Morenoff et al. (2001) Raudenbush & Sampson (1999).

Physical distance is only one geography of social life. A school catchment can bridge adjacent census tracts; a housing complex can split one administrative tract into distinct relational settings; displaced residents may sustain connections across long distances. Neighborhood boundaries are therefore measurements as well as historical institutions. Repeated network surveys, travel-pattern data, and geocoded interaction episodes could test when collective efficacy is generated across contiguous places and when it follows nonterritorial networks. Any such extension should preserve the difference between the social unit the theory identifies and the convenient units available in a dataset Sampson (2012) Morenoff et al. (2001).

Public Disorder, Systematic Social Observation, and Social Control

Sampson and Raudenbush's 1999 program of systematic social observation documented public settings through standardized field protocols and video-recorded street segments in Chicago. It sought to measure visible physical and social disorder more independently than relying exclusively on residents' reports or police records. Their analysis investigated how disorder relates to neighborhood conditions and crime, identifying concentrated disadvantage and collective efficacy as central contextual considerations. The methodological innovation was to build observable indicators of street conditions that could be analyzed alongside survey-based neighborhood constructs Sampson & Raudenbush (1999).

Disorder indicators are not interchangeable with harm. Graffiti, litter, groups gathering outdoors, and observable confrontations differ in severity and social meaning. Some features index physical neglect; some may record ordinary use of public space; still others may signal immediate danger. Context, neighborhood history, observer assumptions, and time of day affect classification. A valid design specifies coding rules, tests interobserver agreement, and considers whether markers predict independently verified consequences rather than simply reproducing culturally dominant assumptions about acceptable public behavior Sampson & Raudenbush (1999) Sampson & Raudenbush (2004).

The causal problem is particularly acute when disorder and crime are recorded at the same time. Both might arise from structural disadvantage, institutional disinvestment, or low collective efficacy. Crime itself may undermine social trust, generating apparent disorder and fear. The 2020 causal review by Lanfear, Matsueda, and Beach compared these competing directions and concluded that the literature had reached little consensus and that most existing studies did not establish strong causal identification. The rational synthesis thus treats observed disorder as one variable within a dynamic neighborhood system, not an automatic experimental cause of crime Lanfear et al. (2020).

Perceived Disorder, Stigma, and Racialized Neighborhood Judgments

A neighborhood reputation may shape residents' and outsiders' judgments before they encounter a specific street. Sampson and Raudenbush (2004) studied the social construction of perceived disorder in Chicago, asking why people interpret a place as disorderly and how such perceptions relate to observed local conditions and racial and socioeconomic composition. Their findings expose an important measurement challenge: a report that a neighborhood is disorderly may convey information about its social reputation and the perceiver's interpretive assumptions as well as the physical environment Sampson & Raudenbush (2004).

This distinction affects collective efficacy because both predictor and outcome may be measured through perceptions. A resident who describes neighbors as untrustworthy may also rate local disorder as severe, producing shared-method covariance even when observers classify the relevant physical features differently. Conversely, widely repeated place-based stigmas may affect property investment, institutional service decisions, and residents' opportunities for coordinated action. Researchers need multiple observational vantage points, including systematic street audits, independent safety outcomes, and measures of racialized interpretations Sampson & Raudenbush (2004) Raudenbush & Sampson (1999).

Informal social control can be evaluated ethically and scientifically only by identifying who acts and who is acted upon. In some contexts, adult intervention protects children and vulnerable residents; in others, demands for conformity can become exclusionary pressure against racial minorities, young people, newcomers, or unhoused residents. A one-dimensional scale of willingness to intervene cannot determine whether a practice is fair. Disaggregated outcome data, residents' accounts, procedural indicators, and independent coding of conduct are needed to distinguish community protection from stigmatizing regulation Sampson & Raudenbush (2004) Wickes et al. (2013).

Comparative Evidence from Australian Neighborhoods

Mazerolle, Wickes, and McBroom (2010) studied violence across 82 Brisbane communities using survey information from 2,859 residents, official crime data, and census-based contextual measures. Their multilevel analysis separated social ties and collective efficacy and compared Australian results with earlier work in Chicago and Stockholm. The study reported patterns consistent with an association between collective efficacy and lower violent victimization across distinct urban contexts. Its value lies in comparing a theoretical claim across municipal settings with different institutions and settlement histories Mazerolle et al. (2010).

The Australian evidence supports transportability as a research question rather than as an assumption. Similar coefficient directions across cities suggest that the hypothesized process may travel, but survey items, neighborhood scales, and official reporting systems can differ. Administrative crime counts reflect detection and reporting, while victimization surveys depend on recall, willingness to disclose, and sampling. Explicit harmonization of the construct and outcomes is required before comparing effect magnitudes or claiming that one country's neighborhood process is identical to another's Mazerolle et al. (2010) Wickes et al. (2013).

Australian research also supplies a methodological correction to generic measurement. The task-specific study by Wickes and colleagues distinguished residents' capacities to respond to violence, delinquency, and civic or political problems, rather than treating collective efficacy as a single universal propensity. It demonstrated the scientific utility of separating social cohesion and social ties from context-specific intervention expectations. The deeper implication is that neighborhoods may possess differentiated action capacities, much as organizations have different competencies for separate tasks Wickes et al. (2013).

Contradictory and Conditional Findings in Dutch Neighborhood Research

Bruinsma, Pauwels, Weerman, and Bernasco (2013) assessed six models linking social disorganization, social capital, collective efficacy, crime, and offenders in a Dutch urban setting. Their research is important precisely because it tested related explanatory models in a context outside the classic American setting and assessed whether hypothesized paths captured observed variation. Findings were mixed and did not simply reproduce all expected neighborhood-level associations. The article complicates claims of a single, universally adequate causal chain from structural disadvantage through cohesion to reduced crime Bruinsma et al. (2013).

A model can fail to replicate for several scientifically different reasons. The social mechanism might vary across settings; sample sizes and neighborhood partitions may yield limited precision; administrative crime patterns may reflect different enforcement and recording practices; or the measured items may capture a different facet of social regulation. Tests should therefore distinguish theory failure from measurement non-equivalence and lack of statistical power. When studies disagree, an adequate synthesis reports the different outcomes rather than average their conclusions rhetorically Bruinsma et al. (2013) Mazerolle et al. (2010).

Cross-national research also helps identify institutional boundary conditions. Resident intervention can depend on housing administration, social service access, policing arrangements, and the legitimacy of public authority. These conditions may alter the costs of intervention and the likelihood of cooperation even if individual trust is similar. Multisite designs with comparable measurement and contextual covariates offer a way to distinguish a recurring psychological mechanism from its local institutional expression Bruinsma et al. (2013) Wickes et al. (2013).

Crime, Fear, Cohesion, and Reciprocal Causation

Markowitz, Bellair, Liska, and Liu (2001) investigated relationships among neighborhood cohesion, disorder, and fear, extending social-disorganization theory toward a dynamic rather than exclusively one-directional account. Fear can alter how residents use public space, whom they meet, and how willing they are to participate in intervention. Disorder and cohesion can be mutually related over time. These associations complicate models that specify local trust as a fixed prior cause of safety while treating residents' experiences of crime as passive downstream outcomes Markowitz et al. (2001).

Reciprocal processes can create cumulative advantages and disadvantages. An event perceived as dangerous can discourage evening street use and lower opportunities for routine interaction; limited interaction can then make coordinated response less likely. Successful public cooperation may have the reverse effect by making help more credible and visible. The theoretical cycle is plausible but longitudinal identification remains necessary. A study that observes fear, efficacy, and violence only once cannot tell whether loss of trust preceded crime or crime damaged trust Markowitz et al. (2001) Lanfear et al. (2020).

Feedback models also need to separate fear from actual risk. Two neighborhoods with comparable victimization histories may differ in perceived threat because of reputational narratives, communication networks, and public attention. Different demographic groups may experience the same place differently. Research should therefore evaluate incident-level exposure, independently measured events, individual vulnerability, social interpretation, and neighborhood conditions in a multilevel framework Sampson & Raudenbush (2004) Markowitz et al. (2001).

Causal Inference, Selection, and the Neighborhood Effect

A causal claim about collective efficacy is a claim about what would happen to comparable neighborhoods if their shared capacity for informal regulation changed while relevant alternative pathways were held appropriately constant. This counterfactual is difficult to estimate because residents select or are constrained into neighborhoods, neighborhood conditions influence each other, and efficacy is often inferred from reports collected near the outcome date. Measured disadvantage and prior violence reduce some confounding but cannot guarantee exchangeability between high- and low-efficacy settings Sampson et al. (1997) Lanfear et al. (2020).

Longitudinal designs can establish whether shifts in efficacy precede shifts in outcomes, but time-varying confounding remains. An infrastructure improvement may simultaneously increase residents' trust and reduce opportunities for victimization. Fixed effects can remove stable unobserved neighborhood differences while leaving changing confounders unresolved. Natural experiments and carefully designed interventions can sometimes create more informative contrasts, provided treatment assignment, interference between areas, and compliance are recorded. Sensitivity analyses should express how strong an omitted cause must be to overturn a substantive conclusion Lanfear et al. (2020) Morenoff et al. (2001).

Investigators must also avoid conflating statistical mediation with identified causal mediation. The 1997 Chicago study found that collective efficacy statistically accounted for structural differences in violence. To claim that improving efficacy would eliminate a given proportion of inequality requires stronger assumptions: correct temporal order, no omitted exposure–mediator or mediator–outcome confounding, valid measurement, and a well-defined intervention on the mediator. The classic study established a compelling, testable theoretical account; translating its mediation estimates into policy predictions demands additional evidence Sampson et al. (1997) Lanfear et al. (2020).

Neighborhood Boundaries, Sampling, and Statistical Precision

Neighborhood definitions determine what counts as contextual exposure. Census tracts, respondent-defined neighborhoods, service districts, and small geographic grids have different advantages. Administrative units facilitate linkage with census and public records; perceived boundaries may better reflect actual social interaction; networks may traverse all such partitions. When the geographic unit changes, estimated between-neighborhood variance and the apparent strength of collective efficacy can change. A robust analysis pre-specifies the unit or tests conclusions across defensible alternatives Raudenbush & Sampson (1999) Morenoff et al. (2001).

Precision depends on both the number of neighborhoods and the number of informants within them. Hundreds of residents concentrated in one area cannot replace the independent contextual observations obtained from many distinct areas. Likewise, numerous sampled areas with only one or two respondents provide uncertain estimates of local norms. Multilevel measurement models can account for uneven respondents per setting, item differences, and respondent composition. Reporting variance components, uncertainty intervals, and effective contextual sample size allows readers to assess the evidence beyond nominal survey totals Raudenbush & Sampson (1999) Sampson et al. (1997).

Clustered data also introduce interference. One neighborhood's intervention may relocate harm or generate protective spillovers in adjacent areas. Standard causal frameworks that presume independent neighborhood outcomes may therefore be inappropriate. Spatial models and partial-interference designs can express defined channels of spillover, while direct measures of cross-neighborhood relationships help identify where boundaries are permeable. The solution is driven by the actual exchange and mobility processes the research seeks to explain Morenoff et al. (2001) Lanfear et al. (2020).

Psychometric Validity and the Meaning of Shared Expectations

Construct validation begins with the interpretation of each survey item. When asked whether neighbors would intervene in a fight, respondents may picture a verbal disagreement, a weapon, or a threat involving their own family. These imagined scenarios differ in danger and feasibility. Item wording, reference period, and population composition therefore influence meaning. Multi-item scales improve coverage, but averaging cannot eliminate ambiguity in what participants believed they were rating Raudenbush & Sampson (1999) Wickes et al. (2013).

Measurement invariance is relevant across age groups, gender identities, cultural communities, and cities. If residents in one setting use the upper response categories only when intervention is virtually certain, while others use them to indicate that intervention is desirable, equal numeric values fail to represent equal latent capacity. Cognitive interviews, anchoring vignettes, item-response models, and multigroup factor analysis offer different strategies for testing that possibility. Such checks are necessary before declaring national or historical trends from repeated questionnaire scores Wickes et al. (2013) Raudenbush & Sampson (1999).

Convergent validation can compare reported intervention expectations with recorded neighborhood meetings, mutual-aid requests, safe bystander actions, and independently observed responses to defined local problems. Discriminant validation should establish that the scale predicts action capacity beyond generic satisfaction, perceived safety, resident income, or institutional trust. Predictive validation requires outcomes measured separately from the same respondents' affective judgments. These strategies turn collective efficacy into an empirically accountable construct Sampson et al. (1997) Wickes et al. (2013).

Methodological Triangulation and Research Ethics

No single method captures all layers of informal social control. Surveys assess expectations and experiences; interviews uncover what residents mean by care or intervention; direct observation documents public conduct; network analysis identifies channels of support; administrative statistics index recorded incidents; and longitudinal research follows change. Triangulation matters because each method has a distinct error structure. Convergence across independently measured processes is stronger evidence for a neighborhood mechanism than repetition of one questionnaire in several datasets Raudenbush & Sampson (1999) Sampson & Raudenbush (1999).

Field observation raises privacy and consent questions, especially when public-space video, location traces, or neighborhood platform data are collected. Researchers should limit personally identifying records, specify observation purposes, protect vulnerable residents, and avoid publishing maps that expose sensitive locations. Survey questions about willingness to intervene may put respondents at risk if a dataset associates identifiable opinions with local conflicts. Methodological rigor includes secure data management, transparent coding, careful community consultation, and reporting designed to prevent stigmatization Sampson & Raudenbush (2004) Sampson (2012).

Community research also benefits from meaningful participation in formulating outcome measures. Residents may prioritize safe access to services, youth belonging, or freedom from harassment rather than aggregate official crime counts. A scientifically accountable project makes those preferences visible while still testing the reliability and accuracy of the chosen indicators. Local input informs the meaning of the construct; independent measurement and transparent methods evaluate the evidence. The two functions reinforce each other when carried out explicitly Wickes et al. (2013) Lanfear et al. (2020).

Structural Inequality and the Conditions of Collective Action

Collective efficacy has a structural history. Residential turnover, concentrated disadvantage, housing precarity, exclusionary investment practices, and access to institutions affect the opportunities residents have to form stable relationships and expect coordinated responses. Sampson, Raudenbush, and Earls (1997) examined the association of concentrated disadvantage and residential instability with violence through a measured neighborhood process. Later syntheses emphasize that structural and cultural processes unfold across periods longer than one survey cycle Sampson et al. (1997) Sampson (2012).

Theoretical explanation must keep structural constraints within view. A resident whose work schedule, disability access, housing insecurity, or exposure to retaliation inhibits intervention may still support shared neighborhood goals. Weak expressed willingness is not automatically evidence of apathy. Trust can be undermined by repeated abandonment of public services or by unequal treatment. Consequently, programs that ask residents to police each other without supplying safe, legitimate resources risk measuring deprivation as deficient civic virtue Sampson (2012) Sampson & Raudenbush (2004).

Research on disparity needs to examine who benefits from a reported neighborhood improvement. A decline in recorded crime might accompany displacement or changes in enforcement rather than lower underlying harm. An increase in complaints might reflect increased trust in reporting mechanisms. Researchers should distinguish victimization, reporting, enforcement, and institutional responses, and disaggregate benefits across groups. A community's ability to act together becomes more informative when investigated alongside the fairness and inclusiveness of its goals Sampson (2012) Lanfear et al. (2020).

Institutions, Police, and the Boundaries of Informal Control

Formal law enforcement, city services, schools, housing organizations, and informal neighborhood networks can interact in several ways. Residents may call public agencies, coordinate service requests, mediate disputes, or organize safe supervision. These behaviors differ from residents exercising governmental authority; they nevertheless rely on expectations that institutions will respond and that joint action will be legitimate. A theory of collective efficacy should model such institutional linkages as candidate enabling or constraining mechanisms Sampson (2012) Wickes et al. (2013).

Formal intervention can alter informal expectations. Reliable responses may make residents more willing to report threats and coordinate preventative activities. Arbitrary or discriminatory responses can make shared intervention dangerous, particularly for groups disproportionately subjected to scrutiny. These propositions require empirical tests with measures of perceived legitimacy, procedural fairness, response times, and independently ascertained harms. Aggregate neighborhood crime alone cannot determine how cooperation with institutions is experienced by different residents Sampson & Raudenbush (2004) Lanfear et al. (2020).

Research should distinguish contacting authorities from direct confrontation, monitoring from harassment, and prevention from punishment. The target behavior and the means of intervention are part of the psychological object being measured. A neighborhood may have strong capacity to seek safe public repairs and weak capacity for direct confrontation, a pattern that can be adaptive rather than deficient. Task-specific measurement therefore helps link collective efficacy theory with proportional and accountable policy design Wickes et al. (2013).

Community Programs and Intervention Evaluation

Programs intended to strengthen collective efficacy can include neighborhood associations, safe public-space activities, resident-led problem identification, youth participation, opportunities for reciprocal exchange, and predictable communication with service institutions. These are plausible strategies for enhancing social coordination. The classic observational studies do not license a blanket claim that any such program will lower violence. Program evaluation must test an explicit causal pathway from implementation to changed social expectations, actual cooperation, and independently measured outcomes Sampson et al. (1997) Lanfear et al. (2020).

An evaluation should document treatment components, assignment or selection into the program, baseline conditions, implementation fidelity, costs, and unintended consequences. Neighborhood-level randomized designs may be informative where assignment is feasible and interference is addressed; quasi-experimental designs require defensible assumptions about trends and alternative influences. Measuring changes in both perceived cohesion and task-specific action can reveal whether an intervention changes survey attitudes without improving actual coordinated capability Wickes et al. (2013) Raudenbush & Sampson (1999).

Program success has several dimensions: preventing victimization, supporting residents at risk, reducing fear, preserving procedural fairness, and improving the ability to obtain public services. These outcomes can move in different directions. A community intervention that strengthens relationships while displacing vulnerable residents fails an inclusive evaluation even if average survey ratings improve. Explicitly reporting distributional and adverse effects is essential to assessing collective efficacy as a social process rather than as an unqualified policy slogan Sampson & Raudenbush (2004) Sampson (2012).

Life Course, Youth Participation, and Intergenerational Context

Neighborhood relationships are learned and practiced across developmental stages. Adult supervision of public space, communication across families, school routines, and opportunities for youth contribution create occasions to observe whether others are dependable. Sampson, Morenoff, and Earls (1999) treated intergenerational closure as a distinct resource relevant to collective efficacy for children. The concept indicates a potential channel for child supervision and shared care, while keeping neighborhood-level expectations conceptually distinct from individual developmental outcomes Sampson et al. (1999).

Children and adolescents also possess local knowledge inaccessible to adult informants. They may recognize informal safety routes, contested locations, changing peer norms, and experiences of unequal surveillance. Youth involvement in measurement can therefore improve construct coverage; it also raises consent and safeguarding requirements. The research design should avoid interpreting every youth gathering as disorder or every adult intervention as beneficial. Adolescent participation in defining locally meaningful outcomes is a promising avenue for community psychology Sampson & Raudenbush (2004) Sampson et al. (1999).

Longitudinal life-course research could examine whether stable neighborhood relationships influence expectations of mutual assistance as residents become parents, older adults, newcomers, or caregivers. Residential mobility complicates exposure measurement: a participant may have lived in several neighborhoods, attended a school in another, and maintained long-distance family support. Estimating cumulative exposure requires time-varying residential histories and care in separating neighborhood processes from family resources and selective migration Sampson (2012) Morenoff et al. (2001).

Collective Efficacy Beyond Crime and Public Order

The underlying concept applies to tasks broader than controlling violence, provided researchers define the collective objective and mechanism. Neighbors may coordinate on emergency preparedness, public-space maintenance, accessibility, shared caregiving, or local service access. Wickes and colleagues (2013) directly analyzed separate intervention expectations for violence, delinquency, and civic/political issues, supplying empirical support for multidimensional task specification. Such applications should test new indicators rather than merely relabel the original crime-control instrument Wickes et al. (2013).

A wider approach must distinguish efficacy from political mobilization. Organizing a campaign for municipal funds may require shared capacities for public deliberation, representation, and sustained advocacy, whereas responding to immediate interpersonal violence may depend more on perceived safety, bystander coordination, and trust. The same networks can support both, but the costs, opportunities, and success criteria differ. A testable research program asks whether task-specific efficacy factors are correlated, what resources they share, and which outcomes each predicts Wickes et al. (2013) Bandura (1997).

Generalizing to health or disaster preparedness is an explicit interdisciplinary hypothesis until a domain-appropriate study supports it. The availability of neighbors for mutual support could plausibly affect stress appraisal or access to assistance, yet no claim about intervention effect sizes follows from the Chicago violence study alone. Each extension needs defined populations, validated measures, a temporal model, and outcomes matched to the domain. This preserves the explanatory reach of collective efficacy without dissolving it into a universal label for beneficial communities Sampson et al. (1997) Wickes et al. (2013).

Digital Neighborhood Infrastructures and Platform-Mediated Coordination

Digital message groups, local forums, neighborhood information services, and emergency communication channels create additional pathways through which residents can observe concerns and request coordinated action. They may make previously invisible needs legible, lower the effort required to contact multiple households, and document whether a response occurred. They may also concentrate attention on vivid but unrepresentative threats, amplify reputational judgments, and allow users outside a neighborhood to influence local perceptions. These propositions extend theories of communication and social context and require separate tests; classical neighborhood-efficacy studies did not evaluate contemporary digital platforms Sampson (2012) Sampson & Raudenbush (2004).

A digital collective-efficacy measure should specify which participants can act in the physical setting, which can coordinate remotely, and what intervention is anticipated. Counting platform members or posts is insufficient because access, participation, credibility, moderation rules, and response capacity differ. A successful design would trace requests from initial message to verified local action and independent outcome, while accounting for demographic differences in platform adoption. A network could have high message volume but low practical ability to support a vulnerable resident; the reverse is also possible Wickes et al. (2013) Raudenbush & Sampson (1999).

Privacy protection and platform governance are central when digital discussions mention suspected wrongdoing or individual identities. Rumor propagation can reproduce neighborhood stigma, while intrusive surveillance can discourage safe use of public space. Researchers should compare communication quality, accuracy of reports, fairness of moderation, and residents' opportunities to correct misinformation. This agenda tests whether technology changes the relational mechanism of collective efficacy rather than assuming more digital connection means stronger community control Sampson & Raudenbush (2004) Lanfear et al. (2020).

Artificial Systems as Participants in Community Coordination

In the Artificial Era, systems capable of summarizing public notices, translating messages, routing service requests, and helping residents compare response options may alter the information structure of community coordination. Their possible effects include reducing communication friction, clarifying available resources, and making conflicting claims easier to compare. These are proposed mechanisms at the intersection of community psychology and human–artificial relations. Existing neighborhood collective-efficacy studies do not establish that such systems themselves possess neighborhood trust, civic authority, or independently validated social-control capacity Bandura (1997) Wickes et al. (2013).

An empirical framework would distinguish human residents' confidence in one another from confidence in a technical coordinator. It would measure what information an artificial system receives, how it represents uncertainty, what instructions it gives, which humans retain authority over actions, and whether the resulting coordination is safe and equitable. A shared expectation that neighbors will respond is a psychological judgment about people situated in a relational setting. The performance of a digital system is separately assessable through accuracy, reliability, and authorized action Raudenbush & Sampson (1999) Lanfear et al. (2020).

Controlled studies could compare resident groups assigned different forms of decision support, with privacy protections and explicit accountability. Outcomes would include observed coordinated responses, perceived legitimacy, intervention safety, and distribution of benefits. This research would extend collective-efficacy theory into new infrastructures while preserving its original operational criteria. Any assertion that artificial coordination improves community safety would require direct evidence, reproducible measures, and an analysis of unintended effects Sampson (2012) Sampson & Raudenbush (2004).

Integrative Model: Conditions, Expectancies, Actions, and Outcomes

A useful synthesis organizes the evidence into four analytical layers: contextual conditions; shared expectations; actual local practices; and social consequences. Contextual conditions include material resources, residential stability, accessible institutions, network channels, and the history of victimization. Shared expectations are residents' beliefs about trust, mutual assistance, and responses to concrete problems. Practices include contacting neighbors, assisting someone at risk, coordinating services, and intervening safely. Consequences include harm, support, fear, inclusion, and perceptions of public order. Each layer requires its own indicators Sampson et al. (1997) Wickes et al. (2013).

The relationships among these layers are temporally reciprocal. Context may shape expectations, but outcomes can change context through migration, institutional responses, reputational effects, and network adaptation. Observed action may reinforce or undermine efficacy expectations. Spatial dependence means that processes in adjacent areas can influence the local trajectory. This integrative model is a conceptual organization of existing research findings and methodological challenges, not an independently established universal causal equation Morenoff et al. (2001) Markowitz et al. (2001) Lanfear et al. (2020).

The model makes competing hypotheses explicit. A cohesion-first account predicts that strengthened social ties generate intervention expectations that subsequently change behavior. A risk-response account predicts that reduced victimization first increases trust. An institutional-capacity account predicts that improved service response enables residents' cooperation. A stigma account predicts that reputational perceptions shape both reported efficacy and reported disorder. Multiple accounts may operate simultaneously; temporal, spatial, and multimethod evidence is required to estimate their relative importance Sampson & Raudenbush (2004) Bruinsma et al. (2013) Lanfear et al. (2020).

Open Questions and a Research Agenda

The field needs studies that validate collective efficacy against actual behavior, separate its task-specific components, and follow communities through changes in resources and safety. Carefully collected repeated survey panels, incident-level observation, spatial exposure histories, and independent outcome records would help identify how expectations turn into action. Research must measure both successful coordination and nonparticipation, including whether residents intended to intervene but lacked safe means to do so. Negative and contradictory findings should remain visible in cumulative synthesis Wickes et al. (2013) Bruinsma et al. (2013).

Multisite comparative studies should examine whether social-cohesion items retain the same meaning across languages, cultural traditions, housing types, and municipal institutions. The literature can then determine which pathways are generalizable and which are contingent on local political and social arrangements. Designs should report neighborhood definitions, cluster sizes, missingness, uncertainty, spillovers, and alternative causal models. An explicit plan for reproducibility would make different studies easier to compare without presuming identical mechanisms Mazerolle et al. (2010) Raudenbush & Sampson (1999).

An inclusive research agenda also asks whose interests community intervention advances, how actions are governed, and whether enhanced coordination produces unwanted surveillance or exclusion. Digital tools and artificial systems introduce new questions about authority, accuracy, privacy, and responsibility. Such questions are empirical research objectives whose answers depend on observed practices, not on the mere presence of technology. Future interventions should be evaluated using pre-specified safety and inclusion outcomes alongside conventional violence and disorder measures Sampson & Raudenbush (2004) Lanfear et al. (2020).

Conclusion

Collective efficacy and informal social control describe a consequential dimension of neighborhood social organization: shared expectations that residents can cooperate in responding to defined challenges and the practical capacity to enact such responses. The empirical tradition established by Sampson, Raudenbush, and Earls moved community theory toward explicit neighborhood measurement and demonstrated robust associations between an ecometrically assessed local capacity and violence in the 1995 Chicago setting. Historical social-disorganization research, subsequent spatial analyses, and comparative studies provide the intellectual and empirical architecture of the subject Sampson & Groves (1989) Sampson et al. (1997) Morenoff et al. (2001).

The evidence also sets precise limits. Cohesion, social ties, task-specific efficacy, reported control, and actual intervention are separable; their associations with crime can vary by setting, outcome, and model. Australian work supports careful cross-context comparison, while Dutch findings challenge uniform generalizations. Studies of disorder and stigma reveal that observed conditions and interpreted neighborhood reputations can diverge. Causal reviews caution that cross-sectional and even complex observational models do not automatically settle reciprocal pathways between social control, disorder, and victimization Mazerolle et al. (2010) Bruinsma et al. (2013) Sampson & Raudenbush (2004) Lanfear et al. (2020).

The cumulative scientific conclusion is that collective efficacy offers a productive, testable framework for studying how community conditions become organized expectations and coordinated conduct. Its further development depends on reliable contextual measurement, design-sensitive causal analysis, inclusive definitions of intervention, and realistic institutional and spatial models. New technological infrastructures, including artificial coordination systems, belong to the future empirical horizon of the field where their effects can be measured independently. The central scientific task remains to connect shared perceptions to observed capacities and consequences in specific communities Raudenbush & Sampson (1999) Wickes et al. (2013).

References

  • Bandura, A. (1997). Self-efficacy: The exercise of control. W. H. Freeman. https://books.google.com/books/about/Self_Efficacy.html?id=eJ-PN9g_o-EC
  • Browning, C. R., Dietz, R. D., & Feinberg, S. L. (2004). The paradox of social organization: Networks, collective efficacy, and violent crime in urban neighborhoods. Social Forces, 83(2), 503–534. https://doi.org/10.1353/sof.2005.0006
  • Bruinsma, G. J. N., Pauwels, L. J. R., Weerman, F. M., & Bernasco, W. (2013). Social disorganization, social capital, collective efficacy and the spatial distribution of crime and offenders: An empirical test of six neighbourhood models for a Dutch city. British Journal of Criminology, 53(5), 942–963. https://doi.org/10.1093/bjc/azt030
  • Lanfear, C. C., Matsueda, R. L., & Beach, L. R. (2020). Broken windows, informal social control, and crime: Assessing causality in empirical studies. Annual Review of Criminology, 3, 97–120. https://doi.org/10.1146/annurev-criminol-011419-041541
  • Markowitz, F. E., Bellair, P. E., Liska, A. E., & Liu, J. (2001). Extending social disorganization theory: Modeling the relationships between cohesion, disorder, and fear. Criminology, 39(2), 293–319. https://doi.org/10.1111/j.1745-9125.2001.tb00924.x
  • Mazerolle, L., Wickes, R., & McBroom, J. (2010). Community variations in violence: The role of social ties and collective efficacy in comparative context. Journal of Research in Crime and Delinquency, 47(1), 3–30. https://doi.org/10.1177/0022427809348898
  • Morenoff, J. D., Sampson, R. J., & Raudenbush, S. W. (2001). Neighborhood inequality, collective efficacy, and the spatial dynamics of urban violence. Criminology, 39(3), 517–558. https://doi.org/10.1111/j.1745-9125.2001.tb00932.x
  • Raudenbush, S. W., & Sampson, R. J. (1999). Ecometrics: Toward a science of assessing ecological settings, with application to the systematic social observation of neighborhoods. Sociological Methodology, 29(1), 1–41. https://doi.org/10.1111/0081-1750.00059
  • Sampson, R. J. (2012). Great American City: Chicago and the enduring neighborhood effect. University of Chicago Press. https://press.uchicago.edu/books/excerpt/2012/sampson_great_american_city.html
  • Sampson, R. J., & Groves, W. B. (1989). Community structure and crime: Testing social-disorganization theory. American Journal of Sociology, 94(4), 774–802. https://doi.org/10.1086/229068
  • Sampson, R. J., Morenoff, J. D., & Earls, F. (1999). Beyond social capital: Spatial dynamics of collective efficacy for children. American Sociological Review, 64(5), 633–660. https://doi.org/10.1177/000312249906400501
  • Sampson, R. J., & Raudenbush, S. W. (1999). Systematic social observation of public spaces: A new look at disorder in urban neighborhoods. American Journal of Sociology, 105(3), 603–651. https://doi.org/10.1086/210356
  • Sampson, R. J., & Raudenbush, S. W. (2004). Seeing disorder: Neighborhood stigma and the social construction of “broken windows.” Social Psychology Quarterly, 67(4), 319–342. https://doi.org/10.1177/019027250406700401
  • Sampson, R. J., Raudenbush, S. W., & Earls, F. (1997). Neighborhoods and violent crime: A multilevel study of collective efficacy. Science, 277(5328), 918–924. https://doi.org/10.1126/science.277.5328.918
  • Shaw, C. R., & McKay, H. D. (1942). Juvenile delinquency and urban areas: A study of rates of delinquents in relation to differential characteristics of local communities in American cities. University of Chicago Press. https://books.google.com/books/about/Juvenile_Delinquency_and_Urban_Areas.html?id=GKRkMQEACAAJ
  • Wickes, R., Hipp, J. R., Sargeant, E., & Homel, R. (2013). Collective efficacy as a task specific process: Examining the relationship between social ties, neighborhood cohesion and the capacity to respond to violence, delinquency and civic problems. American Journal of Community Psychology, 52(1–2), 115–127. https://doi.org/10.1007/s10464-013-9582-6