No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
The Concentration and Stability of Gun Violence at Micro Places in Boston, 1980–2008
Abstract
Boston, like many other major U.S. cities, experienced an epidemic of gun violence during the late 1980s and early 1990s that
was followed by a sudden large downturn in gun violence in the mid 1990s. The gun violence drop continued until the early
part of the new millennium. Recent advances in criminological research suggest that there is significant clustering of crime
in micro places, or “hot spots,” that generate a disproportionate amount of criminal events in a city. In this paper, we use
growth curve regression models to uncover distinctive developmental trends in gun assault incidents at street segments and
intersections in Boston over a 29-year period. We find that Boston gun violence is intensely concentrated at a small number
of street segments and intersections rather than spread evenly across the urban landscape between 1980 and 2008. Gun violence
trends at these high-activity micro places follow two general trajectories: stable concentrations of gun assaults incidents
over time and volatile concentrations of gun assault incidents over time. Micro places with volatile trajectories represent
less than 3% of street segments and intersections, generate more than half of all gun violence incidents, and seem to be the
primary drivers of overall gun violence trends in Boston. Our findings suggest that the urban gun violence epidemic, and sudden
downturn in urban gun violence in the late 1990s, may be best understood by examining highly volatile micro-level trends at
a relatively small number of places in urban environments.
KeywordsGuns-Gun violence-Hot spots-Epidemic
... Gun violence remains a critical issue in many U.S. cities, influenced by a complex interplay of social and economic factors [1]. Understanding the patterns and determinants of gun violence is essential for local policymakers to develop and implement effective intervention strategies. ...
... Spatio-temporal clustering patterns of crime events have been documented across various crime types, including violent crime [1,2,20,23] and property crime [11,19,28]. Early work by Messner et al. [16] identified a non-random clustering pattern of homicides at the county level, while Ratcliffe and Rengert [23] observed similar non-random spatiotemporal clustering at the block level. ...
... Previous studies [7,14,31] have suggested the presence of gun violence contagion, which can be driven by various factors, such as the rapid diffusion of human behaviors [3] or the structure of social networks [8,22,26,29]. Another line of research highlights a strong relationship between gun violence and the characteristics of local communities, including socioeconomic factors [1,12,20,22] and physical infrastructure [34]. The motivation of this work is that by considering both the contagion of gunshot events and the influence of various exogenous spatial factors across different communities on gun violence risk, we can provide a comprehensive depiction of the dynamics of gun violence in the city of Atlanta. ...
Analysis of gun violence in the United States has utilized various models based on spatiotemporal point processes. Previous studies have identified a contagion effect in gun violence, characterized by bursts of diffusion across urban environments, which can be effectively represented using the self-excitatory spatiotemporal Hawkes process. The Hawkes process and its variants have been successful in modeling self-excitatory events, including earthquakes, disease outbreaks, financial market movements, neural activity, and the viral spread of memes on social networks. However, existing Hawkes models applied to gun violence often rely on simplistic stationary kernels, which fail to account for the complex, non-homogeneous spread of influence and impact over space and time. To address this limitation, we adopt a non-stationary spatiotemporal point process model that incorporates a neural network-based kernel to better represent the varied correlations among events of gun violence. Our study analyzes a comprehensive dataset of approximately 16,000 gunshot events in the Atlanta metropolitan area from 2021 to 2023. The cornerstone of our approach is the innovative non-stationary kernel, designed to enhance the model's expressiveness while preserving its interpretability. This approach not only demonstrates strong predictive performance but also provides insights into the spatiotemporal dynamics of gun violence and its propagation within urban settings.
... The first research design investigates longitudinal patterns of crime at micro-places over extended periods of time. These studies find stability of temporal patterns and observe between years, most micro-places within cities have similar amounts of crime (Braga et al. 2010;Curman et al. 2015;Andresen et al. 2017). These findings suggest most locations which experience crime events are likely to experience revictimization during the ensuing years. ...
... (Weisburd et al. 2004). There have been several replications of this study which have supported the general conclusion that most micro-places experience revictimization or stability of crime patterns over time (Braga et al. 2010;Curman et al. 2015;Hibdon et al. 2017). These studies use observational periods which have ranged up to 50 years (see Wheeler et al. 2021a, b). ...
... Longitudinal research designs primarily use latent class analyses and examine years as the predominant unit of temporal observation (Curman et al. 2015). These designs do not offer much insight on temporal patterns outside of identifying groups of developmental trajectories which are often just characterized as either stable or volatile (see Braga et al. 2010). Repeat and near-repeat research designs provide much attention to the specific temporal units within observational periods to examine revictimization, but they are explicitly cross-sectional analyses. ...
This study presents a new perspective on the influence of time and the reoccurrence of crime problems at micro-places. I examined 342,690 aggravated assault incidents reported to the Chicago Police Department from 2001 to 2020 using a longitudinal repeat and near-repeat research design combined with cumulative incidence graphs across different temporal windows. There are two distinct periods to observe the revictimization of violence at micro-places. There is immediate risk after an incident within a week followed by a longer period with lower risk across 2 to 4 years when crime often routinely circles back to the same locations. Future research should continue to refine understanding of cyclical patterns or the “life course” of crime at micro-places to enhance the efficacy of place-based crime preventions strategies.
... " Spatial concentration also emerged as an important characteristic of firearm violence. An influential study by Spatial de-concentration of fatal and nonfatal firearm violence in Boston, MA, 2007MA, -2021 Faizah Shareef 1 , Emma L. Gause 2,3 , Suzanne McLone 4 , Erika Gebo 5 and Jonathan Jay 4,6* Braga and colleagues found that declining incidence at a small subset of microplaces accounted for firearm violence reductions in Boston, MA from 1980 to 2008 [7]. In that analysis, the 5% of microplaces where crime rates fluctuated most dramatically over time accounted for 74% of the firearm violence decline from 1980 to 2008. ...
... Societal changes since the early 2000s (e.g., gentrification [16], social media [17,18], and other factors) could have influenced the spatial distribution of violence; yet little recent work has examined temporal trends in the spatial concentration of firearm violence. The current study revisits these trends in Boston, MA, where earlier research contributed to the dominant understanding of spatial concentration [7]. Assessing the persistence of spatial concentration is necessary to balance the benefits and harms of place-based practices and to determine whether strategies and guidelines should be revised. ...
Background
It is a “law” of criminology that urban crime chronically recurs at the same microplaces (i.e., street segments and intersections). An influential study found high concentrations of firearm violence at microplaces in Boston, MA, from 1980 to 2008. The current study assessed whether this strong spatial concentration has persisted.
Approach
Fatal and nonfatal shooting incidents with one or more victims from January 2007 through September 2021 were included, obtained from the Boston Police Department. We matched shootings to the closest microplaces, i.e., intersections and street segments in Boston ( n = 32,267). We operationalized spatial concentration as the probability of shootings occurring at the same microplace. We employed a case-only design, with shootings as the units of analysis; the outcome of interest was a binary indicator for whether another shooting in the dataset occurred at the same microplace in the past or future. We used log-linear regression to estimate this outcome as a function of study year.
Results
Annual shootings decreased over the study period, except for a spike in 2020. Spatial concentration of shootings declined from 62% in 2007 to 55% in 2021. We estimated that spatial concentration declined by an average of 1.8% per year [95% CI (-1.1, -3.4), p < 0.001].
Implications
This declining trend in the spatial concentration of firearm violence has important implications for place-based interventions and underscores the need to monitor this trend over time. Social media, which reshapes social interactions and is linked to community violence, warrants further study as a potential cause.
... Perhaps the key finding in this area of study is that there is significant clustering of crime at places, irrespective of the specific unit of analysis that is defined (e.g. see Pierce et al. 1988;Sherman et al. 1989;Weisburd et al. 1992Weisburd et al. , 2009Weisburd and Green 1995;Brantingham and Brantingham 1999;Roncek 2000;Braga, Papachristos, and Hureau 2010;Andresen and Malleson 2011;Weisburd and Amram 2014;Curman, Andresen, and Brantingham 2015;Weisburd 2015;Gill, Wooditch, and Weisburd 2017;Haberman, Sorg, and Ratcliffe 2017). Such concentrations have been found across time within cities and across cities (Weisburd 2015). ...
... Weisburd et al. 2012aWeisburd et al. , 2021Groff and Lockwood 2014;Weisburd, Groff, and Yang 2014;Hipp and Kim 2019). While some scholars (Braga et al. 2010;Braga and Clarke 2014) have found support for a general application of opportunity principles to micro-geographic places, several others argue that opportunity theory is by its nature a specific theory approach since it sees particular types of opportunities as leading to specific kinds of crime. This latter concern forms the basis for Clarke and Cornish's critique of conventional sociological theories of crime causation, such as social disorganization theory (Clarke and Cornish 1985;Cornish and Clarke 1987). ...
We use data on street segments in Tel Aviv-Yafo to examine whether general or specific crime factors are key to understanding crime rates on street segments. We pose two questions. (1) What causal factors explain the property and violent crimes at the same locations? (2) What are the differences and similarities between the risk factors of the two crime types? Our study capitalizes on data drawn from the Israeli Central Bureau of Statistics (CBS) to identify social and opportunity data at the street segment level. The dependent variable is the average counts of violent and property crimes in Tel Aviv-Yafo between 2010 and 2014. Zero-inflated negative binomial (ZINB) regression explains violent and property crime variations. While many of the significant factors that explain crime are similar between the two crime types, there are also distinct criminogenic factors predicting violent and property crime. Overall, our results support the position of common or general crime causes at places, but at the same time, they suggest the importance of understanding specific causes for specific crime types.
... Research conducted by Braga et al. (2010) in urban settings has demonstrated that the concentration and stability of property crimes, including residential burglaries, can vary significantly across micro places within high-rise buildings. The spatial distribution of crimes within and around these buildings further highlights the importance of understanding the specific crime dynamics associated with vertical density. ...
... For instance, some studies have suggested that factors such as social isolation, anonymity, and access to escape routes within high-rise buildings may contribute to an increased risk of certain violent offenses, such as domestic violence or sexual assault (Braga et al., 2010). ...
... In particular, it has become the leading cause of death for children and teenagers in the US (Peña and Jena, 2022; Rossen et al., 2024). Many studies have con rmed that gun violence tends to occur in neighborhoods within relatively small areas (i.e., hotspots) and remains unchanged over time (Braga et al., 2010;Koper et al., 2015;Mazeika, 2023;Situ, 2024;Weisburd, 2015). This spatial nature has been explained in terms of routine activity and crime pattern theories (Eck and Weisburd, 1995;Felson, 2002), which emphasize the important role of both place and routine activities in crime. ...
Gun violence is a leading cause of death in the US. Understanding the spatial patterns of gun violence and how they may have been affected by the COVID-19 pandemic is essential for developing evidence-based prevention strategies. This study investigates whether the COVID-19 pandemic altered the spatial patterns of gun violence in Syracuse, New York. Using data from 2009 to 2023, we analyzed the annual incidence of gunshots at the census block group level and applied geospatial techniques including mean center, standard distance, Moran’s I, and Getis-Ord Gi* to assess its temporal trends and spatial clustering. Despite a reversal in the linear trends of annual gunshot counts, which declined before and rose during the pandemic, various spatial patterns of gun violence remained remarkably stable. The geographic center and dispersion of gun violence showed minimal variation, and gun violence continued to cluster within consistent “hot spot” neighborhoods throughout the entire study period of 14 years. Moreover, the intensity of clustering increased logarithmically with gunshot counts, suggesting that higher gun violence rates intensified within established areas rather than spreading citywide. A distance-decay effect further revealed that incidence of gunshots diminished with increasing distance from core hot spots, challenging assumptions of spatial spillover or contagion models. These findings suggest that entrenched structural conditions, such as racial segregation and neighborhood-level socioeconomic disparities, may be primary drivers of gun violence patterns, rather than temporary disruptions like pandemic-related policies. Methodologically, the study highlights the importance of long-term, fine-scale geospatial analyses to uncover persistent violence dynamics and guide preventive interventions. We argue that future violence prevention strategies should focus on enduring spatial patterns of gun violence and their underlying structural determinants, rather than reacting solely to short-term fluctuations in incident frequency.
... Broadly, the evidence regarding the law of crime concentration has been demonstrated across variously sized jurisdictions in the United States in cities such as Seattle, Minneapolis, Chicago, and Boston (Braga et al., 2010;Schnell et al., 2017;Sherman et al., 1989;, as well as in other countries such as Canada, England, Sweden, and India Marklund & Merenius, 2014;Mazeika & Kumar, 2017;Nelson et al., 2001). Thus, concentrating crime prevention resources in hot spots is compelling for both short-and long-term improvements (Andresen & Malleson, 2011;Braga et al., 2017;, and it should be expected to impact levels of violent crime in various contexts. ...
Hot spots policing is an effective, evidence-based strategy that reduces violent crime within small geographic units, or “hot spots,” in urban areas. A strong body of research demonstrates that these hot spots disproportionately contribute to cities’ overall crime rates. However, the existing literature has yet to answer a critical question: Can the localized crime reductions achieved through hot spots policing make a significant enough contribution to lower overall crime rates across an entire city? To address this gap, we examined seven years of violent crime data from three major Texas cities. One city underwent a two-year targeted hot spots policing intervention, while the other two cities served as counter-factual, similarly situated jurisdictions. We used interrupted time series analysis (ITSA) to examine longitudinal crime patterns in all three cities. The findings indicate that, compared to the control cities, the treatment city experienced a significant and substantial reduction in overall violent crime at the point of intervention. This decrease suggests that the hot spots intervention directly contributed to city-wide crime reduction, extending its impact beyond the designated hot spots.
... First and foremost, we wanted to use a more micro-level spatial unit than previous studies, which tended to use larger areal units. This is based on previous EC research that shows large amounts of variation across micro-level units like street blocks (e.g., Braga et al., 2010Braga et al., , 2011Groff et al., 2009Groff et al., , 2010Weisburd et al., 2004Weisburd et al., , 2012. Additionally, street blocks are theoretically important to the study of crime and place as they approximate behavior settings for human activity (Taylor, 1997(Taylor, , 1998(Taylor, , 1988Wicker, 1987). ...
Weather variables, like temperature and precipitation, have long been established as predictors of criminal behavior. So too have researchers established the importance of controlling for potentially criminogenic places when predicting when and where crimes will occur at micro-level units of analysis. The current study examines the role that temperature and precipitation play, along with places, in the odds of street robbery occurrence on street blocks in Cincinnati, Ohio. Using multilevel modeling, with days clustered within street blocks, our results showed that temperature, but not precipitation, predicted increased odds of street robbery occurrence, even after controlling for the presence of potentially criminogenic places. We conclude that research should continue to examine these important relationships, specifically how weather impacts the role of places in the formation of criminal opportunities.
... First, structural factors such as socioeconomic disadvantage, segregation, and marginalization are associated with patterns of violence across place (Peterson & Krivo, 2010;Sampson, 2012). The incidence of gun violence tends to concentrate in relatively small areas within communities (Braga et al., 2010), whereas the risk for violent injury and death spreads within social networks of high-risk individuals (Papachristos et al., 2012) through cycles of retaliation and normalized violence (Anderson, 2000). These empirical regularities have pushed prevention efforts toward multifaceted strategies that leverage acute intervention models to "treat" high-risk individuals and disrupt ongoing cycles of violence, as well as broader initiatives to build community capacity, increase equitable access to services (e.g., education, housing, mental health), and address community norms around violence and conflict resolution. ...
Research Summary
Firearm violence continues to be a leading cause of death in the United States. As alternatives to law enforcement intervention, community‐based violence prevention programs, such as Cure Violence, have become increasingly popular across U.S. cities. This article documents the results of a multiyear, mixed‐methods, quasi‐experimental study of the implementation and impact of Cure Violence in St. Louis, Missouri, from 2020 to 2023. We analyzed data from semistructured interviews, a two‐wave community survey, and police records of neighborhood violence, finding limited evidence of program effectiveness. Although some interview participants stated that the program had a positive impact, we observed no change in community norms/perceptions of violence, and our analysis of police data suggests that program implementation was associated with declining violence in only one of the three intervention sites.
Policy Implications
This article highlights challenges of implementing and evaluating community‐based violence prevention. More work is needed to unpack the mechanisms responsible for turning program activities into measurable impacts in both the short and long term. Implications for planning and evaluating community‐based violence prevention programs are discussed.
... The list of longitudinal analyses of hotspots policing is short yet informative. Braga et al. (2010) found that gun violence remained stable and concentrated in 5% of street segments and intersections in Boston between 1980 and 2008. Braga et al. (2011) observed that 1% of street segments and 8% of intersections accounted for 50% of commercial robberies and 66% of street robberies in Boston between 1980 and 2008. ...
A rich body of literature suggests that crime is concentrated in hotspots, some consistently ‘hot’ over long periods. However, whether there are spatial and temporal concentrations of domestic abuse (DA) is presently unknown. While it is plausible that DA data follow similar Pareto curves as general crime, it is equally reasonable to assume stochasticity, especially regarding year-to-year consistency. We conducted a retrospective longitudinal analysis of 1.7 million DA initial reports to the police (as opposed to ‘crime incidents’) over 13 years (2007–19) in London, UK. We also examine crime harm patterns, which provide a more nuanced risk estimate for victims based on a crime harm index. We utilize a combination of spatial statistics and trajectory modelling approaches. We find that a small percentage of addresses are responsible for an outsized proportion of DA counts but half the bandwidth for crime harm generated. Year-to-year repeat victimization at specific addresses is 69.9%, and the mean probability of receiving another DA report from the same address in the following month is 41%. For both crime count and harm models, locations with either low or high DA reportage remained as such throughout the study. Changes in less than 1% of locations will drive DA trends in London. We conclude that concentrating on place-based emergency-calls-for-service data rather than crime reports unmasks a substantially greater likelihood of repeat DA victimization than previously assumed. The discovery of a spatiotemporal DA hotspot allows law enforcement to ‘zero in’ prevention efforts on a small number of premises relative to the overall scale of the capital. Future DA research should place greater weight on micro-place factors associated with DA to calibrate prevention efforts’ accuracy and efficiency.
... Facing the problem of fixed grids, researchers have been seeking to predict crime on street networks. There is growing evidence that crime variability can be attributed to street segments [26], because the shape and structure of the street network are thought to play a crucial role in influencing the distribution of crime [27,28]. More importantly, predicting crime on the street is practical and effective for police patrolling in urban cities [29,30]. ...
In contemporary research, the street network emerges as a prominent and recurring theme in crime prediction studies. Meanwhile, graph representation learning shows considerable success, which motivates us to apply the methodology to crime prediction research. In this article, a graph representation learning approach is utilized to derive topological structure embeddings within the street network. Subsequently, a heterogeneous information network that incorporates both the street network and urban facilities is constructed, and embeddings through link prediction tasks are obtained. Finally, the two types of high-order embeddings, along with other spatio-temporal features, are fed into a deep neural network for street-level crime prediction. The proposed framework is tested using data from Beijing, and the outcomes demonstrate that both types of embeddings have a positive impact on crime prediction, with the second embedding showing a more significant contribution. Comparative experiments indicate that the proposed deep neural network offers superior efficiency in crime prediction.
... By processing and learning from this data, machine learning models can identify complex patterns, trends, and relationships that may not be apparent through traditional analysis methods. [5] II. ...
In contemporary law enforcement, the need for proactive strategies to combat crime and ensure public safety is paramount. This paper presents the development and implementation of a predictive modeling framework aimed at identifying crime hotspots and optimizing resource allocation for law enforcement agencies. The model leverages historical crime data, geographical information, and socio- economic factors to forecast areas at elevated risk of criminal activity. Through a multistage process encompassing data collection, preprocessing, feature engineering, and model training, the predictive model enables law enforcement agencies to anticipate and prioritize areas with the highest likelihood of crime occurrence. By strategically deploying resources to these identified hotspots, law enforcement agencies can intervene early, deter criminal activity, and reduce overall crime rates. This research contributes to the advancement of evidence-based policing practices by offering a scalable framework for crime hotspot mapping that prioritizes efficiency, effectiveness, and community partnership.
... As noted by Papachristos et al. (2012), for example, the gun violence spike that occurred in Boston during the mid-2000s was driven by gang activity and represented a problem of high-risk places and people. For context, about 5 percent of the street faces and intersections in Boston generated approximately 74% of the shootings that occurred in the city between 1980 and 2008 (Braga et al., 2010). Similarly, in their study of non-fatal gunshot injuries in Chicago that occurred between 2006 and 2012, Papachristos et al. (2015) found that 25% of all gunshot victimization occurred in 6% of all beats, while 70% of all non-fatal shooting incidents occurred in networks comprising less than 6% of the total population; thus, indicating a subset of risky places and people. ...
A growing body of literature has explored the ‘life course’ of crime guns, with particular focus on the time between initial point of sale of firearms and their eventual recovery by police following a crime. We contend that this examination is incomplete, with limited consideration given to the period between a firearm’s first known use in a criminal offense and its recovery by police—which we refer to as time in crime. Increased understanding of this time frame is important given that crime guns are frequently recirculated among criminally involved groups and the recent finding that time in circulation following first known use in a crime is a significant predictor of multiple uses of crime guns. We add to the literature through application of negative binomial regression to a sample of 310 crime guns used in offenses in a city in the Southeastern United States to examine how neighborhood context and initial incident characteristics influence the number of days that firearms remain in circulation after their first known use in a crime. We find that increased levels of concentrated disadvantage and gang involvement during the original incident correspond with significant increases in time in crime, while increased levels of residential stability and the ability of police to identify suspects are linked with more rapid recovery of crime guns. Notably, these findings hold even after inclusion of popular time to crime covariates, including firearm quality, caliber, and status as a stolen gun.
... This aligns with Clarke and Newman's (2006) argument that counterterrorism through situational prevention is one of the most effective preventative measures. There is robust evidence that placebased interventions can be effective without displacing crime to nearby areas (Braga, Papachristos, and Hureau 2010). In the few studies that have examined the effectiveness of place-based interventions in counterterrorism (in addition to overall effectiveness, e.g. ...
One of the areas in terrorism research that has not been adequately examined is the geographic concentration of pyro-terrorism (PT) attacks. The literature related to the geographical distribution of terror generally focuses on macro-places (continents, countries and regions) and micro-places – “hot spots”. However, to our knowledge, the geographical distribution of terror in forests has yet to be studied. The findings of “hot spots” studies have great relevance for law enforcement practitioners. This study examines the spatial characteristics of forest PT and whether there is a concentration of PT attacks in a limited number of “hot forests” that are stable over time. This study examined the entire population of suspected forest PT in Israel between 2008 and 2022 (excluding 2016), totalling 2,297 attacks. The research revealed high spatial clustering of forest PT attacks concentrated in specific “hot forests” and that they were stable over time. In conclusion, the concentration of PT attacks calls for a specialized counterterrorism response, equivalent to “hot spots policing”, based on the characteristics of potential terrorist hot forests.
... These lapses in public safety are not equally distributed. They concentrate geographically and are persistent historically (Braga, Papachristos, and Hureau 2010;Grosjean 2014; Matthew R. Lee, Thomas, and Ousey 2009). For example, Louisiana had about four times more homicides per capita than Massachusetts in the 1800s, and it still has about four times more today (Asher, Horwitz, and Monkovic 2021). ...
Why do high rates of interpersonal violence sometimes persist despite the presence of a strong state? We argue that this is because individuals who grow up in unsafe places learn to distrust criminal justice institutions, such as police, and to rely on themselves and their family for protection. Even if institutions improve, they persist in these self-defense habits. To examine the empirical viability of this argument, we study millions of internal white US migrants. We find that migrants from historically unsafe states carry much of their risk of violent death with them, including the risk of police violence. They do so even when compared to similar migrants living in the same region. In a large-scale national survey, we find that they distrust law enforcement and are more likely to rely on themselves for protection in dangerous situations. Our results have implications for which strategies governments could adopt to reduce violence.
... While researchers have historically examined spatial crime concentrations at coarser levels of spatial aggregation (see Bursik and Grasmick 1993;Kubrin and Weitzer 2003;Sampson et al. 1997), there is a growing realization that these larger places are spatially heterogenous (see Braga et al. 2017). This increased acknowledgement has led to a plethora of studies examining crime at the street segment level (see Braga et al. 2010;Groff et al. 2010;Kim and Hipp 2018;Weisburd et al. 2009) as well as studies delineating the safest and shortest routes between locations based on the underlying crime risk of the street network (see Galbrun et al. 2016;Garvey et al. 2016;Goel et al. 2017;Mata et al. 2016;among others). In terms of graph theory, street networks are a collection of nodes (intersections) and edges (street segments linking intersections) which can either be topological (nonspatial) graphs (with topological distance in terms of number of street segments along a path (Goddard and Oellermann 2011) or spatial graphs (with spatial/metric distances in meters or kilometers (Crucitti et al. 2006). ...
Safety is arguably the single biggest daily concern for commuters. Popular navigation software systems such as Google Maps, TomTom and Waze do not however currently give the same weight of consideration to crime risk when directing road users to the safest and shortest routes as they do existing traffic knowledge. In this study we propose three simple methods to delineate the safest and shortest route between any two locations in South Africa considering crime risk. The first method uses historical crime data to delineate the safest and shortest route; the second method, uses only the location of so-called ‘risky facilities’ to determine the safest and shortest route; while the third method uses a combination of both historical crime data and facility data to delineate the optimal route. We contrast the resultant routes delineating by these three methods and speculate on the policy implications of integrating this knowledge into future navigation software systems.
... Our study involved partnership among four police agencies across the Roanoke Valley with Blue Ridge A place-based randomized controlled trial • 3 Behavioural Healthcare, a mental health service provider, to establish a standard crisis response protocol that provides police with peak-hour access to trained MHPs who can offer stabilization services and guide individuals into further treatment if needed. Following the recommendations of a recent evaluation of co-response models and the hotspots policing literature (Braga et al., 2010;Curman et al., 2015;Gill et al., 2017;Hibdon et al., 2017;Koziarski, 2021;Steenbeek and Weisburd, 2016;Vaughan et al., 2016Vaughan et al., , 2018Weisburd and Amram, 2014;Weisburd et al., 2004Weisburd et al., , 2009Wheeler et al., 2016;White and Goldberg, 2018;Yang et al. 2018b), we designed and implemented the co-response program on the mental health hotspots, defined by the amount of mental health CFSs, from 16 May 2022 to 31 December 2022. The Roanoke Valley, which includes Roanoke County, the cities of Roanoke and Salem, and the town of Vinton, is a mixture between urban-suburban-rural region in southwest Virginia. ...
Responding to incidents involving individuals with mental illness has been a challenge for police officers. While co-response teams have been embraced as an effective police response strategy, most prior evaluation studies on co-response teams focused on outcomes that are not directly related to individuals’ subsequent mental health state. Additionally, the lack of experimental research hinders our ability to draw causal conclusions on the effects of co-response teams. To address this knowledge gap, this study evaluated the effectiveness of co-response teams on hospitalization outcomes of individuals in crisis using a place-based randomized controlled trial in southwest Virginia. Street segments with high volumes of mental health calls for service were randomly assigned to either treatment hotspots (n = 113) or control hotspots (n = 115). The results suggest that the co-response teams had a significant effect on reducing subsequent hospitalizations, with an estimated effect size of −0.22. The findings, challenges, and recommendations for future co-response team implementations were discussed.
... Firearm violence is spatially concentrated within micro geographic spaces that have higher incidences of structural disadvantage and racial disparities (Braga et al., 2010;Magee, 2020). Black children are at a substantially higher risk of firearm injury and exposure to firearm violence compared to White children, largely within low socioeconomic communities Magee, 2020;Martin et al., 2022;Pino et al., 2023). ...
... Il s'agit de non seulement cibler les points chauds, mais également d'étudier en profondeur leur évolution temporelle. Plusieurs auteurs ont déjà examiné ce type d'analyse (Braga et al., 2010;Curman et al., 2015;Weisburd et al., 2004). Pour obtenir une compréhension complète des dynamiques spatio-temporelles de la criminalité, il est nécessaire d'explorer plus en détail les analyses qui combinent à la fois les aspects spatiaux et temporels. ...
Mémoire présenté en vue de l'obtention du grade de Maîtrise ès sciences (M. Sc.) en criminologie, option mémoire Août 2023.
Le contexte particulier de la pandémie de la COVID-19 a engendré pour tous des bouleversements sociaux et économiques et plusieurs mesures exceptionnelles ont été mises en place, dont celle du couvre-feu, imposée au Québec en date du 9 janvier 2021 à 20 h 00. Les politiques adoptées pour atténuer la propagation de la COVID-19 constituent une expérience sociale sans précédent. La théorie des activités routinières prédit que les mesures qui influencent les activités sociales auront un impact certain sur la criminalité. Le couvre-feu a considérablement réduit les probabilités d’un tel regroupement, du moins pendant une période définie (par exemple, de 20 h 00 à 5 h 00). Il est possible que le couvre-feu ait pu avoir un impact sur la délinquance, mais aussi sur son déplacement dans l'espace et dans le temps. L’objectif de cette recherche est d’évaluer l’impact du couvre-feu imposé en janvier 2021 sur la criminalité dans la ville de Montréal. Plus spécifiquement, ce mémoire vise à évaluer : (1) s’il y a eu une augmentation ou une diminution de la criminalité, tout en examinant l’impact sur les différents types de crimes, (2) les heures de commission des délits et les types de lieux affectés par ces variations. Les analyses de type spatio-temporelles sont basées sur les données officielles de la criminalité du Service de police de la ville de Montréal (SPVM). Alors que la plupart des études disponibles jusqu'à présent se sont penchées sur l'impact du COVID sur le nombre de crimes (enregistrés), celle-ci recherche spécifiquement l'existence de déplacement dans l'espace et dans le temps. Ce mémoire apporte donc une contribution importante à la littérature sur les impacts d’un couvre-feu sur le crime. En effet, les résultats indiquent que la criminalité diminue initialement à la suite de l’imposition du couvre-feu, cependant cette perturbation est temporaire uniquement. De plus, un déplacement temporel est constaté puisque les taux de crimes sont plus élevés dans les heures où la mesure n’est pas en vigueur. Au niveau géographique, il n’y a pas de déplacement spatial de la criminalité observé à la suite du couvre-feu. Il est intéressant pour les décideurs de connaître les impacts de ce type de mesure sur la criminalité.
The particular context of the COVID-19 pandemic has caused social and economic upheavals for all and several exceptional measures have been put in place, including a curfew, imposed in Quebec on January 9, 2021 at 8:00 p.m. The policies adopted to mitigate the propagation of COVID-19 are an unprecedented social experiment. Routine activity theory predicts that measures that influence social activities will significantly impact crime. The curfew has greatly reduced the likelihood of such a gathering, at least for a defined period (e.g., from 8:00 p.m. to 5:00 a.m.). It is possible that the curfew could have impacted not only crime but also its movement in space and time. This research aims to assess the impact of the curfew on crime in the city of Montreal. More specifically, it aims to assess: (1) whether there has been an increase or a decrease in crime while examining the impact on different types of crime, (2) the times of commission of the offences and the types of locations affected by these variations. Spatio-temporal analyses are based on official crime data from the police department of Montreal. While most studies available so far have looked at the impact of COVID on the number of (recorded) crimes, this one specifically seeks the existence of displacement in space and in time. This dissertation contributes significantly to the literature on the impacts of a curfew on crime. Indeed, the results indicate that crime initially decreases following the imposition of the curfew, but this disruption is only temporary. Additionally, a temporal displacement is observed as crime rates are higher during the hours when the curfew is not in effect. There are no indicators of spatial displacement due to the curfew. Understanding the impacts of such measures on crime is important for policymakers.
Research shows that crime is spatially concentrated. However, most research on crime concentration has been conducted in U.S. or European cities. This study assessed whether the "law of crime concentration at place" applies to non-Western settings. This study intends to extend the law of crime concentration’s applicability to a different urban and cultural settings. I tested this by focusing on burglary incidents in Taipei City, Taiwan, from 2015 to 2019. I used data from the Taipei City Police Department, focusing exclusively on burglary over 5 years. I employed descriptive statistics, Lorenz curves, and Gini coefficients to quantify and visualize crime concentration at the micro-place level. The units of analysis are clusters of 90 housing units, similar to street segments used in other studies. Consistent with the law of crime concentration at place, burglary in Taipei City is significantly concentrated. Specifically, 51.3% of burglaries occurred in just 5% of micro-place units and 25.9% in only 1.7% of units. Over 80% of spatial units were burglary-free. The generalized Gini coefficient of 0.4967 indicates a noteworthy disparity in the distribution of burglary incidents across different micro-places. I found evidence of crime concentration in a non-Western setting, suggesting that similar patterns of crime distribution may exist across different cultural and geographic contexts. However, socio-economic factors such as residential density and poverty may influence the lower burglary rates in Taiwan than in the U.S. The study site is a densely populated urban setting that may not represent other regions or rural areas in East Asia. Further research is needed to explore the applicability of the law of crime concentration in diverse geographic and socio-economic contexts.
Objectives
In the United States, firearm injuries disproportionately occur in low-income communities and among racial and ethnic minority populations. Recognizing these patterns across social conditions is vital for effective public health interventions. Using timely and localized data, we examined the association between social vulnerability and firearm injuries in King County, Washington.
Methods
For this ecological, cross-sectional study, we used health reporting areas (HRAs) (n = 61), a subcounty geography of King County. We obtained HRA-level counts of firearm injuries by using responses from King County emergency medical services (EMS) from 2019 through 2023. We measured HRA-level social vulnerability by using the Social Vulnerability Index (SVI) from the Centers for Disease Control and Prevention/Agency for Toxic Substances and Disease Registry, categorized into tertiles (low, moderate, and high SVI). We used bivariate choropleth mapping to illustrate spatial associations between SVI and rates of firearm injuries per 10 000 residents. We used Bayesian spatial negative binomial regression to quantify the strength of these associations.
Results
Bivariate choropleth mapping showed a correlation between SVI and rates of firearm injuries. In spatial models, HRAs categorized as high SVI had a 3 times higher rate of firearm injuries than HRAs categorized as low SVI (incidence rate ratio = 3.01; 95% credible interval, 2.02-4.47). Rates of firearm injuries were also higher in HRAs categorized as moderate versus low SVI (incidence rate ratio = 1.72; 95% credible interval, 1.23-2.40).
Conclusion
In King County, areas with high social vulnerability had high rates of EMS responses to firearm injuries. SVI can help identify geographic areas for intervention and provide a framework for identifying upstream factors that might contribute to spatial disparities in firearm injuries.
Microgeographic units of analysis have moved to the center of criminological inquiry. This Element brings together leading crime-and-place scholars to identify promising areas for future study. Section 1 introduces the Element and the importance of focusing on the future of studies of crime and place. Section 2 examines the development of hot-spots policing and the importance of focusing on its impact on communities. It also looks at how 'pracademics' can advance the science and practice of place-based policing. Section 3 focuses on place managers as prevention agents and examines how city government can influence crime at place. It further contends that rural communities need to become a key focus of crime-and-place scholarship. Section 4 emphasizes the importance of the connection of health, crime, and place. It also argues for the importance of expanding the methodological tools of crime and place to include careful ethnographic and qualitative research.
The US has alarmingly high, and rising, gun violence and fatality rates. While it is generally believed that supply side policy is critical in addressing the problem, legislation has been relatively ineffective, for example the Gun Violence Prevention and Community Safety Act of 2020. Previous literature has identified hotspots and proximate determinants, particularly urban locations. Effective supply side policy would require application of optimal transport methods. We apply novel and current generation structural models to identify the spatial network for supply of guns and propagation of violence risks. This provides clean quantification of spatial supply chains and associated costs of transport between nodes, which are necessary for mitigation using optimal transport and logistics network management methods.
This study explores the impact of gunshot detection technology (GDT) in Denver, CO, Milwaukee, WI, and Richmond, CA on officer workload and response times to firearm shooting notifications. All three departments experienced a twofold increase of being alerted to shooting incidents following GDT implementation, leading to a two- to threefold increase in staff time allocated to responding to these incidents. Response times to specific events were generally quicker for GDT alerts compared to traditional 911 calls for service regarding shots heard events. However, the time savings are primarily from the duration between notification and assignment of an officer, rather than the actual time taken by officers to arrive at the scene after being assigned. Insights from 46 stakeholder interviews offer optimal approaches for investigating shooting notifications and GDT program management. While GDT programs effectively reduce response times, agencies must anticipate increased workloads and implement specific program characteristics for optimal effectiveness.
Objectives
The importance of peer relations is rooted in decades of policing research; however, scholars have largely overlooked the role of peers in officers’ use-of-force behaviors. The current study investigates the “connected” nature of police use of force.
Methods
Data on officers’ networks are reconstructed from 11,834 use-of-force reports involving 1,894 officers in seven departments in New Jersey. Exponential Random Graph Models evaluate which officer-level attributes and network dependencies are associated with officers’ co-involvement in police use-of-force incidents.
Results
Findings indicate the police use of force is not evenly distributed but concentrated on a subset of officers and partnerships. Variation in officers’ likelihood of using force together is driven by individual characteristics, including officer race/ethnicity, rank, and tenure. In addition, co-involvement in force clusters among officers, with officers likely to engage in force together when they share a connection.
Conclusion
This study highlights an alternative starting point for understanding police use of force. By paying greater attention to the structural makeup of the department, such as the connectivity of the force network, agencies can design efforts that aim to reduce incidents of force through relational properties such as assignments and partnerships.
Cities across the country have increasingly turned toward community violence interventions (CVI) to address community safety without relying on criminal legal strategies. This article inspects beneath the veneer of present-day CVI approaches to examine how their work is dedicated as much to neighborhood social organization as it is to responding to gun violence. Underneath contemporary definitions of outreach workers as mediators of violence, earlier sociologists and criminologists conceived of these workers as frontline builders of community charged with mending breaks in the social fabric. Acknowledging this past is important because it re-centers criminology's contributions to the practice of street outreach and provides insights that help to comprehend the challenging present moment in American public safety. We offer directions for a reinvigorated social science of street outreach that re-centers community processes, structures, and institutions and, in so doing, might better inform contemporary practice and policy.
Purpose
The objectives of this study are to (a) identify spatial and temporal crime concentrations, (b) supplement the traditional place-based analysis that defines hot spots based on counted incidents with an analysis of crime severity and (c) add to the research of hot spots with an analysis of offender data.
Design/methodology/approach
This study explores crime concentration in mass transit settings, focusing on Edmonton’s Light Rail Transit (LRT) stations in 2017–2022. Pareto curves are used to observe the degree of concentration of crime in certain locations using multiple estimates; trajectory analysis is then used to observe crime patterns in the data on both places and offenders.
Findings
A total of 16.3% of stations accounted for 50% of recorded incidents. Train stations with high or low crime counts and severity remained as such consistently over time. Additionally, 3.6% of offenders accounted for 50% of incident count, while 5% accounted for 50% of harm. We did not observe differences in the patterns and distributions of crime concentrations when comparing crime counts and harm.
Research limitations/implications
Hot spots and harm spots are synonymous in low-crime-harm environments: high-harm incidents are outliers, and their weight in the average crime severity score is limited. More sensitive severity measures are needed for high-frequenty, low-harm enviornments.
Practical implications
The findings underscore the benefits of integrating offender data in place-based applied research.
Originality/value
The findings provide additional evidence on the utility of place-based criminology and potentially cost-effective interventions.
Retail or commercial businesses in a neighborhood may act as crime generators or attractors, providing opportunities for crime, including homicide. However, different types of businesses may be more likely to be associated with homicide relative to others. Businesses such as restaurants that encourage clientele to stay on the property tend to have more active forms of guardianship, such as waitstaff or bouncers, whereas off-site or takeaway businesses like gas stations have more passive surveillance measures such as video cameras in place. Further, the relationship between businesses and homicide might vary based on neighborhood characteristics, including disadvantage. Using negative binomial regression models, we predict the number of homicides across several different types of businesses, controlling for neighborhood factors. We find that both off-site and on-site businesses are associated with more homicides, but find limited evidence that the effects are worsened when these businesses are located in socially disorganized neighborhoods. This suggests that businesses associated with lower levels of guardianship may contribute to homicide, but do not alone explain the concentration of homicide in disadvantaged neighborhoods.
: Incarceration-based approaches to illegal gun possession have not proven effective at reducing gun violence, and they have created dramatic racial disparities. Within this context, a small number of jurisdictions have developed prosecutor-led gun diversion programs (PLGDPs), which offer diversion from prosecution and an opportunity to engage in community-based services with a common goal of reducing illegal gun possession. The purpose of this paper is to explore the factors that contribute to illegal gun possession among PLGDP participants, and the extent to which PLGDP programming addresses these complex factors. Semi-structured interviews were conducted with 32 PLGDP stakeholders (8 PLGDP participants, 15 prosecutors, and 9 service providers), and qualitative analyses identified themes related to illegal gun possession and PLGDP programming connections and disconnections. Findings indicate that safety concerns related to structural issues of community violence are a primary factor driving gun possession among PLGDP participants. While PLGDPs were viewed as having some benefits, disconnects in PLGDP programming centered on assumptions made about the target population and the limitations of individual-level interventions to curb gun violence. If PLGDPs are to have an impact on gun violence, trauma-focused approaches must be incorporated, and efforts should be made to better understand and address environmental factors.
O objetivo geral deste artigo é explorar o mapeamento de crimes por microespaços sob a perspectiva da lei de concentração do crime, tendo como recorte espacial a cidade de Belo Horizonte com recorte temporal dos dados de 2018 a 2021. O percurso metodológico contou com o mapeamento dos crimes violentos e furtos para Belo Horizonte com agregação dos pontos de ocorrência em linhas para análise da concentração do crime por segmentos de rua. Os resultados indicaram que, para o caso de Belo Horizonte, há um ajustamento entre os dados coletados e a Lei de Concentração do Crime. Em 2021 houve concentração de 25% dos furtos ocorridos em 0,40% dos segmentos de rua em Belo Horizonte e 50% deste mesmo tipo de delito ocorreu em 2,56% dos segmentos. Os crimes violentos também apresentaram comportamento semelhante. Para o ano de 2021, 25% dos crimes violentos ocorreu em 0,96% dos segmentos de rua e 50% destes delitos ocorreu em 3,42% dos segmentos de rua. De maneira complementar, verificou-se que há uma desigualdade muito grande em termos de concentração criminal, os coeficientes de Gini apresentaram valores próximos de 0,90, o que significa que o crime no município está muito concentrado em pequenas porções do seu território. Outra previsão teórica que se confirmou foi a estabilidade dos pontos quentes. Em Belo Horizonte os dados indicaram que houve uma estabilidade do status em termos de concentração criminal para os segmentos de rua entre o período de análise (2018 a 2021) acima de 70% para ambos os crimes analisados, o que demonstra que, apesar de diminuição geral nas taxas criminais durante o período, os pontos quentes permaneceram quentes. Esta constatação além de fornecer melhor subsídio para a implementação de estratégias de prevenção criminal por meio de Policiamento de Pontos Quentes também robustece o campo de pesquisa da geografia do crime e criminologia ambiental com trabalho de análise empírica.
While police officers must adapt behavior between places to effectively do their jobs, these decisions could result in some communities receiving different levels of exposure to the police. This study explores a new spatial measure of police contacts to observe these differences. We calculate neighborhood-specific Gini coefficients based upon the spatial distribution of 77,752 police-civilian stops at street segments and intersections nested within census tracts in Oakland, California. This coefficient presents a contrast between two divergent distributional patterns—the diffusion of police contacts to more places across neighborhoods and the concentration of contacts at fewer “hot spot” places within neighborhoods. The most consistent environmental explanation for these differences was the race/ethnicity of neighborhood residents, which was associated with the police stopping people across more places. Future research should continue to investigate this finding and examine the mechanisms that explain why spatial exposure to police contacts changes between places.
The current study uses the CDC’s social vulnerability index (SVI), an underutilized measure in criminology, to examine patterns of fatal and nonfatal firearm violence. The SVI is an index that captures the level of social vulnerability within communities that can reflect a community’s ability to withstand many stressors, including firearm violence. Results indicate that vulnerable communities, as measured by the SVI, experience disproportionate firearm violence. Results also show that neighborhood firearm violence is influenced by nearby neighborhoods’ levels of social vulnerability. That is, neighborhoods that are surrounded by other neighborhoods with high levels of social vulnerability experience more firearm violence. These results are important for policymakers who seek to address the root causes of vulnerabilities that impact firearm violence indirectly, and the results provide a foundation for future application of the SVI within criminology.
Despite abundant research on spatial predictors of violent victimization, few studies have explored the varying effects of these predictors on victims of different demographic backgrounds. This study seeks to bridge this gap by integrating perspectives from social disorganization, routine activity, risky lifestyles, and broken windows theories. Utilizing 2017-2019 crime data from Arlington, TX, the authors employ generalized spatial two-stage least squares models to dissect the nuanced effects of place characteristics across different racial and ethnic groups. The findings highlight the importance of victim race/ethnicity in predicting robbery victimization within high-risk areas, uncovering the disparate effects the same place features have on people of different racial or ethnic background.
Community gun violence is a serious public safety issue. Unfortunately, limitations surrounding data on shootings have made it difficult for members of the public and the research community to understand the nature and trends associated with gun violence and have also hampered policy officials' attempts to reduce the incidence of shootings. This paper makes use of shooting data from the Boston Police Department that contains the location of the shooting as well as whether the shooting was fatal or non-fatal. Results show that two police districts, Roxbury and Mattapan, account for over half of the shootings and that the risk of being fatally shot in these two districts for a black male is almost double that in other districts. Findings have implications for police allocations as well as policy interventions.
Crime is highly concentrated at places that lack capable place managers (i.e., landlords and their delegates). In response, numerous cities have instituted problem property interventions that pressure landowners to better manage properties suffering from decay, nuisance, or crime. This approach is distinctive in that it both targets a place and incentivizes those legally responsible to improve its management, yet little is known about the efficacy of such interventions. We assess the short‐ and long‐term impacts of such interventions in Boston, Massachusetts, using matched difference‐in‐difference analyses. Problem property interventions reduced crime and disorder relative to comparable matched properties. They also led to property investment and landowner turnover, suggesting strengthened place management. In addition, drops in crime and disorder were observed at other properties on the same street, although not at other properties with the same owner throughout the city. This study, therefore, provides evidence that problem property interventions compel landowners to better manage the targeted property and that these effects have a diffusion of benefits on surrounding properties. The effect on place management, however, was limited to the target property and did not reliably generalize to the landowner's other holdings. This study reveals nuance in the ways that problem property interventions can benefit communities.
The epidemic of youth violence in the United States peaked in 1993 and has been followed by a rapid, sustained drop. We assess two types of explanation for this drop-those that focus on "cohort" effects (including the effects of abortion legalization) and those that focus on "period" effects (including the effects of the changing crack-cocaine trade). We are able to reject the cohort-type explanations yet also find contradictions with an account based on the dynamics of crack markets. The "way out" of this epidemic has not been the same as the "way in." The relative importance in homicide of youths, racial minorities, and guns, all of which increased greatly during the epidemic, has remained high during the drop. Arrest patterns tell a somewhat different story, in part because of changing police practice with respect to aggravated assault. Finally, we demonstrate that the rise and fall of youth violence has been narrowly confined with respect to race, sex, and age, but not geography. Given the volatility in the rates of juvenile violence, forecasting rates is a risky business indeed. Effectively narrowing the range of plausible explanations for the recent ups and downs may require a long time horizon, consideration of a broader array of problem behaviors, and comparisons with trends in other countries.
This paper makes problematic the situational conditions under which policemen write official crime reports in field encounters with complainants. These reports are the raw materials for official crime rates--"crimes known to the police." They also are a prerequisite of further investigation of the crime by the detective bureau and thus of apprehension of the offender. They constitute official recognition of crimes. The findings derive from a three-city observation study of routine police encounters. Among the conditions that relate to the production of official crime reports are the following: the legal seriousness of the complaint, the complainant's observable preference for police action, the relational distance between the complainant and the suspect, the complainant's degree of deference toward the police, and the complainant's social-class status. However, there is no evidence of racial discrimination in crime-reporting. We interpret these empirical patterns not only from the standpoint of crime rates as such but also from the standpoint of the relation between police work and other aspects of social organization.
The epidemic of youth violence that began in the mid-1980s has been demographically concentrated among black male youths: the homicide-commission rate for this group increased by a factor of about 4.5. A number of patterns stand out: one of every four or five serious crimes of violence, and one of ten homicides, are committed by juveniles who are less than age eighteen; the proportion of arrests for violent crimes, however, that involved juveniles (20 percent) was about the same in 1994 as in 1965. A decline in the adolescent population has been balanced by an increase in rates of arrest. Youths kill more often than they are killed, and there is a great deal of crossover killing (in both directions) between adolescents and older people. The claim that the explosion in youth violence can be attributed to "superpredators," with each cohort having greater prevalence of such fiends than the last, does not accord well with available data. There is a clear indication of increased gun availability during the epidemic: every category of homicide, as well as other violent crimes, exhibited an increase in gun use. While the sizes of successive youth cohorts are increasing and will continue to do so for the foreseeable future, that is not a sound basis for predicting that the volume of youth violence will also increase.
revenge, while another may be associated with expansion of a drug business into new territory. Consequently, street gangs and the crimes in which they engage can- not be viewed as monolithic in nature. This Research in Brief describes these and other patterns of street gang-related vio- lence in a major U.S. city-Chicago. All available information, including Chicago police records of illegal street gang-moti- vated activity -from vandalism to drug offenses to violent offenses (both lethal and nonlethal)- was examined across time, neighborhood, and street gang affiliation. Individual, gang-level, and neighborhood- level characteristics were also analyzed to determine the relationships among these three factors. The results of the analysis give one of the most complete pictures of street gang crime available today.
Most police calls for service come from especially dangerous locations, or "hot spots." If risks at these locations are stable—hot spots stay hot—community problem-solving techniques may reduce crimes and disor- ders substantially. If locations run high risks only temporarily or sporadi- cally, location-based strategies may not work. Analysis of calls for service at high schools, housing projects, subway stations and parks in Boston shows that risks remain fairly constant over time; most apparent changes may be attributed to random processes. Autoregression and displacement in space and call type are statistically significant but unimportant indicators of call rates. In addition to verifying the effectiveness of community problem- solving strategies, these results have practical implications for problem-solv- ing techniques.
Why do some blocks have more crime, or their residents have higher fear levels, than other blocks? In an effort to answer this question we proposed a model that incorporated physical defensible space features, local social ties, and territorial functioning. The model was tested using data from a multistage, stratified sample of 687 Baltimore households on 63 blocks. At each household, surveys were completed and on-site physical features were photographed and subsequently rated. Records of police activity on each block were also obtained. Our model explained significant portions of crimes of violence to persons (18%) and block fear (37%). It was also able to predict a significant amount (13%) of the variation in individual-level fear. At the block level: defensible space features dampened crime and fear but not as strongly as expected; and local social ties dampened crime and fear directly, and indirectly via an enhancement of territorial functioning. A model predicting individual fear levels, controlling for block context, was also supported. Our successful modeling of block dynamics suggests that these entities may profitably be treated as small-scale social units or groups. The pattern of findings has also confirmed suggestions made by others that physical factors alone cannot be relied on to preserve local order and feelings of security. Finally, the block-level linkages between local social ties and territorial attitudes clarify how territorial attitudes reflect, and may contribute to, the development of group-based norms regarding appropriate behaviors in on-block settings.
The authors review research on police effectiveness in reducing crime, disorder, and fear in the context of a typology of innovation in police practices. That typology emphasizes two dimensions: one concerning the diversity of approaches, and the other, the level of focus. The authors find that little evidence supports the standard model of policing—low on both of these dimensions. In contrast, research evidence does support continued investment in police innovations that call for greater focus and tailoring of police efforts, combined with an expansion of the tool box of policing beyond simple law enforcement. The strongest evidence of police effectiveness in reducing crime and disorder is found in the case of geographically focused police practices, such as hot-spots policing. Community policing practices are found to reduce fear of crime, but the authors do not find consistent evidence that community policing (when it is implemented without models of problem-oriented policing) affects either crime or disorder. A developing body of evidence points to the effectiveness of problemoriented policing in reducing crime, disorder, and fear. More generally, the authors find that many policing practices applied broadly throughout the United States either have not been the subject of systematic research or have been examined in the context of research designs that do not allow practitioners or policy makers to draw very strong conclusions.
This research blends ideas and concepts from social disorganization theory with recent development in growth-curve methodology to examine the association between neighborhood structure and homicide level over time. Using 15 years of sequential data, we estimate growth-curve models that emphasize the effects of neighborhood socio-economic and demographic characteristics on changing levels of three types of homicide — General Altercation, Felony, and Domestic — in St. Louis from 1980-1994. The analyses reveal several important findings. First, trends for each subtype differ, with the shape of the trajectory depending on the type of homicide under consideration. Second, while neighborhood factors are significantly associated with trends in General Altercation, Felony and Domestic killings, the factors that influence the rate of change vary for each type of homicide.
The abstract for this document is available on CSA Illumina.To view the Abstract, click the Abstract button above the document title.
Operation Ceasefire is a problem-oriented policing intervention aimed at reducing youth homicide and youth firearms violence in Boston. It represented an innovative partnership between researchers and practitioners to assess the city's youth homicide problem and implement an intervention designed to have a substantial near-term impact on the problem. Operation Ceasefire was based on the “pulling levers” deterrence strategy that focused criminal justice attention on a small number of chronically offending gang-involved youth responsible for much of Boston's youth homicide problem. Our impact evaluation suggests that the Ceasefire intervention was associated with significant reductions in youth homicide victimization, shots-fired calls for service, and gun assault incidents in Boston. A comparative analysis of youth homicide trends in Boston relative to youth homicide trends in other major U.S. and New England cities also supports a unique program effect associated with the Ceasefire intervention.
We tested the block-level deterrent effects on crime of uniformed police raids of crack houses. Court-authorized raids were legally possible on 207 blocks with at least five calls for police service in the preceding 30 days. Raids were assigned randomly to 104 locations and were conducted at 98 of those sites; the other 109 were left alone. Experimental blocks, in relation to controls, showed reductions in both calls for service and offense reports, but effects were quite small and decayed in two weeks. Raids in which arrests were made (23 of 104 assigned) had no consistently different impact from raids in which no arrests were made. Raids had more effect on calls for service in the winter than in the spring, but we found little seasonal or period difference in effects of raids on offense reports. Alternative police methods may be far more cost-effective than raids in “harm reduction” for crack houses.
A developmental trajectory describes the course of a behavior over age or time. A group-based method for identifying distinctive groups of individual trajectories within the population and for profiling the characteristics of group members is demonstrated. Such clusters might include groups of "increasers," "decreasers," and "no changers." Suitably defined probability distributions are used to handle 3 data types—count, binary, and psychometric scale data. Four capabilities are demonstrated: (a) the capability to identify rather than assume distinctive groups of trajectories, (b) the capability to estimate the proportion of the population following each such trajectory group, (c) the capability to relate group membership probability to individual characteristics and circumstances, and (d) the capability to use the group membership probabilities for various other purposes such as creating profiles of group members. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
The possibility that homicides can spread from one geographic area toanother has been entertained for some time by social scientists, yetsystematic efforts to demonstrate the existence, or estimate the strength,of such a diffusion process are just beginning. This paper uses exploratoryspatial data analysis (ESDA) to examine the distribution of homicides in 78counties in, or around, the St. Louis metropolitan area for two timeperiods: a period of relatively stable homicide (1984–1988) and aperiod of generally increasing homicide (1988–1993). The findingsreveal that homicides are distributed nonrandomly, suggestive of positivespatial autocorrelation. Moreover, changes over time in the distribution ofhomicides suggest the possible diffusion of lethal violence out of onecounty containing a medium-sized city (Macon County) into two nearbycounties (Morgan and Sangamon Counties) located to the west. Althoughtraditional correlates of homicide do not account for its nonrandom spatialdistribution across counties, we find some evidence that more affluentareas, or those more rural or agricultural areas, serve as barriers againstthe diffusion of homicides. The patterns of spatial distribution revealedthrough ESDA provide an empirical foundation for the specification ofmultivariate models which can provide formal tests for diffusion processes.
Analysis of variance (ANOVA) is an extremely important method in exploratory and confirmatory data analysis. Unfortunately, in complex problems (e.g., split-plot designs), it is not always easy to set up an appropriate ANOVA. We propose a hierarchical analysis that automatically gives the correct ANOVA comparisons even in complex scenarios. The inferences for all means and variances are performed under a model with a separate batch of effects for each row of the ANOVA table. We connect to classical ANOVA by working with finite-sample variance components: fixed and random effects models are characterized by inferences about existing levels of a factor and new levels, respectively. We also introduce a new graphical display showing inferences about the standard deviations of each batch of effects. We illustrate with two examples from our applied data analysis, first illustrating the usefulness of our hierarchical computations and displays, and second showing how the ideas of ANOVA are helpful in understanding a previously fit hierarchical model.
* List of Tables and Figures * Acknowledgments * List of Acronyms Part 1: Background on the Place, the Theory, and Policies * 1. Introduction * Philadelphia, Pennsylvania * Oakland, California * Focus * Incivilities, Disorder, Social Disorganization, Collective Efficacy, and Social Capital * Broader Theoretical and Empirical Context of Current Approaches * Evidence * The Argument and the Chapters Ahead * The "Bottom Line," * Notes * References * 2. The Baltimore Context, and Its Context * With Charles David Linne * The War Has Been Won? * Purpose * Changes in People, Housing, and Jobs * Changes in Crime: The City as a Whole * Baltimore Neighborhood Crime Rates * Shifting Incivilities, 1981-1994 * Summary Comments on Changes * The Questions of Fear and Neighborhood Problems * In the News * Notes * References * 3. The Incivilities Thesis: Theory, Measurement, and Policy * Organization * Variations on a Theme * Empirical Support for Hypotheses * A Theoretical Aside on Demographic and Structural Issues * From Theory to Research: Incivilities Indicators * Implications for Policy, Practice, and Theory * Notes * References Part 2: Quantitative Evidence on Origins and Impacts * 4. Origins of Incivilities * A Story About One Broken Window * Focus and Organization * Perspectives on the Origins of Incivilities * An Unexciting, but Necessary, Methodological Aside on Change * Another Necessary, but Unexciting, Aside on Multilevel Models * Overview of Indicators, Outcomes, and Controls * Incivilities Observed * Incivilities As Perceived by Residents * Discussion * Notes * References * 5. Impacts of Incivilities on Later Crime and Decline * A Systemic Perspective * Focus * Data and Analysis, * Changes on Decline Indicators in the 1980s * Crime Rate Changes * Predicting Decline * Discussion * Notes * References * 6. Longitudinal Impacts of Incivilities on Reactions to Crime and Local Commitment * Reactions to Crime * Focus * Data and Analysis * Impacts of Specific Predictors * Closing Thoughts, * Support for Longitudinal Impacts of Incivilities * Notes * References Part 3: Qualitative Evidence from Community Leaders * 7. The Community Perspective: Views About Incivilities and Responses to Incivilities in the Context of Collective Crime Prevention Initiatives * Organization of the Chapter and Questions Addressed * What Influences the Type of Collective Strategies Adopted? Podolefsky's Model * Data Sources * Responses to Drug Sales and Use and Related Crime Problems * Neighborhood Fabric and Responses to Crime and Drug Sales and Use * Closing Comments * Appendix: Sample Selection Procedures and Contact Attempts * Notes * References * 8. Place Power and Implications for Coproduced Safety: Changes and Stability in Neighborhood Names, Boundaries, and Organizations * Neighborhood Mapping and Current Data Sources * Organization * Naming and Bounding * Service Delivery Issues and Community Policing * Stability and Changes * Implications: Can Police-Community Partnerships Organize Around Neighborhood Units? * Summary * Notes * References * 9. Closing Thoughts * Context and Ironies * Does the Theory Get Support? * The Context Outside the Theory, * Notes * References * Index
Analysis of variance (ANOVA) is an extremely important method in exploratory and confirmatory data analysis. Unfortunately, in complex problems (e.g., split-plot designs), it is not always easy to set up an appropriate ANOVA. We propose a hierarchical analysis that automatically gives the correct ANOVA comparisons even in complex scenarios. The inferences for all means and variances are performed under a model with a separate batch of effects for each row of the ANOVA table. We connect to classical ANOVA by working with finite-sample variance components: fixed and random effects models are characterized by inferences about existing levels of a factor and new levels, respectively. We also introduce a new graphical display showing inferences about the standard deviations of each batch of effects. We illustrate with two examples from our applied data analysis, first illustrating the usefulness of our hierarchical computations and displays, and second showing how the ideas of ANOVA are helpful in understapding a previously fit hierarchical model.
Change is constant in everyday life. Infants crawl and then walk, children learn to read and write, teenagers mature in myriad ways, and the elderly become frail and forgetful. Beyond these natural processes and events, external forces and interventions instigate and disrupt change: test scores may rise after a coaching course, drug abusers may remain abstinent after residential treatment. By charting changes over time and investigating whether and when events occur, researchers reveal the temporal rhythms of our lives. This book is concerned with behavioral, social, and biomedical sciences. It offers a presentation of two of today's most popular statistical methods: multilevel models for individual change and hazard/survival models for event occurrence (in both discrete- and continuous-time). Using data sets from published studies, the book takes you step by step through complete analyses, from simple exploratory displays that reveal underlying patterns through sophisticated specifications of complex statistical models.
In recent years, researchers have argued that police actions should be focused on high-risk crime places rather than spread thinly across the urban landscape. This review examines the available evaluation evidence on the effects of concentrating police enforcement efforts on crime hot spots. Five randomized experiments and four nonequivalent control group quasi-experiments were identified. The findings of these evaluations suggest that focused police actions can prevent crime and disorder in crime hot spots. These studies also suggest that focused police actions at specific locations do not necessarily result in crime displacement. Unintended crime prevention benefits were also associated with the hot spots policing programs. Although these evaluations reveal that these programs work in preventing crime, additional research is needed to unravel other important policy-relevant issues such as community reaction to focused police enforcement efforts.
Property offenders construct cognitive images of the physical environment to decide where to commit crimes. At least three levels of target selection occur-neighborhoods, or regions; street blocks; and specific sites. Information on the physical characteristics of neighborhoods, such as ease of entry and exit, the number of internal boundaries limiting ease of circulation, and signs of guardianship or of incivilities, are weighed by the offender to determine risks, opportunities, and conveniences. This framework for understanding links between offenders and the physical environment provides a typology for evaluating research on environmental design and crime prevention, especially research on defensible-space theory, territorial perspectives, and the incivilities thesis. The incivilities thesis suggests that offenders take into account social and physical incivilities when deciding which neighborhood to choose as crime targets. Three recent studies involving neighborhoods in Atlanta, Georgia, Baltimore, Maryland, and Hartford, Connecticut, show that the effects of environmental design on crime range from small to moderate. Available research suggests that crime prevention efforts should be directed at the street-block rather than at the neighborhood level of analysis.
Violent crime in the United States has traveled a striking path in the past 20 years, from a sharp spike in the late 1980s to a dramatic drop beginning around 1993 that lasted until a flattening occurred at the beginning of the new century. The upturn of the 1980s was driven almost exclusively by a sharp increase in gun violence among young people, whereas the downturn was a joint product of a turndown in youth violence and a steady drop in violence among adults over 30 that has been ongoing for at least 25 years. These trends have intrigued lay observers and students of crime alike, with advocates of various explanations of crime arguing that the key for the rise, fall, or both lies with one or another favored determinant: prisons, drugs, guns, policing, economics, or demography, including abortion. In this review, we discuss both the crime trends and their candidate explanations, concluding that the most reasonable accounts involve a complex interaction among several of these factors.
We tested the hypothesis that greater enforcement of existing laws against carrying concealed weapons could reduce firearms violence with a quasi-experimental, target beat/comparison beat design. Over a six-month period in a ten-by-eight-block area with a homicide rate 20 times higher than the national average, intensive patrol near gun crime hot spots produced a 65 percent increase in firearms seized by police. Gun crimes declined in the target area by 49 percent, with no significant displacement to any patrol beat surrounding the target area. Neither gun crimes nor guns seized changed significantly in the comparison beat several miles away.
A developmental trajectory describes the course of a behavior over age or time. A group-based method for identifying distinctive groups of individual trajectories within the population and for profiling the characteristics of group members is demonstrated. Such clusters might include groups of "increasers," "decreasers," and "no changers." Suitably defined probability distributions are used to handle 3 data types-count, binary, and psychometric scale data. Four capabilities are demonstrated: (a) the capability to identify rather than assume distinctive groups of trajectories, (b) the capability to estimate the proportion of the population following each such trajectory group, (c) the capability to relate group membership probability to individual characteristics and circumstances, and (d) the capability to use the group membership probabilities for various other purposes such as creating profiles of group members.
Minority gangs of the 1980s are examined as groups of juveniles and young adults and as a fraction of the forming underclass. Gangs today no longer disappear as their ethnic groups rises socially. Instead, gangs are becoming institutionalized in the central cities. This discussion of prior research in Los Angeles (California) and more recent studies in Milwaukee (Wisconsin) demonstrates that gangs are quite different than they were in the past. Underclass in this discussion refers to people permanently excluded from participation in mainstream occupations who survive through a combination of economic resources that can include welfare and profits from crime. In neither of the Chicano communities studied in Los Angeles or Milwaukee have the gangs become involved in an organized crime infrastructure. Instead, a fraction of each gang clique moves into the underclass as the gang becomes an institutionalized feature of some poverty communities and plays a role in the perpetuation of the underclass. In both cities gangs have also become support groups in correctional settings. Interviews with 47 gang members from 19 of Milwaukee's gangs form the basis of this exploration. Five appendixes present the study questionnaire and four tables of gang data. (Contains 3 figures, 6 tables, 4 appendix tables, and 144 references.) (SLD)
This book presents the results of 20 years of ecological research into the nature of the relationship between the distribution of delinquency and the pattern of physical structure and social organization of 21 American cities. Uniform findings in every city confirm the hypothesis that the physical deterioration of residential areas accompanied by social disorganization is greatest in a central zone in the business district, intermediate in a middle zone, and lowest in the other zones, and that there is a progressive decline in the incidence of delinquency from the innermost zone where it is most concentrated to the peripheral areas. Delinquency is found to be highly correlated with changes in population, inadequate housing, poverty, presence of Negroes and foreign-born, tuberculosis, mental disorders, and adult criminality. The common basic factor is social disorganization or the absence of community effort to cope with these conditions. Causation of juvenile delinquency is to be sought more in terms of the community than of the individual. 107 maps pertaining to the cities studied and 118 tables relating to population and delinquency rates are included as well as a chapter describing the Chicago Area Project as a demonstration of the effective mobilization of community forces to combat delinquency and crime. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
The volume's 19 chapters and 2 appendices summarize the author's research in his sociological analysis of criminal homicide in which he used Philadelphia as a community case study. "Analysis has been made of 588 criminal homicides listed by the police in this city between January 1, 1948, and December 31, 1952. A critical review of the important homicide literature in this country is provided, and whenever feasible, comparison is made of criminal homicides in Philadelphia with research elsewhere." Consideration is given to such problems as alcohol, motivation, temporal and spatial patterns. 4 chapters discuss the relationship between the victim and the offender. This sociological work is held to be of major interest for the criminologist and the police administrator. 20-page references. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Several recent studies have used records of calls-for-service (CFS) to police 911 centers to measure crime at the address, neighborhood, and city level. This article examines the limitations of this “new” indicator of crime. After pointing out several types of error in dispatch records, we use data from an observational study of policing in 60 neighborhoods to examine empirically how these errors might bias CFS-based crime counts and discuss the consequences of such bias. We conclude with suggestions for future research on the validity of CFS as an indicator of crime.
Research Summary: This evaluation of a directed police patrol project utilizes a pre-post quasi-experimental design with a non-equivalent control group as well as an interrupted time series analysis. The results suggest that directed patrol had an impact on firearms crime in one of the target areas but not the other.
Policy Implications: The results suggest that a specific deterrence strategy whereby the police utilize directed patrol to focus on suspicious activities and locations reduced violent gun crime. In contrast, a general deterrence strategy focused on maximizing vehicle stops did not have an apparent effect.
We merge Exploratory Spatial Data Analysis (ESDA) and a semi-parametric, group-based trajectory procedure (TRAJ) to classify communities in Chicago by violence trajectories across space. Total, street gun and other weapon homicide trajectories are identified across 831 census tracts between 1980 and 1995. We find evidence consistent with a weapon substitution effect in violent neighborhoods that are proximate to one another, a defensive diffusion effect of exclusively street gun-specific homicide increases in neighborhoods bordering the most violent areas, and a spatial decay effect of temporal homicide trends in which the most violent areas are buffered from the least violent by places experiencing mid-range levels of lethal violence over time. In merging these two methods of data analysis, we provide a more efficient way to describe both spatial and temporal trends and make significant advances in furthering applications of space-time methodologies.
Drawing on the work of Jacobs, Newman, and Gardiner, among others, this paper investigates fear of crime by urban residents as a consequence of two interrelated characteristics of neighborhoods: 1) the perceived volume of street usage and 2) the degree of residents' social integration into the neighborhood. Secondary analysis of a 1975 survey shows that, counter to previous hypotheses, perception of increased street traffic leads to greater fear. However, when controlling for social integration, we find that for those who are socially integrated perceived volume of street traffic has no relationship to fear, while for those not socially integrated the greater the perceived street usage the greater the fear. Three mechanisms by which social integration may reduce fear of people on the streets are considered: 1) reducing the proportion of strangers versus acquaintances on the street; 2) providing networks of potential assistance; and 3) reducing the strangeness of the streets' daily rhythms and routines. We conclude that both physical design and social factors must be interrelated in attempts to understand fear of crime and in designing ameliorative programs.
This article addresses three issues that are central to the criminal career debate. First, is the life course of individual offending patterns marked by distinctive periods of quiescence? Second, at the level of the individual, do offending rates vary systematically with age? In particular, is the age-crime curve single peaked or flat? Third, are chronic offenders different from less active offenders? Do offenders themselves differ in systematic ways? Using a new approach to the analysis of individual criminal careers—based on nested, mixed Poisson models in which the mixing distribution is estimated nonparametrically—we analyze a panel data set that tracks a sample of males for more than 20 years. Our results provide empirical evidence in support of some features of criminal propensity theory and some in support of conventional criminal careers theory. In support of latent-trait criminal propensity theory, the individual-level average offense rate (per unit of time) varies as a function of observable individual-level characteristics and unobservable heterogeneity among individuals, and the age trajectory of the offense rate is generally single peaked rather than flat. On the other hand, in support of conventional criminal careers theory, models that incorporate a parameter that permits periods of active as well as inactive offending across age have greater explanatory power than those that do not. In addition, the nonparametric, discrete approximation to the population distribution of unobservable heterogeneity in the individual-level mean offense rate facilitates identification of four classes of offenders—nonoffenders as well as individual-level characteristics that are unique to each group. Problems of theoretical explanation and empirical generalizability of these results are described.
The experience, observations, local knowledge, and historical perspective of working police officers and others with routine contact with offenders, communities, and criminal organizations may represent an im-portant underutilized resource for describing, understanding, and crafting interventions aimed at crime problems. Mapping and other information-collecting and -ordering techniques, usually aimed at formal police data, can also be used to good effect to capture and organize these experiential assets. This chapter describes one such exercise carried out as part of a project to apply problem-solving techniques to youth gun violence and gun markets in Boston. A working group comprised of Harvard University re-searchers, police officers from the Boston Police Department's Youth Vio-lence Strike Force, probation officers covering high-risk neighborhoods, and city-employed gang-mediation "street workers": estimated the number and size of the city's gangs; mapped their turf; mapped their antagonisms and alliances; and classified five years of youth victimization events ac-cording to their connection (or lack thereof) to this gang geography. The products of these exercises provide: a "snapshot" of Boston's gang turf; an estimate of gang involvement in high-risk neighborhoods; a sociogram of gang relationships; and an estimate of Boston gangs' direct contribution to youth homicide victimization.
A leading sociological theory of crime is the “routine activities” approach (Cohen and Felson, 1979). The premise of this ecological theory is that criminal events result from likely offenders, suitable targets, and the absence of capable guardians against crime converging nonrandomly in time and space. Yet prior research has been unable to employ spatial data, relying instead on individual- and household-level data, to test that basic premise. This analysis supports the premise with spatial data on 323,979 calls to police over all 115,000 addresses and intersections in Minneapolis over 1 year. Relatively few “hot spots” produce most calls to Police (50% of calls in 3% of places) and calls reporting predatory crimes (all robberies at 2.2% of places, all rapes at 1.2% of places, and all auto thefts at 2.7% of places), because crime is both rare (only 3.6% of the city could have had a robbery with no repeat addresses) and concentrated, although the magnitude of concentration varies by offense type. These distributions all deviate significantly, and with ample magnitude, from the simple Poisson model of chance, which raises basic questions about the criminogenic nature of places, as distinct from neighborhoods or collectivities.
Studies of crime at micro places have generally relied on cross-sectional data and reported the distributions of crime statistics over short periods of time. In this paper we use official crime data to examine the distribution of crime at street segments in Seattle, Washington, over a 14-year period. We go beyond prior research in two ways. First, we view crime trends at places over a much longer period than other studies that have examined micro places. Second, we use group-based trajectory analysis to uncover distinctive developmental trends in our data. Our findings support the view that micro places generally have stable concentrations of crime events over time. However, we also find that a relatively small proportion of places belong to groups with steeply rising or declining crime trajectories and that these places are primarily responsible for overall city trends in crime. These findings are particularly important given the more general decline in crime rates observed in Seattle and many other American cities in the 1990s. Our study suggests that the crime drop can be understood not as a general process that occurred across the city landscape but one that was generated in a relatively small group of micro places with strong declining crime trajectories over time.
This article proposes a new method for examining dynamic changes in thespatial distribution of a phenomenon. Recently introduced exploratoryspatial data analysis (ESDA) techniques provide social scientists with anew set of tools for distinguishing between random and nonrandom spatialpatterns of events (Anselin, 1998). Existing ESDA measures, however, arestatic and do not permit comparisons of distributions of events in the samespace but across different time periods. One ESDA method—the Moranscatterplot—has special heuristic value because it visually displayslocal spatial relationships between each spatial unit and its neighbors. Weextend this static cross-sectional view of the spatial distribution ofevents to consider dynamic features of changes over time in spatialdependencies. The method distinguishes between contagious diffusion betweenadjoining units and hierarchical diffusion that spreads broadly throughcommonly shared influences. We apply the method to homicide data, lookingfor evidence of spatial diffusion of youth-gang homicides acrossneighborhoods in a city. Contagious diffusion between neighboring censustracts is evident only during the year of peak growth in total homicides,when high local rates of youth-gang homicides are followed by significantincreases in neighboring youth- nongang rates. This pattern is consistentwith a spread of homicides from gang youth to nongang youth. Otherwise, theincreases in both youth-gang and youth- nongang homicides generally occursimultaneously in nonneighboring tracts.