Showing papers in "Journal of Quantitative Criminology in 2021"
TL;DR: In this article, the authors report the first experimental evidence on the effect of street lighting on crime and find evidence that communities that were assigned more lighting experienced sizable reductions in nighttime outdoor index crimes.
Abstract: This paper offers novel experimental evidence that violent crimes can be successfully reduced by changing the situational environment that potential victims and offenders face. We focus on a ubiquitous but understudied feature of the urban landscape—street lighting—and report the first experimental evidence on the effect of street lighting on crime. Through a unique public partnership in New York City, temporary street lights were randomly allocated to 40 of the city’s public housing developments. We find evidence that communities that were assigned more lighting experienced sizable reductions in nighttime outdoor index crimes. We also observe a large decline in arrests indicating that deterrence is the most likely mechanism through which the intervention reduced crime. Results suggests that street lighting, when deployed tactically, may be a means through which policymakers can control crime without widening the net of the criminal justice system.
53 citations
TL;DR: It is illustrated how a machine learning algorithm, Random Forests, can provide accurate long-term predictions of crime at micro places relative to other popular techniques, and how recent advances in model summaries can help to open the ‘black box’ of Randomforests, considerably improving their interpretability.
Abstract: We illustrate how a machine learning algorithm, Random Forests, can provide accurate long-term predictions of crime at micro places relative to other popular techniques. We also show how recent advances in model summaries can help to open the ‘black box’ of Random Forests, considerably improving their interpretability. We generate long-term crime forecasts for robberies in Dallas at 200 by 200 feet grid cells that allow spatially varying associations of crime generators and demographic factors across the study area. We then show how using interpretable model summaries facilitate understanding the model’s inner workings. We find that Random Forests greatly outperform Risk Terrain Models and Kernel Density Estimation in terms of forecasting future crimes using different measures of predictive accuracy, but only slightly outperform using prior counts of crime. We find different factors that predict crime are highly non-linear and vary over space. We show how using black-box machine learning models can provide accurate micro placed based crime predictions, but still be interpreted in a manner that fosters understanding of why a place is predicted to be risky.
32 citations
TL;DR: In this article, the authors investigate the influence of neighborhood-level criminal opportunity on the relationship between crime generators and block-level crime, and find that the effects of crime generators on crime on blocks are significantly tempered in census block groups with high levels of civic engagement.
Abstract: The present study tests hypotheses regarding the moderating influence of neighborhood-level criminal opportunity on the relationship between crime generators and block-level crime. We first estimated multilevel negative binomial regression models for violent, property, and drug crimes to identify crime-type specific crime generators on each block. We then estimated a series of crime-type specific models to examine whether the effects of violent, property, and drug crime generators are moderated by three census block group-level indicators of neighborhood criminal opportunity—concentrated disadvantage, vehicular traffic activity, and civic engagement. The positive relationship between crime generators and crime on blocks was exacerbated in census block groups with high levels of concentrated disadvantage and high levels of traffic activity for all three crime types. The effects of crime generators on block-level crime were significantly tempered in census block groups with high levels of civic engagement. Particular place types do not generate crime similarly across varying neighborhood contexts. Rather, the criminogenic effects of micro-places appear to be exacerbated in neighborhoods with extensive criminal opportunity and tempered in neighborhoods with less criminal opportunity.
32 citations
TL;DR: In this article, a two-wave panel survey of a nationally representative sample of Australian adults measured people's beliefs about police trustworthiness (procedural fairness and effectiveness), their duty to obey the police, their contact with the police between the two waves, and their evaluation of those encounters in terms of process and outcome.
Abstract: Test the asymmetry thesis of police-citizen contact that police trustworthiness and legitimacy are affected more by negative than by positive experiences of interactions with legal agents by analyzing changes in attitudes towards the police after an encounter with the police. Test whether prior attitudes moderate the impact of contact on changes in attitudes towards the police. A two-wave panel survey of a nationally representative sample of Australian adults measured people’s beliefs about police trustworthiness (procedural fairness and effectiveness), their duty to obey the police, their contact with the police between the two waves, and their evaluation of those encounters in terms of process and outcome. Analysis is carried out using autoregressive structural equation modeling and latent moderated structural models. The association between both process and outcome evaluation of police-citizen encounters and changes in attitudes towards the police is asymmetrical for trust in police effectiveness, symmetrical for trust in procedural fairness, and asymmetrical (in the opposite direction expected) for duty to obey the police. Little evidence of heterogeneity in the association between encounters and trust in procedural fairness and duty to obey, but prior levels of perceived effectiveness moderate the association between outcome evaluation and changes in trust in police effectiveness. The association between police-citizen encounters and attitudes towards the police may not be as asymmetrical as previously thought, particularly for changes in trust in procedural fairness and legitimacy. Policy implications include considering public-police interactions as ‘teachable moments’ and potential sources for enhancing police trustworthiness and legitimacy.
26 citations
TL;DR: In this paper, the authors examined whether the law of crime concentration applies in the context of sub-Saharan Africa using primary data and found that crime was concentrated at all spatial scales, and having accounted for expectation, given the distribution of opportunity, crime was most concentrated at the household level, closely followed by street segments.
Abstract: Research demonstrates that crime is concentrated. This finding is so consistent that David Weisburd refers to this as the “law of crime concentration at place”. However, most research on crime concentration has been conducted in the US or European cities and has used secondary data sources. In this study, we examine whether the law of crime concentration applies in the context of sub-Saharan Africa using primary data. A crime victimization survey was used to collect data in the city of Kaduna (Nigeria). Using these data, the concentration of crime (breaking-and-entering and domestic theft) was examined at the household, street segment, and neighborhood levels. Specifically, variants of a Lorenz curve and the Gini index (GI) were used to examine whether crime concentrates at these different spatial scales and if such concentration reflects anything beyond the spatial distribution of opportunity for these types of offenses. Crime was found to concentrate at all spatial scales, and having accounted for expectation, given the distribution of opportunity, crime was most concentrated at the household level, closely followed by street segments. It was relatively less concentrated at the neighborhood level. The current study extends previous research in a number of ways. It shows that the law of crime concentration at place applies in a very different context to most previous work. Unlike previous studies, we use primary data collected specifically to test the law, avoiding problems associated with the dark figure of crime. Moreover, the findings persist after accounting for crime opportunity.
25 citations
TL;DR: The authors used a multivariate self-exciting point process model to estimate the extent of contagious spread of violent crime for both gang-related and non-gang aggravated assaults and homicides in recent data from Los Angeles.
Abstract: Gangs are thought to enhance participation in violence. It is expected then that gang-related violent crimes trigger additional crimes in a contagious manner, above and beyond what is typical for non-gang violent crime. This paper uses a multivariate self-exciting point process model to estimate the extent of contagious spread of violent crime for both gang-related and non-gang aggravated assaults and homicides in recent data from Los Angeles. The degree of contagious cross-triggering between gang-related and non-gang violent crime is also estimated. Gang-related violence triggers twice as many offspring events as non-gang violence and there is little or no cross-triggering. Gang-related offspring events are significantly more lethal than non-gang offspring events, but no more lethal than non-contagious background gang crimes. Contagious spread of gang-related violent crime is different from contagion in non-gang violence. The results support crime prevention policies that target the disruption of gang retaliations.
24 citations
TL;DR: In this paper, the effects of victimization on several aspects of well-being in a longitudinal study of a general population sample were examined, including the effect of crime, generalized trust, and neighborhood satisfaction.
Abstract: We examined the effects of victimization on several aspects of well-being in a longitudinal study of a general population sample. Previous research has often been inconclusive, as it was largely based on cross-sectional data and prone to problems of unobserved heterogeneity and selection bias. We examined both between-person differences and within-person changes in well-being in relation to property and violent victimization. We investigated psychological and behavioral dimensions of well-being, controlling for and comparing with the effects of other negative life events. We used data from a two-wave panel survey of 2928 respondents aged 25–89 nested in 140 neighborhoods in two large German cities. We applied random-effects modeling to separate between-person from within-person effects. The within-person detrimental effects of victimization were considerably smaller than between-person effects, which reflected preexisting, time-stable factors that distinguish individuals who have experienced victimization from individuals who have not. Detrimental effects concerned fear of crime, generalized trust, and neighborhood satisfaction, but did not extend to emotional well-being or life satisfaction, in contrast to other negative life events. We found empirical support both for adaptation (‘recovery’) effects as well as for anticipation effects. Violent victimization had stronger effects than property victimization, and victimization near the home had stronger effects than victimization elsewhere. The findings indicate that violent victimization has palpable detrimental effects on security perceptions, trust and neighborhood satisfaction—but not on emotional well-being and life satisfaction—and that individuals largely recover from the victimization within 18 months.
23 citations
TL;DR: In this article, the authors used Google Street View images from every 20 meters in every street segment in the city of Santa Ana, CA, and then used machine learning to detect features of the environment.
Abstract: Despite theoretical interest in how dimensions of the built environment can help explain the location of crime in micro−geographic units, measuring this is difficult. This study adopts a strategy that first scrapes images from Google Street View every 20 meters in every street segment in the city of Santa Ana, CA, and then uses machine learning to detect features of the environment. We capture eleven different features across four main dimensions, and demonstrate that their relative presence across street segments considerably increases the explanatory power of models of five different Part 1 crimes. The presence of more persons in the environment is associated with higher levels of crime. The auto−oriented measures—vehicles and pavement—were positively associated with crime rates. For the defensible space measures, the presence of walls has a slowing negative relationship with most crime types, whereas fences did not. And for our two greenspace measures, although terrain was positively associated with crime rates, vegetation exhibited an inverted−U relationship with two crime types. The results demonstrate the efficacy of this approach for measuring the built environment.
22 citations
TL;DR: In this article, an incident rate ratio and relative incident rate rate ratio effect size and associated overdispersion parameter are developed and advocated as the preferred effect size for count-based outcomes in impact evaluations and meta-analyses of such studies.
Abstract: Area-based prevention studies often produce results that can be represented in a 2-by-2 table of counts. For example, a table may show the crime counts during a 12-month period prior to the intervention compared to a 12-month period during the intervention for a treatment and control area or areas. Studies of this type have used either Cohen’s d or the odds ratio as an effect size index. The former is unsuitable and the latter is a misnomer when used on data of this type. Based on the quasi-Poisson regression model, an incident rate ratio and relative incident rate ratio effect size and associated overdispersion parameter are developed and advocated as the preferred effect size for count-based outcomes in impact evaluations and meta-analyses of such studies.
21 citations
TL;DR: In this paper, a survey of 1,057 individuals in 98 relatively high-crime English neighborhoods defined at a small spatial scale measured willingness to cooperate using a hypothetical crime vignette and legitimacy using indicators of normative alignment between police and citizen values.
Abstract: Objectives: Test whether cooperation with the police can be modelled as a place-based norm that varies in strength from one neighborhood to the next. Estimate whether perceived police legitimacy predicts an individual’s willingness to cooperate in weak-norm neighborhoods, but not in strong-norm neighborhoods where most people are either willing or unwilling to cooperate, irrespective of their perceptions of police legitimacy.
Methods: A survey of 1,057 individuals in 98 relatively high-crime English neighborhoods defined at a small spatial scale measured (a) willingness to cooperate using a hypothetical crime vignette and (b) legitimacy using indicators of normative alignment between police and citizen values. A mixed-effects, location-scale model estimated the cluster-level mean and cluster-level variance of willingness to cooperate as a neighborhood-level latent variable. A cross-level interaction tested whether legitimacy predicts individual-level willingness to cooperate only in neighborhoods where the norm is weak.
Results: Willingness to cooperate clustered strongly by neighborhood. There were neighborhoods with (i) high mean and low variance, (ii) high mean and high variance, (iii) (relatively) low mean and low variance, and (iv) (relatively) low mean and high variance. Legitimacy was only a positive predictor of cooperation in neighborhoods that had a (relatively) low mean and high variance. There was little variance left to explain in neighborhoods where the norm was strong.
Conclusions: Findings support a boundary condition of procedural justice theory: namely, that cooperation can be modelled as a place-based norm that varies in strength from neighborhood to neighborhood and that legitimacy only predicts an individual’s willingness to cooperate in neighborhoods where the norm is relatively weak.
20 citations
TL;DR: In this article, the authors used the Google Vision API to extract objects from sampled Google Street View (GSV) images in each census block of Los Angeles and then compute indices for object diversity, diversity related to commonly employed census variables, and crime diversity from reports provided by the Los Angeles Police Department.
Abstract: Crime diversity is a measure of the variety of criminal offenses in a local environment, similar to ecological diversity. While crime diversity distributions have been explained via neutral models, to date the environmental and social mechanisms behind crime diversity have not been investigated. Building on recent work demonstrating that crime rates can be inferred from street level imagery with neural network computer vision models, in this paper we consider the task of inferring crime diversity through street level imagery. We use the Google Vision API, a deep learning image tagging service, to extract objects from sampled Google Street View (GSV) images in each census block of Los Angeles. For each census block we then compute indices for (1) object diversity, (2) diversity related to commonly employed census variables, and (3) crime diversity from reports provided by the Los Angeles Police Department. We then build ordinary least squares and geographically weighted regression models to explain crime diversity as a function of environmental diversity, population diversity, and population size. We show that crime diversity arises via a combination of environmental diversity (as measured through street view object diversity), household diversity (as measured through the census), and population size. Population size and area of the census block both lend credence to the neutral model proposed by Brantingham for crime diversity. However, environmental and demographic diversity combined play an equally important role in explaining variation in crime diversity. Our study has two primary implications for research on crime and place. First, Google Street View (via the Google Vision API) can provide important, cost-effective empirical insights to best understand distinct geographic environments of crime. Second, environmental diversity, as measured by image tagging in GSV, was observed to be more predictive of crime diversity (variety of crime types) than commonly used census measures.
TL;DR: In this paper, the authors examined whether unstructured socializing with specific friends can explain within-individual changes in adolescents' degree of specialization in delinquency and substance use, and they found that involvement in un-structured socially-oriented activities with friends who steal, vandalize, commit violence, use alcohol, use cigarettes, or use drugs enhances adolescents' risks for engaging in those respective behaviors.
Abstract: Despite abundant attention to offending specialization in criminology, scholars have only recently started to explore opportunity-driven explanations for within-individual patterns of specialization. The current study examines whether unstructured socializing with specific friends can explain within-individual changes in adolescents’ degree of specialization in delinquency and substance use. Data were derived from the PROSPER Peers Project, a longitudinal study consisting of five waves of data on 11,183 adolescents (aged 10 to 17). The data include self-reports about engagement in delinquency and substance use, sociometric information, and information on the time respondents reported spending in unstructured socializing with their nominated friends. Hypotheses were tested with negative binomial and binomial logit multilevel models. The findings indicate that involvement in unstructured socializing with friends who steal, vandalize, commit violence, use alcohol, use cigarettes, or use drugs enhances adolescents’ risks for engagement in those respective behaviors. Such activity affects adolescents’ quantitative engagement as well as their level of specialization in these behaviors. The study indicates that routine activity—in particular involvement in unstructured socializing—explains within-individual changes in deviance specialization among adolescents. Thus, exposure to opportunities can explain why adolescents specialize in certain types of delinquency and substance use in one time-period, and in other types of behavior in other time-periods. This adds a proximate explanation for this phenomenon to other explanations that focus on local life circumstances and peer group affiliation.
TL;DR: In this paper, the authors explore the influence of micro-facilities (e.g., pubs and fast-food restaurants) and superfacilities on area level counts of crime.
Abstract: The aim of this study was to explore the influence of “micro-” (e.g., pubs and fast-food restaurants) and “super-facilities” on area level counts of crime. Soccer stadia were selected as an example of a super-facility as their episodic use provides conditions not unlike a natural experiment. Of particular interest was whether the presence of such facilities, and their influence on the flow of people through neighborhoods on match days affects crime. Consideration was also given to how the social composition of a neighborhood might influence crime. Crime, street network, and points of interest data were obtained for the areas around five UK soccer stadia. Counts of crime were computed for small areal units and the spatial distribution of crime examined for match and non-match days. Variables derived from graph theory were generated to estimate how micro-facilities might influence the movement flows of people on match days. Spatial econometric analyses were used to test hypotheses. Mixed support was found for the influence of neighborhood social composition on crime for both match and non-match days. Considering the influence of facilities, a selective pattern emerged with crime being elevated in those neighborhoods closest to stadia on match but not non-match days. Micro-facilities too were found to influence crime levels. Particularly clear was the finding that the influence of pubs and fast-food restaurants on estimated movement flows to and from stadia on match (but not non-match) days was associated with area level crime. Our findings provide further support for ecological theories of crime and how factors that influence the likely convergence of people in urban spaces affect levels of crime.
TL;DR: This paper examined the time-varying consequences of paternal incarceration for adolescent behavior and found that exposure to paternal incarceration during early childhood, but not during middle childhood or early adolescence, is positively associated with behavior problems, partially explained by family adversities stemming from paternal incarceration.
Abstract: I draw on general strain theory, a framework often used to understand adolescent behavior, and augment it with aspects of the stress process perspective to examine the time-varying consequences of paternal incarceration for adolescent behavior. I use six waves of data from the Fragile Families and Child Wellbeing Study, a cohort of children born around the turn of the twenty-first century, and inverse probability of treatment weighting models to estimate the time-varying relationship between paternal incarceration and adolescent behavior problems and the mechanisms underlying this relationship. Results document three main findings. First, adolescents exposed to paternal incarceration at any point in the life course have more behavior problems than their counterparts not exposed to paternal incarceration. Second, exposure to paternal incarceration during early childhood, but not during middle childhood or early adolescence, is positively associated with behavior problems. Third, this relationship is partially explained by family adversities stemming from paternal incarceration. This research builds on our criminological understanding of how strains, such as paternal incarceration, can facilitate inequalities in adolescent behavior by considering dynamic selection into paternal incarceration, the time-varying repercussions of paternal incarceration, and the mechanisms linking paternal incarceration to adolescent behavior. Early life course paternal incarceration facilitates chains of adversity that accumulate throughout early childhood, middle childhood, and adolescence.
TL;DR: Techniques to incorporate two types of proximity, geographic proximity and commuting proximity in spatial generalized linear mixed models (SGLMM) in order to estimate domestic and sexual violence in Detroit, Michigan and Arlington County, Virginia show that incorporating information on commuting ties contributes to better deviance information criteria (DIC) scores in Arlington.
Abstract: Our goal is to understand the social dynamics affecting domestic and sexual violence in urban areas by investigating the role of connections between area nodes, or communities. We use innovative methods adapted from spatial statistics to investigate the importance of social proximity measured based on connectedness pathways between area nodes. In doing so, we seek to extend the standard treatment in the neighborhoods and crime literature of areas like census blocks as independent analytical units or as interdependent primarily due to geographic proximity. In this paper, we develop techniques to incorporate two types of proximity, geographic proximity and commuting proximity in spatial generalized linear mixed models (SGLMM) in order to estimate domestic and sexual violence in Detroit, Michigan and Arlington County, Virginia. Analyses are based on three types of CAR models (the Besag, York, and Mollie (BYM), Leroux, and the sparse SGLMM models) and two types of SAR models (the spatial lag and spatial error models) to examine how results vary with different model assumptions. We use data from local and federal sources such as the Police Data Initiative and American Community Survey. Analyses show that incorporating information on commuting ties, a non-spatially bounded form of social proximity, to spatial models contributes to better deviance information criteria scores (a metric which explicitly accounts for model fit and complexity) in Arlington for sexual and domestic crime as well as overall crime. In Detroit, the fit is improved only for overall crime. The distinctions in model fit are less pronounced when using cross-validated mean absolute error as a comparison criteria. Overall, the results indicate variations across crime type, urban contexts, and modeling approaches. Nonetheless, in important contexts, commuting ties among neighborhoods are observed to greatly improve our understanding of urban crime. If such ties contribute to the transfer of norms, social support, resources, and behaviors between places, they may then transfer also the effects of crime prevention efforts.
TL;DR: In this paper, DBSCAN* was used to estimate the home clusters of 54,249 Twitter users who sent at least one geotagged tweet in Boston and found that the proportion of Twitter users on a block at a given time who are local residents, intermetro commuters, or tourists is correlated with incidences of public violence and private conflict for four different time periods: weekday days, weekday nights, weekend days, and weekend nights.
Abstract: Test the reliability of geotagged Twitter data for estimating block-level population metrics across place types. Evaluate whether the proportion of Twitter users on a block at a given time who are local residents, inter-metro commuters, or tourists is correlated with incidences of public violence and private conflict for four different time periods: weekday days, weekday nights, weekend days, and weekend nights. DBSCAN* machine learning technique is used to estimate the home clusters of 54,249 Twitter users who sent at least one geotagged tweet in Boston. Public violence and private conflict are measured using geocoded 911 dispatches. ANOVA models are used to evaluate how the presence of our three groups of interests varies across three types of block-level land usage. Hierarchical linear regression models are used to evaluate whether the proportion of commuters and tourists at census tract- and block-levels are predictive of crime events across the four time periods of interest. We find evidence that Twitter data has limited reliability across residential blocks due to data sparseness. For non-residential blocks, we find that commuter and tourist presence at the block-level are positively associated with both public violence and private conflict, but that these effects are not stable across time periods. Commuters and tourists only effect violence during weekday days, and the effects of commuters and tourists on private conflict are only statistically significant during weekday days and weekend days. Consistent with routine activities and crime pattern theories, the influx of outsiders in a given location impacts the likelihood of crime occurring there. While we find that data from Twitter users can be valuable for measuring block-level ambient populations, it appears this is not true for residential blocks. Future research may further consider how the characteristics of Twitter users may inform spatial patterns in crime.
TL;DR: In this paper, the authors assess the sensitivity of racial disparities in UOF severity to a series of analytic choices, using a 5'×'2'2 '2' '3' '4' '5' '6' '7' '8' '9' '10' '11' '12' '13' '14' '15' '16' '17' '18' '20' '21' '24' '25' '26' '27' '28' '29' '30' '31' '34
Abstract: To understand the impact of measurement and analytic choices on assessments of police use of force (UOF) and racial disparities therein. We collected and standardized UOF data (N = 9982 incidents) from a diverse set of 11 police departments, and measured departments’ aggregate force severity in five ways. We assessed the sensitivity of racial disparities in UOF severity to a series of analytic choices, using a 5 × 2 × 2 × 2 design comparing force severity to population and arrest benchmarks, using two definitions of minority group (Black/Nonwhite), and two modes of comparison (ratios/differences). Significant racial disparities were observed under most analytic choices in most departments. However, lethal force was rare, and estimates of lethal force disparities were statistically uncertain, as were departments’ relative ranks as equitable or disparate. Ratios of minority to White force severity were less sensitive to measurement differences within measures including nonlethal force. The choice of a population or arrest benchmark had implications for which departments emerged as highly disparate, while focal minority group and mode of comparison had less systematic effects. Given increased scrutiny of police activity by advocates and policymakers, it is important to understand how measurement and other analytic choices affect our understanding of equity in police practices. Our findings demonstrate that analytical decisions interact in complex ways and that standardization is essential when comparing multiple departments. We recommend comprehensive data collection that includes nonlethal as well as lethal force, and make recommendations for measuring and contextualizing racial disparities in UOF and other police activity.
TL;DR: In this article, negative binomial regression models were used to estimate the influence of street block slope on robbery net of betweenness, facility composition, and socio-demographics in Cincinnati, Ohio.
Abstract: To examine the influence of street block slope on robbery in Cincinnati, Ohio. Data visualizations were used to examine how street block slope varies across the city. Negative binomial regression models were used to estimate the influence of street block slope on robbery net of betweenness, facility composition, and socio-demographics. A 1% increase in street block slope was associated with roughly 4.5% fewer street block robberies per foot of street block length. Street blocks with a higher expected usage potential, measured via betweenness, were also observed to have higher expected robbery levels. In addition, numerous facilities and neighborhood socio-demographic characteristics linked to higher robbery levels. Steeper street blocks may have fewer robberies because they make the physical costs for committing robberies too high, are too difficult to escape from, and/or provide fewer robbery opportunities due to relatively lower usage. Moreover, more robberies appear to occur on street blocks with higher betweenness due to more potential opportunities there. Finally, the influence of facilities and community characteristics were largely consistent with theoretical expectations and past studies. Future studies should continue to examine how topography influences aggregate crime levels and offender decision making in other settings to bolster the external validity of the present findings.
TL;DR: The method develops a method for determining a granularity that provides a compromise between internal uniformity and robustness to error and shows that finer is not necessarily better in the micro-analysis of crime, and that units coarser than street segments might be better for this type of study.
Abstract: Crime counts are sensitive to granularity choice. There is an increasing interest in analyzing crime at very fine granularities, such as street segments, with one of the reasons being that coarse granularities mask hot spots of crime. However, if granularities are too fine, counts may become unstable and unrepresentative. In this paper, we develop a method for determining a granularity that provides a compromise between these two criteria. Our method starts by estimating internal uniformity and robustness to error for different granularities, then deciding on the granularity offering the best balance between the two. Internal uniformity is measured as the proportion of areal units that pass a test of complete spatial randomness for their internal crime distribution. Robustness to error is measured based on the average of the estimated coefficient of variation for each crime count. Our method was tested for burglaries, robberies and homicides in the city of Belo Horizonte, Brazil. Estimated “optimal” granularities were coarser than street segments but finer than neighborhoods. The proportion of units concentrating 50% of all crime was between 11% and 23%. By balancing internal uniformity and robustness to error, our method is capable of producing more reliable crime maps. Our methodology shows that finer is not necessarily better in the micro-analysis of crime, and that units coarser than street segments might be better for this type of study. Finally, the observed crime clustering in our study was less intense than the expected from the law of crime concentration.
TL;DR: In this article, the authors explored the merits of commercially-based survey data on crime through cross-validation with established crime metrics and found that African and Latin American countries suffer from the highest levels of various types of crime across the board, followed by countries in Asia.
Abstract: This article explores the merits of commercially-based survey data on crime through cross-validation with established crime metrics. Using unpublished data from 166 countries covering the period between 2006 and 2019, the article describes the geographical distribution across global regions and trends over time of three types of common crime, homicide, and organised crime. The article then explores possible determinants of the geographical distributions through regressing prevalence rates against indices of poverty, inequality, proportion of youth, presence of criminal opportunities (wealth and urbanisation), and governance/rule of law. The results show that African and Latin American countries suffer from the highest levels of various types of crime across the board, followed by countries in Asia. European, North American and Australian countries experience intermediate or relatively low levels of most types of crime. Levels of common crime have dropped or stabilized globally except in Africa where they went up. Homicides have fallen almost universally. Trends in organised crime are diverging. Dimensions of governance emerged as powerful determinants of levels of all types of crime. Important determinants of common crime besides governance were poverty, inequality, and proportion of youth. To some extent changes in these same characteristics of countries were found to be correlated with changes in levels of crime over the past fifteen years. The article concludes with a discussion of the study’s limitations and suggestions for further research.
TL;DR: In this article, the authors conducted an event-study analysis with a robust set of controls to determine the impact of racial uprisings on the subsequent police killings of African American civilians.
Abstract: There is a long-standing history of protests in response to police killings of African American citizens. However, it remains a largely unanswered question as to whether these protest events have had any impact on subsequent police killings of African American civilians. To answer this question, we turn to the over 700 racial uprisings that occurred in the 1960s and early 1970s that were largely triggered by negative and often violent interactions between the African American community and police. To determine the impact of racial uprisings on police killings of civilians, we conduct an event-study analysis with a robust set of controls. We employ data on civilian deaths by legal intervention by race, county-level uprising occurrence, and county demographic characteristics. We take advantage of variation in the location and timing of a county’s first uprising to determine the impact of uprisings on police killings of civilians. Our identification strategy relies on pre-existing trends in deaths by legal intervention being uncorrelated with the date of the first uprising in a county. The results show that counties saw a marked increase in both non-white and white deaths due to legal intervention in the years immediately following an uprising. This initial increase is substantially larger for non-whites relative to white civilians. Deaths due to legal intervention for non-white and white civilians diverge over the medium-to-long run. Non-white deaths resulting from legal intervention remain elevated after nearly a decade while deaths of whites revert to their pre-existing trend after a handful of years. Additional analysis regarding the impact of uprisings on policing shows that total crime and police employment do not change in a significant manner over the long run, however, officers are more likely to be killed or injured on duty. The results clearly show that historical protest resulted in an increase in civilian deaths by legal intervention regardless of race in the short-run and a seemingly permanent increase in killings of non-white over the medium-to-long run. These results paint a depressing picture in which uprisings represent a structural change in police-civilian relations, adversely affecting white civilians in the short-run and non-white civilians in the short and long-run.
TL;DR: In this paper, the authors examined the correlation between attrition bias and recidivism in the LoneStar Project, a longitudinal sample of reentering men in Texas, using four sample conditions: full sample, listwise deleted sample, multiply imputed sample, and two-stage corrected sample.
Abstract: Longitudinal data offer many advantages to criminological research yet suffer from attrition, namely in the form of sample selection bias. Attrition may undermine reaching valid inferences by introducing systematic differences between the retained and attrited samples. We explored (1) if attrition biases correlates of recidivism, (2) the magnitude of bias, and (3) how well methods of correction account for such bias. Using data from the LoneStar Project, a representative longitudinal sample of reentering men in Texas, we examined correlates of recidivism using official measures of recidivism under four sample conditions: full sample, listwise deleted sample, multiply imputed sample, and two-stage corrected sample. We compare and contrast the results regressing rearrest on a range of covariates derived from a pre-release baseline interview across the four sample conditions. Attrition bias was present in 44% of variables and null hypothesis significance tests differed for the correlates of recidivism in the full and retained samples. The bias was substantial, altering effect sizes for recidivism by a factor as large as 1.6. Neither the Heckman correction nor multiple imputation adequately corrected for bias. Instead, results from listwise deletion most closely mirrored the results of the full sample with 89% concordance. It is vital that researchers examine attrition-based selection bias and recognize the implications it has on their data when generating evidence of theoretical, policy, or practical significance. We outline best practices for examining the magnitude of attrition and analyzing longitudinal data affected by sample selection.
TL;DR: In this article, causal mediation analysis is used as a versatile technique of effect decomposition, which can derive the indirect effect even in the presence of a treatment-mediator interaction.
Abstract: Review causal mediation analysis as a method for estimating and assessing direct and indirect effects. Re-examine a field experiment with an apparent implementation failure. Test procedural justice theory by examining to which extent procedural justice mediates the impact of contact with the police on police legitimacy and social identity. Data from a block-randomised controlled trial of procedural justice policing (the Scottish Community Engagement Trial) were analysed. All constructs were measured using surveys distributed during roadside police checks. Treatment implementation was assessed by analysing the treatment effect’s consistency and heterogeneity. Causal mediation analysis, which can derive the indirect effect even in the presence of a treatment–mediator interaction, was used as a versatile technique of effect decomposition. Sensitivity analysis was carried out to assess the robustness of the mediating role of procedural justice. First, the treatment effect was fairly consistent and homogeneous, indicating that the treatment’s effect is attributable to the design. Second, there is evidence that procedural justice channels the treatment’s effect towards normative alignment (NIE = − 0.207), duty to obey (NIE = − 0.153), and social identity (NIE = − 0.052), all of which are moderately robust to unmeasured confounding (ρ = 0.3–0.6, LOVE = 0.5–0.7). The effect’s consistency and homogeneity should be examined in future block-randomised designs. Causal mediation analysis is a versatile tool that can salvage experiments with systematic yet ambiguous treatment effects by allowing researchers to “pry open” the black box of causality. The theoretical propositions of procedural justice policing were supported. Future studies are needed with more discernible causal mediation effects.
TL;DR: This article used a multilevel negative binomial-logit hurdle model to examine whether the factors that explain the likelihood of becoming a victim of extortion also explain the number of incidents suffered by victimized businesses.
Abstract: Research consistently shows that crime concentrates on a few repeatedly victimized places and targets. In this paper we examine whether the same is true for extortion against businesses. We then test whether the factors that explain the likelihood of becoming a victim of extortion also explain the number of incidents suffered by victimized businesses. The alternative is that extortion concentration is a function of event dependence. Drawing on Mexico’s commercial victimization survey, we determine whether repeat victimization occurs by chance by comparing the observed distribution to that expected under a Poisson process. Next, we utilize a multilevel negative binomial-logit hurdle model to examine whether area- and business-level predictors of victimization are also associated with the number of repeat extortions suffered by businesses. Findings suggest that extortion is highly concentrated, and that the predictors of repeated extortion differ from those that predict the likelihood of becoming a victim of extortion. While area-level variables showed a modest association with the likelihood of extortion victimization, they were not significant predictors of repeat incidents. Similarly, most business-level variables significantly associated with victimization risk showed insignificant (and sometimes contrary) associations with victimization concentration. Overall, unexplained differences in extortion concentration at the business-level were unaffected by predictors of extortion prevalence. The inconsistent associations of predictors across the hurdle components suggest that extortion prevalence and concentration are fueled by two distinct processes, an interpretation congruent with theoretical expectations regarding extortion that considers that repeats are likely fueled by a process of event dependence.
TL;DR: In this paper, the effects of body-worn cameras (BWCs) on rates of fatalities arising from police-citizen encounters were evaluated using difference-in-difference (DID) analyses using Poisson models.
Abstract: This study assesses the effects of body-worn cameras (BWCs) on rates of fatalities arising from police-citizen encounters. While existing experimental research has not examined this outcome because it is so rare, the staggered roll-out of BWCs across the nation’s law enforcement agencies provides an opportunity for quasi-experimental analysis. Difference-in-difference (DID) analyses using Poisson models compare changes in U.S. law enforcement agencies’ fatality counts with changes in BWC acquisition. Using a federal law enforcement survey focused on body worn cameras (LEMAS-BWCS) and media-sourced data on fatal encounters from fatalencounters.org (FE), the research examines agencies acquiring BWCs between 2013/14 and 2015/16 and those that did not acquire them up to 2016 and had no plans to do so. It includes a fixed effects annual panel data analysis with data from 2005/06 to 2018/19 and two two-group analyses focusing on a pre-treatment period (2010/11 to 2012/13) and a post-treatment period (2016/17 to 2018/19). The latter includes a propensity score matched comparison. Two out of three DID analyses showed statistically significant negative effects of BWCs on citizen fatalities. The propensity score matched two-group analysis returned a non-significant negative effect. The research finds some evidence for BWC effects on citizen fatalities. However, there are important validity threats to this conclusion. These include the possibility that BWC acquisition serves as a marker for other policy changes focused on BWC-acquiring agencies in the 2013/14 to 2015/16 period and beyond.
TL;DR: This article conducted an anonymous survey to authors of articles published in criminology journals to examine the prevalence of Questionable Research Practices (QRPs) and Open Science Practices (OSPs).
Abstract: Questionable research practices (QRPs) lead to incorrect research results and contribute to irreproducibility in science. Researchers and institutions have proposed open science practices (OSPs) to improve the detectability of QRPs and the credibility of science. We examine the prevalence of QRPs and OSPs in criminology, and researchers’ opinions of those practices. We administered an anonymous survey to authors of articles published in criminology journals. Respondents self-reported their own use of 10 QRPs and 5 OSPs. They also estimated the prevalence of use by others, and reported their attitudes toward the practices. QRPs and OSPs are both common in quantitative criminology, about as common as they are in other fields. Criminologists who responded to our survey support using QRPs in some circumstances, but are even more supportive of using OSPs. We did not detect a significant relationship between methodological training and either QRP or OSP use. Support for QRPs is negatively and significantly associated with support for OSPs. Perceived prevalence estimates for some practices resembled a uniform distribution, suggesting criminologists have little knowledge of the proportion of researchers that engage in certain questionable practices. Most quantitative criminologists in our sample have used QRPs, and many have used multiple QRPs. Moreover, there was substantial support for QRPs, raising questions about the validity and reproducibility of published criminological research. We found promising levels of OSP use, albeit at levels lagging what researchers endorse. The findings thus suggest that additional reforms are needed to decrease QRP use and increase the use of OSPs.
TL;DR: The finding that recency and proximity of prior burglaries do not contribute to the performance of the forecast, probably indicates that relevant spatio-temporal interaction is limited to fine-grained spatial and temporal units of analysis, such as days and street blocks.
Abstract: Objectives - We investigate the spatio-temporal variation of monthly residential burglary frequencies across neighborhoods as a function of crime generators, street network features and temporally and spatially lagged burglary frequencies. In addition, we evaluate the per-formance of the model as a forecasting tool.
Methods - We analyze 48 months of police-recorded residential burglaries across 20 neigh-borhoods in Amsterdam, the Netherlands, in combination with data on the locations of urban facilities (crime generators), frequencies of other crime types, and street network data. We apply the Integrated Laplace Approximation method, a Bayesian forecasting framework that is less computationally demanding than prior frameworks.
Results - The local number of retail stores, the number of street robberies perpetrated and the closeness of the local street network are positively related to residential burglary. Inclu-sion of a general spatio-temporal interaction component significantly improves forecasting performance, but inclusion of spatial proximity or temporal recency components does not.DiscussionOur findings on crime generators and street network characteristics support evi-dence in the literature on environmental correlates of burglary. The significance of spatio-temporal interaction indicates that residential burglary is spatio-temporally concentrated. Our finding that recency and proximity of prior burglaries do not contribute to the perfor-mance of the forecast, probably indicates that relevant spatio-temporal interaction is lim-ited to fine-grained spatial and temporal units of analysis, such as days and street blocks.
Discussion - Our findings on crime generators and street network characteristics support evidence in the literature on environmental correlates of burglary. The significance of spatio-temporal interaction indicates that residential burglary is spatio-temporally concentrated. Our finding that recency and proximity of prior burglaries do not contribute to the performance of the forecast, probably indicates that relevant spatio-temporal interaction is limited to fine-grained spatial and temporal units of analysis, such as days and street blocks.
TL;DR: In this paper, the authors introduced psychopathology network modeling as an analytic strategy capable of addressing these challenges through a more nuanced description of the structural and statistical association between features of psychopathy and offending.
Abstract: Concerns about the value of features of psychopathy to explanations of offending may be driven by challenges with testing this relationship as opposed to the construct’s limited predictive validity. The current study introduced psychopathology network modeling as an analytic strategy capable of addressing these challenges through a more nuanced description of the structural and statistical association between features of psychopathy and offending. To provide more direct implications for criminological theory and research, this included examining whether within-individual changes in features of psychopathy were associated with within-individual change in offending. Data on male and female young offenders from the Pathways to Desistance Study (n = 1354) were used to examine the association between features of psychopathy and offending versatility, as measured by the Youth Psychopathic Traits Inventory and the Self Report of Offending scale, respectively. Three network structures were modeled that separated the variance in the relationship between features of psychopathy and offending into between-subjects and within-individual networks. Between-subjects and within-individual analyses indicated that interpersonal and affective features of psychopathy were positively associated with offending versatility. The network approach indicated that remorselessness and manipulativeness were central features of psychopathy that were also associated with offending versatility. Remorselessness in particular helped bridge together conceptually distinct features of psychopathy. Psychopathology network modeling illustrated the value of features of psychopathy to mainstream criminological theory and research. These features included interpersonal and affective deficits, which previously were identified as poor predictors of offending outcomes.
TL;DR: In this article, a systematic review and meta-analysis have been carried out to fill the gap of systematic knowledge about the effectiveness of sports programs in crime and delinquency, finding that participants showed a significant decrease in outcomes such as aggressiveness or anti-social behavior when they participated in sports programs.
Abstract: Sports programs are widely implemented as measures of crime prevention. In contrast to their popularity, there is little systematic knowledge about their effectiveness. This systematic review and meta-analysis have been carried out to fill this gap. In a systematic review, we gathered data on evaluated prevention programs specifically designed to prevent crime and delinquency. We then conducted a meta-analytic integration with studies using at least roughly equivalent control groups for the program evaluation. To retrieve relevant literature, we conducted a comprehensive international literature search until June 2021 drawing on scientific databases. We also applied snow-balling searches and contacted practitioners in the field. Studies were eligible if they evaluated sports programs designed to prevent delinquency on primary, secondary, and/or tertiary level. We focused on crime-related outcomes and potentially underlying psycho-social factors. We made no restrictions regarding characteristics of the participants or other aspects such as duration of the program. 24 studies were eligible for our systematic review, from which only thirteen were included into our meta-analytic integration. We found a moderate effect of participation in sports programs on crime-related outcomes (d = 0.36, p < .001). Participants showed a significant decrease in outcomes such as aggressiveness or anti-social behavior. We also analyzed psychological outcomes such as self-esteem or mental well-being, which also significantly improved when participating in sports programs (d = 0.87, p < ..05). Sports programs seem to be an effective measure of crime prevention. However, future research needs more sound evaluation designs and moderator analyses to better understand the functioning and improve the implementation of sports programs.
TL;DR: In this paper, Luo et al. used the age-period-cohort interaction (APC-I) model developed by Luo and Hodges (Sociol Methods Res 2020) to analyze the UCR age-specific arrest statistics for robbery, aggravated assaults, and homicide from 1960 to 2014.
Abstract: Previous research in criminology has overlooked that cohort effects on crime should be age-time-specific (Ryder in Am Sociol Rev 30(6):843–861, 1965) and consequently assumed cohort effects to be the same across the life course. The current study addresses these limitations by modeling cohort effects as the differential impacts of social change depending on age groups. With this new operationalization that is closely tied to Ryder’s conceptualization, we examine both inter-cohort differences and intra-cohort dynamics in violent crime. We use the age–period–cohort-interaction (APC-I) model developed by Luo and Hodges (Sociol Methods Res 2020) to analyze the UCR age-specific arrest statistics for robbery, aggravated assaults, and homicide from 1960 to 2014. Specifically, we estimate and test two types of cohort variation: average cohort deviations and life-course dynamics. Our findings reveal varying degrees of cohort deviations at different ages. The early boomers (born between 1945 and 1954) and the late boomers (born between 1955 and 1964) demonstrate different intra-cohort dynamics of robbery arrest, and the violence epidemic cohorts’ (born between 1975 and 1984) high risks of homicide arrest appear to be driven by cohort deviations at young ages. The APC-I framework introduced in this study provides new insights into the dynamic aspect of cohort effects on violent crime that have not been examined in the criminological literature. Criminological studies on cohort effects would benefit by shifting away from the problematic assumption of constant and additive cohort effects to the dynamic and interactive approach represented by the APC-I framework.