Search

In this study, we use 1980–2019 longitudinal age-arrest data from Taiwan and applied the age-period-cohort-interaction (APC-I) model (Luo & Hodges, 2022) to examine the stability or change in the age-arrest distributions across five offenses. We focus on two research questions: (1) whether the shape of age-arrest curves in Taiwan diverges from the Hirschi and Gottfredson's (HG) invariant premise after accounting for period and cohort effects; and (2) whether any observed period or cohort effects on age patterns vary depending on offense type. Findings indicate overall consistency in the shape of Taiwan's age-arrest distributions after adjusting for period and cohort effects, which are characterized by relatively older peak ages and symmetrical spread-out distributions that diverge considerably from HG's invariant projection and prototypical US age-arrest patterns. In addition, we find that period effects have contributed to higher arrest rates in recent years, and cohort effects have impacted somewhat the shape of Taiwan's age-arrest distributions. These findings, along with recent cross-sectional evidence from Taiwan, South Korea, and India (Steffensmeier et al., 2017; 2019; 2020), further confirm that the aggregate age-crime relationship is robustly influenced by country-specific processes and historical and social transformations. [ABSTRACT FROM AUTHOR]

Introduction: Accurate prediction of short-term offending in young people exhibiting antisocial behaviour could support targeted interventions. Here we develop a set of machine learning (ML) models that predict offending status with good accuracy; furthermore, we show interpretable ML analyses can complement models to inform clinical decision-making.This study included 679 individuals aged 11–17 years who displayed moderate-to-severe antisocial behaviour, from a controlled trial of Multisystemic-therapy in England. The outcome was any criminal offence in the 18 months after study baseline. Four types of ML algorithms were trained: logistic regression, elastic net regression, random forest, and gradient boosting machine (GBM). Prediction models were developed (1) using predictors readily available to clinicians (e.g. sociodemographics, previous convictions), and (2) with additional information (e.g. parenting). Model agnostic feature importance values were calculated and the most important predictors identified. Nested cross-validation with 100 iterations of random data splits and 10-fold cross-validation within each iteration was employed, and the average predictive performance was reported.Among the ML models using readily available predictors, the GBM is the strongest model (AUC 0.85, 95% CI 0.85–0.86); the other models have average AUCs of 0.82. This performance was better than using only the total number of previous offences as the predictor (0.67, 0.66–0.68), and the model simply assuming past offending status as the prediction (0.81, 0.80–0.81). Additional predictors slightly increased the performance of logistic regression and random forest models but decreased the performance of elastic net regression and gradient boosting machine-based models.The potential utility of ML approaches for accurately predicting criminal offences in high-risk youth is demonstrated. Interpretable ML-based predictive models could be utilised in youth services or research to help develop and deliver effective interventions.Methods: Accurate prediction of short-term offending in young people exhibiting antisocial behaviour could support targeted interventions. Here we develop a set of machine learning (ML) models that predict offending status with good accuracy; furthermore, we show interpretable ML analyses can complement models to inform clinical decision-making.This study included 679 individuals aged 11–17 years who displayed moderate-to-severe antisocial behaviour, from a controlled trial of Multisystemic-therapy in England. The outcome was any criminal offence in the 18 months after study baseline. Four types of ML algorithms were trained: logistic regression, elastic net regression, random forest, and gradient boosting machine (GBM). Prediction models were developed (1) using predictors readily available to clinicians (e.g. sociodemographics, previous convictions), and (2) with additional information (e.g. parenting). Model agnostic feature importance values were calculated and the most important predictors identified. Nested cross-validation with 100 iterations of random data splits and 10-fold cross-validation within each iteration was employed, and the average predictive performance was reported.Among the ML models using readily available predictors, the GBM is the strongest model (AUC 0.85, 95% CI 0.85–0.86); the other models have average AUCs of 0.82. This performance was better than using only the total number of previous offences as the predictor (0.67, 0.66–0.68), and the model simply assuming past offending status as the prediction (0.81, 0.80–0.81). Additional predictors slightly increased the performance of logistic regression and random forest models but decreased the performance of elastic net regression and gradient boosting machine-based models.The potential utility of ML approaches for accurately predicting criminal offences in high-risk youth is demonstrated. Interpretable ML-based predictive models could be utilised in youth services or research to help develop and deliver effective interventions.Results: Accurate prediction of short-term offending in young people exhibiting antisocial behaviour could support targeted interventions. Here we develop a set of machine learning (ML) models that predict offending status with good accuracy; furthermore, we show interpretable ML analyses can complement models to inform clinical decision-making.This study included 679 individuals aged 11–17 years who displayed moderate-to-severe antisocial behaviour, from a controlled trial of Multisystemic-therapy in England. The outcome was any criminal offence in the 18 months after study baseline. Four types of ML algorithms were trained: logistic regression, elastic net regression, random forest, and gradient boosting machine (GBM). Prediction models were developed (1) using predictors readily available to clinicians (e.g. sociodemographics, previous convictions), and (2) with additional information (e.g. parenting). Model agnostic feature importance values were calculated and the most important predictors identified. Nested cross-validation with 100 iterations of random data splits and 10-fold cross-validation within each iteration was employed, and the average predictive performance was reported.Among the ML models using readily available predictors, the GBM is the strongest model (AUC 0.85, 95% CI 0.85–0.86); the other models have average AUCs of 0.82. This performance was better than using only the total number of previous offences as the predictor (0.67, 0.66–0.68), and the model simply assuming past offending status as the prediction (0.81, 0.80–0.81). Additional predictors slightly increased the performance of logistic regression and random forest models but decreased the performance of elastic net regression and gradient boosting machine-based models.The potential utility of ML approaches for accurately predicting criminal offences in high-risk youth is demonstrated. Interpretable ML-based predictive models could be utilised in youth services or research to help develop and deliver effective interventions.Conclusion: Accurate prediction of short-term offending in young people exhibiting antisocial behaviour could support targeted interventions. Here we develop a set of machine learning (ML) models that predict offending status with good accuracy; furthermore, we show interpretable ML analyses can complement models to inform clinical decision-making.This study included 679 individuals aged 11–17 years who displayed moderate-to-severe antisocial behaviour, from a controlled trial of Multisystemic-therapy in England. The outcome was any criminal offence in the 18 months after study baseline. Four types of ML algorithms were trained: logistic regression, elastic net regression, random forest, and gradient boosting machine (GBM). Prediction models were developed (1) using predictors readily available to clinicians (e.g. sociodemographics, previous convictions), and (2) with additional information (e.g. parenting). Model agnostic feature importance values were calculated and the most important predictors identified. Nested cross-validation with 100 iterations of random data splits and 10-fold cross-validation within each iteration was employed, and the average predictive performance was reported.Among the ML models using readily available predictors, the GBM is the strongest model (AUC 0.85, 95% CI 0.85–0.86); the other models have average AUCs of 0.82. This performance was better than using only the total number of previous offences as the predictor (0.67, 0.66–0.68), and the model simply assuming past offending status as the prediction (0.81, 0.80–0.81). Additional predictors slightly increased the performance of logistic regression and random forest models but decreased the performance of elastic net regression and gradient boosting machine-based models.The potential utility of ML approaches for accurately predicting criminal offences in high-risk youth is demonstrated. Interpretable ML-based predictive models could be utilised in youth services or research to help develop and deliver effective interventions. [ABSTRACT FROM AUTHOR]

This paper examines the impact of local banking market frictions measured by bank failures on youth crime. Using a difference-in-differences framework, we find white and black youth violent crimes increase by 0.11 and 0.66 per 1000 people, respectively, following a bank failure. Further, we find white property crime, petty delinquency and drug crime increase by 0.29, 0.75 and 0.35 per 1000 people. Black youth see a rise in drug crimes by 0.88 per 1000 people. Results by gender reveal male drug crimes rise by 2.4 and female property crimes by 0.66 per 1000 people, respectively. The economic effects are most persistent for petty delinquency crimes committed by white youths. Furthermore, we find evidence that bank failures worsen local labor markets conditions by increasing white youth unemployment and reduce employment in the limited-services restaurant industry; channels that likely alter the opportunity cost of youth crime. [ABSTRACT FROM AUTHOR]

Copyright of Forensische Psychiatrie, Psychologie, Kriminologie is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Recent years have seen a wide discussion of populism in penal policy, which is internationally regarded as a strong drive for establishing punitive tendencies. Generally, 'penal populism' is characterized by an extensive consensus across the most influential political parties, a punitive orientation, and the dismissal of scientific or professional expertise. Recent penal policy therefore appears to be a relatively unified practice strongly oriented toward punitive measures that primarily address the public and its perceived need for protection. Because analyses of Anglophone countries are predominant in this discussion, we contrast them with a reconstruction of debates on youth crime in German parliaments from 1970 to 2012. They exhibit a wide variety of populist articulations. Although they imply a strong punitive bias, they also encompass a very heterogeneous rhetoric of penal policy. In conclusion, we argue that penal populism can (and should) be described as a tactical practice, i.e., as political maneuvering employed to negotiate the prospects of punitive and other styles of politics. [ABSTRACT FROM AUTHOR]

Why have social constructionists remained absent from debates over public sociology? I argue that constructionist scholarship would be particularly amenable to Michael Burawoy's notion of 'organic' public sociology, given the ability of constructionist scholars to orient awareness contexts in order to help engender constructionist imaginations. This approach requires that constructionists take on a different view of the role of the analyst. I also discuss some of the problems Canadian academics have had engaging with the media in their efforts to engage in 'traditional' public sociology, as well as what a constructionist public sociology may look like practice. I conclude by addressing potential challenges to a constructionist public sociology within Canada, including reference to sociology's disciplinary coherence and how we can approach-and what we mean by-'publics'. [ABSTRACT FROM AUTHOR]

Objective: The aim of this study was to examine the psychosocial correlates of various crime types among adolescent males born in Finland in 1981. Methods: Data on crime registered in the Finnish National Police Register between 1998 and 2001 were received for 2,866 boys, of whom 81% ( n = 2,330) filled in a questionnaire at obligatory military call-up at age 18 in 1999. Crime was divided into five types: drug, violent, property, traffic, and drunk driving offences. Results: Of the 2,866 boys, 23% had been registered for offending; 4% for drug, 7% for violent, 11% for property, 11% for traffic, and 5% for drunk driving offences during the 4-year period in late adolescence. All the crime types correlated with each other and shared many of the psychosocial problems. Small community size, parents' divorce, aggressiveness, daily smoking, and weekly drunkenness were generic correlates of crime, being independently related to various crime types. Conclusions: The results support general rather than specific accounts of youth crime. In particular, measures moderating the adverse effects of divorce, alleviating parental adversities and supporting parenthood, and tackling substance abuse seem relevant in social and criminal policy because they address psychosocial problems characterizing youth crime in general. [ABSTRACT FROM AUTHOR]

This study examines the null hypothesis of Granger no-causality between labor force participation (LFP) and juvenile delinquency in Taiwan. In order to explore this issue more thoroughly, this study adopts the approach proposed by Toda and Yamamoto (1995, Journal of Econometrics, 66(1–2), 225–250). It uses official time-series data provided by the Government of Taiwan. After estimating both a four- and five-variable VAR system, one that substitutes both male and female LFP rates for the aggregate LFP rate, the primary findings of this study reveal the following: The higher the past juvenile crime rate, the lower the future aggregate and female LFP rate will be. In addition, the higher the past male LFP rate, the higher the future juvenile crime rate will be. These findings are quite robust in terms of different lag-length structures. [ABSTRACT FROM AUTHOR]

Despite considerable research on juvenile homicide, pre-adolescent homicide offenders have received less attention. This paper reviews the existing literature on preteen murderers in order to characterize the current state of research knowledge about this population, and draws on some of the work on adolescent homicide as well. The analysis of this literature considers historical context, methodological issues, previous attempts to classify youthful homicide offenders, and predictors of preteen homicidal behavior. While there is a high degree of heterogeneity within this population, several developmental similarities emerged across cases that were associated with the perpetration of homicide by preteens. A high percentage of preteen homicide offenders come from homes characterized by physical abuse, domestic violence, poor or absent parenting, and overall instability. Gun availability may have been a facilitating factor. Support for different etiologies of preteen versus adolescent homicide is weak. Recommendations for future research directions are offered. [ABSTRACT FROM AUTHOR]