Abstract Loss and rejection are central to prison life, yet how they crystallize into a more general sense of abandonment remains under‐theorized. Drawing on 64 interviews with men held in prolonged self‐isolation or serving indeterminate forvaring sentences in Denmark, we conceptualize abandonment not simply as hardship but as suffering plus attribution: being left in harm by identifiable others perceived to have the capacity to respond. Building on Stauffer's notion of ethical loneliness, defined as “the experience of being abandoned by humanity compounded by the experience of not being heard” (2015, p. 9), we develop layered abandonment as a typological framework. We specify five distinct layers of abandonment that may occur independently or in combination and show how their co‐occurrence compounds suffering. With abandonment, we foreground the moral and relational processes through which prisoners interpret suffering as the failure of others to respond, rather than treating suffering as an impersonal, arbitrary, or diffuse condition of confinement.
Glucagon-like peptide-1 receptor agonists (GLP-1 RAs), a class of medications widely prescribed for diabetes and obesity, have exhibited emerging effects on substance use, reward processing, and impulse control. This study examines whether current GLP-1 RA use moderates established behavioral pathways to violent crime from impulsivity and alcohol use. Data come from a 2025 nationally representative US survey (n = 7521; 821 lifetime GLP-1 RA users). Negative binomial regression models with overlap weighting compared current (n = 597) and former (n = 224) users on self-reported violent criminal behavior in the past year. Although impulsivity and alcohol use were strongly associated with violent criminality, these associations were significantly weaker among current GLP-1 RA users (GLP-1 RA & times; impulsivity incidence rate ratio (IRR) = 0.38, p = 0.002; GLP-1 RA & times; alcohol use IRR = 0.48, p = 0.023). Robustness checks consistently supported the GLP-1 RA & times; impulsivity interaction, whereas less consistent support was found for the alcohol interaction. Results suggest that GLP-1 RAs may attenuate widely documented behavioral risk mechanisms like impulsivity linked to aggression, pointing to novel biosocial hypotheses regarding pharmacological influences on violent criminality. Future longitudinal and experimental research should confirm these findings and test causal mechanisms.
Rational choice models of the decision to commit a crime have been prominent in the criminological literature for decades. Their theoretical components, however, remain fragmented, leading to misconceptions about the framework's explanatory power and policy relevance. The present paper aims to synthesize these fragmented theoretical elements and demonstrate the flexibility of rational choice models in accommodating various recent advancements in research on crime decision-making. To do so, we construct a series of equations that formalize the core mechanisms underlying the decision to offend. We begin with a baseline equation derived from Beccaria's classical deterrence model-centered on the certainty, severity, and celerity of legal sanctions-which is extended to include noncriminal alternatives to crime. We then build on this structure to account for the situational bases for risk perception, discrete stages of the criminal justice process, nonlinear risk processing, ambiguity, moral disutility, safety concerns, and visceral influences such as emotion. The result is a series of compact yet flexible models that address how incentives and constraints shape criminal behavior and that can be used to specify statistical models to test key model predictions.
Offenders target locations that maximize gains and minimize risks, and these are often tied to the built environment and neighborhood social structural characteristics. But gendered routine activities and differential assessments of risk also likely influence where men and women offend and how far they will travel to get there. We investigate these possibilities using data from over 3000 theft arrests in St. Petersburg, FL, and discrete choice modeling to predict where and how far female relative to male suspects travel. We examine whether businesses that might differentially attract women versus men, such as retail, grocery, and convenience stores, shape offending locations, and whether these relationships vary across neighborhood disorder, disadvantage, and distance. We find that neighborhood context outweighs business density in shaping gendered travel patterns. Female suspects disproportionately target neighborhoods with more disorder and travel further distances than male suspects. While disadvantage is predictive of theft location, there is minimal evidence that it varies across gender. While retail and grocery store density play a role, particularly in affluent areas, distance and neighborhood characteristics emerge as the primary factors in where women offend. These findings suggest that the built environment and neighborhood characteristics drive crime travel patterns in gendered ways.
Criminologists have long described and theorized the relationship between status, respect, and violence within urban communities. Although this finding is generally accepted within criminology, ethnographic empirical illustrations of this phenomenon are sparse. To better understand how the potential for violence shapes interactions, affords interactional privilege, and structures relative positions in the social hierarchy, we develop a theoretical framework based on the work of Pierre Bourdieu, Erving Goffman, and Randall Collins. Drawing on ethnographic fieldwork in a disadvantaged area in Copenhagen, we show that interactional boundaries-what liberties one can take with others-are contingent on familiarity and verbal skill but more fundamentally on one's violence capital (i.e., the resources, capacities, skills, and dispositions that confer advantage in exerting corporal violence against others). By empirically illustrating how interactions are structured by violence capital, we contribute to a deeper understanding of how street hierarchies are maintained and reproduced through violence.
Prior research has focused on marriage as a key relationship associated with crime cessation. Yet particularly within the contemporary context, relationships with parents, peers, and other family members may also foster or inhibit progress toward desistance. We rely on a symbolic interactionist perspective on social learning that not only centers on the behaviors of these significant others but foregrounds the role of communication processes within the context of these close ties. Drawing on data from a 22-year longitudinal study, results of mixed effects models indicate that partner and peer criminal behavior are linked to within-individual variations in crime, net of traditional marriage indicators. Additionally, parents' early antisocial behavior was a modest between-subjects predictor, and a composite measure capturing behaviors across the full network suggests a negative effect of "network encapsulation." In-depth interviews (n = 99) with a subset of respondents, parents, and partners identify four specific dynamics related to communication across these relationships that may influence perspectives and behavioral outcomes: (1) co-constructing meanings, (2) buttressing specific hooks for change, (3) presenting possibilities for a united front, and (4) advancing cognitive contrasts. Results highlight the importance of attending to relationship assets and challenges as individuals attempt to accomplish a consistent pattern of desistance.
Transitional justice trials have become a central mechanism for addressing mass violence and human rights violations, yet little is known about the determinants of case outcomes within these courts-particularly in domestic contexts. This study examines Rwanda's gacaca courts, a localized transitional justice system that tried people suspected of participating in the 1994 genocide. Drawing on a comprehensive dataset of all gacaca trials, we analyze how individual-level factors (e.g., gender, age, foreign-born status) and community-level characteristics (e.g., cumulative participants, urbanicity) influenced guilty verdicts and sentence lengths. Our findings suggest that, beyond offense category and charge counts (and, in sentencing models, net of a proxy for confession), demographic and contextual factors significantly shaped trial outcomes. Specifically, in the absence of racial differences among defendants, we find that age and gender were significantly tied to guilty verdicts and sentence lengths. Additionally, defendants tried in districts with higher cumulative participants were less likely to be found guilty and received shorter sentences, consistent with desensitization and system-burden mechanisms. These insights contribute to understanding the lived impacts of transitional justice in Rwanda and offer broader implications for the design and evaluation of domestic postconflict legal systems.
Research on how delinquent peer associations affect individuals' life courses is limited. This paper addresses this gap by examining delinquent peer network characteristics and their impact on offending trajectories through social network analysis (SNA) and group-based trajectory modeling (GBTM). SNA helps minimize same-source bias and offers a detailed view of delinquent peer networks, whereas GBTM identifies developmental pathways. Using data from the National Longitudinal Study of Adolescent to Adult Health (Add Health), this study explores how the popularity and betweenness centrality of delinquent peers affect an individual's placement in offending trajectories. Four trajectories were identified: non-offending, adolescent-limited (AL), low-offending-desister, and high-offending-desister. The study finds that delinquent friend popularity increases the likelihood of placement in AL and high-offending-desister trajectories, whereas delinquent peer betweenness centrality is also associated with increased placement in the low-offending-desister and high-offending-desister trajectories. These findings enhance theoretical understanding by showing how the unique positionality of delinquent peers influences offending behavior over time. The results suggest that public policy should focus on understanding adolescent friendships and providing appropriate restorative resources.
This article is based on Katheryn Russell-Brown's 2025 presidential address at the 85th annual meeting of the American Society of Criminology, held in Washington, D.C. The article begins with an overview of the Author's approach to research and the highlights of her scholarly contributions. Following this, the article presents a research question-whether contemporary race and crime scholars and their scholarship have been stigmatized by the creation of a moral panic designed to marginalize and criminalize their work. This moral panic specifically targets Black history and other race-related scholarship. A criminological and sociological lens is used to analyze the breadth of the panic. Using moral panic as a base for the discussion, the next section examines the broad contours of stigma, and stigmatization and how race-related stigmas operate. Following this is a look at how negative stigmas assigned to race scholars and race scholarship harms the discipline of criminology. For instance, studies indicate that fewer scholars will embrace race and crime scholarship if its value is diminished. Migration studies and imagined futures are included in the assessment of how racial stigmas impact race scholarship. Academic disciplines benefit from rigorous engagement regarding their core principles and perspectives. The absence or diminishment of critical research creates a void in the criminology literature. This raises questions about the representativeness, integrity, and comprehensiveness of criminological scholarship. The article concludes with a charge to the discipline to ensure that race and crime scholarship remain a robust part of the criminological canon.
Extensive research has established a link between low self-control and child victimization. However, the specific neighborhood conditions under which low self-control most strongly influences victimization have been little examined, and, more importantly, no previous studies have investigated the complex ways in which neighborhood context interacts with self-control and other individual characteristics to impact victimization. Drawing on data from the Project on Human Development in Chicago Neighborhoods, the current study tests whether the effect of low self-control on victimization is lessened in neighborhoods with high collective efficacy and further examines the multifaceted nature of neighborhood effects that interact with individual characteristics-particularly age and poverty. Employing three-level multivariate Rasch models and cross-level three-way interactions, the findings reveal that low self-control increases victimization risk, and that the effect of low self-control is contingent on the combined effect between neighborhood collective efficacy of children, age, and poverty. The study concludes that child victimization is a product of the intricate interplay between neighborhood context-especially collective efficacy-and individual factors such as self-control, age, and socioeconomic conditions. This research contributes importantly to building a developmental and life-course perspective on victimization, by highlighting the intricate nature of neighborhood effects.
The decision to intervene when witnessing a crime is central to understanding bystander behavior and social control, yet there remain large gaps in knowledge about the factors that shape the decision to intervene and how. Drawing on theories of reactive guardianship, procedural justice, and crime reporting, we propose a broader conceptual framework of reactive guardianship and bystander decision-making that considers both official (e.g., calling the police) and unofficial (e.g., confronting the offender) interventions. Using data from a national sample (N = 1901) and a preregistered, video-based, factorial experiment that manipulates both offender and victim race (Black vs. White), we explore the joint roles of race, policing attitudes, danger perceptions, and key individual characteristics in shaping bystander responses to crime. Findings indicate that policing-related cognitions and emotions, gun ownership, and bystander characteristics (e.g., race, gender) are associated with differences in how people respond when they witness criminal offending. This study advances theories of formal and informal social control and offers policy-relevant insights for improving public safety and trust in law enforcement.
Several criminological theories identify employment as an important "turning point" that decreases the likelihood of engaging in crime; however, studies have found the effect of employment on recidivism to be small. Recent scholarship has suggested a compelling explanation for these findings; it is not just any jobs, but good jobs, that improve outcomes, and policies and programs should prioritize helping individuals with criminal records build skills valued in the middle-skill workforce. The current study examines whether postsecondary credentials earned during incarceration-specifically, sub-baccalaureate certificates and bachelor's degrees-improve the likelihood of securing middle-skill employment after release. The study employs a r & eacute;sum & eacute; audit methodology in which the author applied to 2623 middle-skill job advertisements using r & eacute;sum & eacute;s of fictional applicants with criminal records that varied on both education credentials and race. The results suggest that employers are more likely to extend an interview for a middle-skill job to individuals who earn bachelor's degrees-but not sub-baccalaureate certificates-while incarcerated relative to those who earn a GED, and that the odds of receiving an interview increase substantially more when relevant experience is coupled with a bachelor's degree. Additionally, bachelor's degrees appear to increase the likelihood of securing a job that offers a living wage.
This study explores the profound and often overlooked developmental journeys of individuals sentenced to life without parole as minors. Through in-depth life-history interviews with 30 formerly incarcerated people, we examine how they forged personal growth amid the cumulative trauma of long-term confinement. Our analysis reveals developmental paths aligned with the posttraumatic growth framework, which recognizes how strength and transformation can emerge from profound adversity. Participants shared accounts of how incarceration shaped their identities, values, and sense of purpose, even in the absence of institutional support for meaningful change. Findings challenge readers to reflect on the paradox between the human capacity for change over time and long-term incarceration, particularly as it relates to individuals who entered confinement during a critical period of ongoing psychosocial development.
Research on racial disparities in the criminal legal system generally examines isolated sentencing decisions, rather than the "package" of punishment that defendants experience. Using Minnesota court administrative data from 2004 to 2017, we specify multivariate and instrumental variables models to simultaneously estimate the outcomes of three elements of racialized punishment: incarceration, probation, and monetary sanctions. We instrument incarceration using jail capacity, which accounts for confounding and the simultaneity of incarceration and other punishment forms. Our results show racial patterning in the "mix" of punishment for similarly situated defendants. Before accounting for this mix, Black, Hispanic, and Native American defendants appear to receive less probation and lower monetary sanctions, but longer incarceration than White defendants. After accounting for instrumented incarceration, monetary sanctions and probation are racialized beyond incarceration in complex ways: Black, Hispanic, and Native American defendants receive lower monetary sanctions as compared to White defendants, and probation for Black and Native American defendants is higher after adjustment. The contours of this racialized package depend critically upon whether the state guidelines recommend a prison sentence. These results show that punishment can be modeled as experienced-as a constitutive package of costs, surveillance, and confinement constrained by structural features of state sentencing guidelines.
Darkness has long been associated with changes in criminal activity. However, evidence on its underlying mechanisms and distributional effects remains limited. This research uses data from 11 densely populated US cities and leverages the discontinuous shift in ambient light caused by daylight saving time to address this knowledge gap. First, it evaluates whether changes in crime reflect deterrence or incapacitation by examining variation in the probability of apprehension. Second, it analyzes the role of routine activity by incorporating ridership and foot traffic data. Third, it studies whether ambient light affects sleep, potentially impairing decision-making and increasing impulsive behavior. The results show that an extra hour of daylight decreases sunset robberies by 28.8%, but increases aggravated assaults by 21.4%. Clearance rates remain stable, suggesting that deterrence rather than incapacitation drives the observed effects. Crimes outside sunset hours and indoor offenses were not affected either, reinforcing the role of natural light exposure. Overall, ambient light operates through multiple, simultaneous mechanisms, causing differential impacts across crime types and neighborhoods. For instrumental crimes, deterrence and enhanced guardianship appear complementary, while alterations in routine activity and cognitive functioning may amplify aggression. Importantly, targeted investments in street lighting, particularly in disadvantaged communities with disproportionate exposure to darkness and situational risks, can help mitigate crime.
Focusing on the 1994 genocide in Rwanda, we examine how sibling relationships-one of the most salient familial bonds-influence individual engagement in violence during mass atrocity. Drawing on an adaptation of differential association and social learning theories for contexts of mass atrocity, we analyze a novel dataset linking over 300,000 siblings from Rwanda's post-genocide gacaca court records. We find that the probability an individual perpetrated interpersonal violence (rather than property crime) increased significantly with the proportion of their siblings who did so. Complementing this quantitative analysis with 38 in-depth interviews with individuals convicted of genocidal violence, we identify several specific mechanisms that help explain this aggregate relationship, such as imitation, restraint, and coercion. Our findings extend classical criminological theories on sibling delinquency to extraordinary contexts of societal breakdown, carrying broader implications for research at the intersection of violent crime, collective behavior, and social networks.
Research consistently shows that lesbian, gay, bisexual, and transgender (LGBT) individuals experience significantly higher rates of violent victimization than their non-LGBT counterparts. In this study, we address theoretical, empirical, and methodological gaps in this developing body of work. Motivated by intersectionality theory and building on existing research, we examine heterogeneity in rates of violent victimization among the LGBT population compared with their non-LGBT counterparts. We also assess the extent to which subgroup sociodemographic differences account for disparities in violent victimization and whether the effect of LGBT status on violent victimization differs by sex. We address these questions using several waves of data from the National Crime Victimization Survey (NCVS). Consistent with prior work, we find that LGBT subgroups have a higher rate of violent victimization than their non-LGBT counterparts. Our results also reveal heterogeneity within the LGBT population. Rates of violent victimization are highest for bisexual individuals, followed by transgender individuals, lesbian/gay individuals, and non-LGBT individuals. Although sociodemographic differences explain a nontrivial portion of these disparities (similar to 15% for gay/lesbian individuals to 41% for bisexual individuals), the bulk of disparities are not accounted for by sociodemographic controls. The conditional effects of sex are minimal and nuanced, warranting further research. We contribute methodologically by drawing attention to current challenges with the recommended measures of sex, sexual orientation, and gender identity used in the NCVS and other large population surveys. We suggest revisions to foster more inclusive, comprehensive, and precise measurement.
Studies of early case processing outcomes in the United States typically assume that decisions are made unilaterally by the prosecutor, such that prior contact with the legal system is universally associated with harsher outcomes for defendants. However, this approach neglects the ways in which defendants are able to affect case outcomes, and that experienced defendants may be better positioned to do so. Contrasting expectations from the popular prosecutor-focused paradigm, we borrow insights from the legal psychology, deterrence, and rational choice literatures to consider whether prior experiences shape defendants' ability to secure dismissals. Using administrative data collected from an urban prosecutor's office, we find evidence that there is considerable nuance to the impact of defendants' prior record. Although more prior cases generally decrease the likelihood that a defendant's current case will be dismissed, higher rates of prior conviction avoidance notably increase it. Findings are consistent with the possibility of an experiential effect in defendant decision-making, in which defendants use prior court experiences to refine their estimates of the likelihood of case dismissal and exert agency in accordance with those estimates.
We investigate how the affordances of an online context shape the processes of social learning. Using a dataset of more than 11,000 posts from the fraud subdread on the dark web forum Dread, we examine how affordances of platform governance, connectivity, anonymity, invisibility, asynchronicity, and limited oversight influence the components of social learning theory (SLT). We find that the affordances of this context introduce intermediaries who control the accessibility and visibility of possible associations, shape the breadth and other characteristics of these associations, facilitate a novel way to learn definitions, and influence reinforcement by separating offenders from consequences, providing vicarious rewards, and enabling feedback. They also shape imitation by adding ways to assess potential models and to virtually imitate them. Our findings suggest the need to extend SLT in online contexts with the concepts of mediated differential association, the window of associations, vicarious reinforcement, repeated definitional exposure reinforcement, and virtual imitation. We also discuss how our findings inform the applicability of traditional criminological theories to cybercrime, online criminal subcultures, and policies and practices aimed at disrupting online fraud.
Recent tests of deterrence theory have indicated that individuals update perceptions of risk after experiencing an arrest in rational and Bayesian ways. Still, research in psychology has found that the impact of new information and experiences on perceptions is characterized by recency and decay effects, whereby "signals" lead to immediate changes in perceptions, but as time passes, individuals gravitate back toward prior beliefs. Drawing on this phenomenon, we assess the extent to which risk updating post arrest is subject to recency and decay-that is, whether experiencing an arrest leads to immediate changes in perceived arrest risk (a recency effect) but that the signal effect decays over time. We test these ideas using data from the Pathways to Desistance study and the Denver Youth Survey. Consistent with recency effects, the results suggest that more recent arrestees demonstrate relatively large and statistically significant risk updating, while those arrested earlier in a recall period display small and nonsignificant signal effects. Further, lowess smoothing functions of estimated marginal signal effects indicate an immediate risk updating effect, but one that declines with more time between signal and perception. The theoretical and practical implications for deterrence are discussed.
