Robert J. Sampson’s research while affiliated with Harvard University and other places
What is this page?
This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.
Most homicides in the United States are committed using a handgun, but little research examines gun carrying over critical stages of the life course and changing contexts of violence. Notably, although most of the handgun homicides are committed by adults, most research on concealed gun carrying focuses on adolescents in single cohort studies. Using more than 25 years of longitudinal multicohort data from Chicago, 1994–2021, we show that pathways of concealed gun carrying are distinct between adolescence and adulthood. Adolescent carrying is often age-limited and responsive to direct exposure to gun violence (witnessing and victimization), while adult carrying is a persistent behavior that is less tied to direct exposure. The onset of concealed carry is also a strong predictor of later gun use (shooting or brandishing), and we find distinct patterns of gun use between individuals who first carry in adolescence versus adulthood. We discuss the implications of these dual pathways for research and policies on firearm use.
Although racial disparities in criminal justice contact are long‐standing and the subject of continuing public debate, few studies have linked early‐life social conditions to racial disparities in arrest over the life course and in changing times. In this article, we advance and test a theoretical model of racial inequality in long‐term arrest histories on a representative sample of nearly 1,000 individuals from multiple birth cohorts in the Project on Human Development in Chicago Neighborhoods. Large Black–White disparities in arrests from ages 10 to 40 arise from racial inequalities in exposure to cumulative childhood advantages and disadvantages rather than from race‐specific effects. Smaller but meaningful Hispanic–White gaps follow a similar pattern, and the same explanations of racial disparities hold across different offense types and across birth cohorts who came of age at different times during 1995 to 2021. These findings indicate that inequalities in early‐life structural factors, which themselves are historically shaped, trigger processes of cumulative advantage and disadvantage that produce racial disparities in arrests over the life course and that persist across different points in contemporary history.
Intergenerational transmission processes have long been of interest to demographers, but prior research on the intergenerational transmission of criminal justice contact is relatively sparse and limited by its lack of attention to the correlated “family troubles” and familial incarceration that predate criminal justice contact. In this article, we provide a test of the intergenerational transmission of criminal justice contact after adjusting extensively for these factors that predate such contact by linking longitudinal data from the Project on Human Development in Chicago Neighborhoods with official arrest histories from 1995 to 2020. The results provide support for three conclusions. First, parental criminal justice contact is associated with a shorter time to first arrest and a larger number of arrests even after rigorously accounting for selection. Second, robustness checks demonstrate that neither the magnitude nor the significance of the findings is sensitive to model choices. Third, associations are strongest among White individuals and inconsistently significant for African American and Hispanic individuals. Despite large recent crime declines, the results indicate that parental criminal justice contact elevates the criminal justice contact of the adult children of the prison boom, independent of the often-overlooked troubles that predate criminal justice contact, and that these associations are strongest among the White population.
This article presents a theoretical and methodological framework for comparative urban studies grounded in the proposition that a neighborhood depends not only on its own conditions, as typically conceived, but also the conditions of the neighborhoods to which its residents are connected, through networks of everyday urban mobility. Based on this framework, we highlight three arguments and associated applications based on the analyses of networks of movement throughout large American cities. The first is that even though residents of disadvantaged neighborhoods travel far and wide, their relative isolation and segregation persist. Second, mobility-based socioeconomic disadvantage, or what has been termed triple neighborhood disadvantage, explains neighborhood well-being independent of residential-based disadvantage. Third, a city’s level of social connectedness depends on how uneven and concentrated networks of everyday mobility are among its neighborhoods, which in turn are hypothesized to predict social behavior across cities beyond that expected by their residential-based segregation. The results offer a new way of thinking about neighborhood effects, the dynamics of everyday urban mobility, spatial inequality, and social segregation that can be studied in a comparative framework in cities anywhere.
Nearly 35 years ago, Sampson and Laub popularized the concept of desistance from crime and isolated core factors that promote and inhibit this process. In this article, we introduce the concept of intergenerational desistance and provide guidance on measuring and explaining this process, encouraging researchers to think of the life-course of crime in terms of both individuals and generations. We first review research on the intergenerational transmission of family criminality and criminal justice contact, relying also on research outside of criminology to highlight how using broader conceptions of the family, including social parents, entire generations, and three (or more) generations could enliven this area. Bridging these literatures allows us to then introduce the concept of intergenerational desistance and elaborate on the concept of intergenerational escalation and demonstrate how they can be measured using data from the Project on Human Development in Chicago Neighborhoods (PHDCN). We close by developing a research agenda for considering intergenerational desistance and escalation in ways that enhance our understanding of how the life-course of crime, criminal justice contact, and other troubles in life (e.g., with alcohol, drugs, and mental health) progress through families.
Expected final online publication date for the Annual Review of Criminology, Volume 7 is January 2024. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.
Risk assessment instruments (RAIs) are widely used to aid high-stakes decision-making in criminal justice settings and other areas such as health care and child welfare. These tools, whether using machine learning or simpler algorithms, typically assume a time-invariant relationship between predictors and outcome. Because societies are themselves changing and not just individuals, this assumption may be violated in many behavioral settings, generating what we call cohort bias. Analyzing criminal histories in a cohort-sequential longitudinal study of children, we demonstrate that regardless of model type or predictor sets, a tool trained to predict the likelihood of arrest between the ages of 17 and 24 y on older birth cohorts systematically overpredicts the likelihood of arrest for younger birth cohorts over the period 1995 to 2020. Cohort bias is found for both relative and absolute risks, and it persists for all racial groups and within groups at highest risk for arrest. The results imply that cohort bias is an underappreciated mechanism generating inequality in contacts with the criminal legal system that is distinct from racial bias. Cohort bias is a challenge not only for predictive instruments with respect to crime and justice, but also for RAIs more broadly.
Importance:
The past quarter-century has seen both sharp declines and increases in firearm violence in the United States. Yet, little is known about the age of first exposure to firearm violence and how it may differ by race, sex, and cohort.
Objective:
To examine race, sex, and cohort differences in exposure to firearm violence in a representative longitudinal study of children who grew up in periods with varying rates of firearm violence in the United States and to examine spatial proximity to firearm violence in adulthood.
Design, setting, and participants:
This population-based representative cohort study included multiple cohorts of children followed-up from 1995 through 2021 in the Project on Human Development in Chicago Neighborhoods (PHDCN). Participants included Black, Hispanic, and White respondents from 4 age cohorts of Chicago, Illinois, residents, with modal birth years of 1981, 1984, 1987, and 1996. Data analyses were conducted from May 2022 to March 2023.
Main outcomes and measures:
Firearm violence exposure, including age when first shot, age when first saw someone shot, and past-year frequency of fatal and nonfatal shootings within 250 m of residence.
Results:
There were 2418 participants in wave 1 (in the mid-1990s), and they were evenly split by sex, with 1209 males (50.00%) and 1209 females (50.00%). There were 890 Black respondents, 1146 Hispanic respondents, and 382 White respondents. Male respondents were much more likely than female respondents to have been shot (adjusted hazard ratio [aHR], 4.23; 95% CI, 2.28-7.84), but only moderately more likely to have seen someone shot (aHR, 1.48; 95% CI, 1.27-1.72). Compared with White individuals, Black individuals experienced higher rates of all 3 forms of exposure (been shot: aHR, 3.05; 95% CI, 1.22-7.60; seen someone shot: aHR, 4.69; 95% CI, 3.41-6.46; nearby shootings: adjusted incidence rate ratio [aIRR], 12.40; 95% CI, 6.88-22.35), and Hispanic respondents experienced higher rates of 2 forms of violence exposure (seen someone shot: aHR, 2.59; 95% CI, 1.85-3.62; nearby shootings: aIRR, 3.77; 95% CI, 2.08-6.84). Respondents born in the mid-1990s who grew up amidst large declines in homicide but reached adulthood during city and national spikes in firearm violence in 2016 were less likely to have seen someone shot than those born in the early 1980s who grew up during the peak of homicide in the early 1990s (aHR, 0.49; 95% CI, 0.35-0.69). However, the likelihood of having been shot did not significantly differ between these cohorts (aHR, 0.81; 95% CI, 0.40-1.63).
Conclusions and relevance:
In this longitudinal multicohort study of exposure to firearm violence, there were stark differences by race and sex, yet the extent of exposure to violence was not simply the product of these characteristics. These findings on cohort differences suggest changing societal conditions were key factors associated with whether and at what life stage individuals from all race and sex groups were exposed to firearm violence.
... The widespread shift toward "shall-issue" or permit-free carry laws has altered the legal context of gun carrying, potentially influencing age-related trends and making the distinction between legal and illegal carrying more complex. As of 2023, more than half of US states have adopted laws that make it easier for individuals to carry concealed weapons without a permit (6). This legal evolution may contribute to changes in who carries guns, at what age, and for what reasons. ...
... Controlling for fifteen other explanations for crime, they found a statistically significant relationship between lead pollution and homicide across counties (see also Stretesky & Lynch 2004). These results have also been discovered at lower levels of aggregation in the cities of Chicago (Barrett 2017, Sampson & Winter 2018, Winter & Sampson 2017 and St. Louis (Boutwell et al. 2016). Much research remains to be performed on the effect of pollution on behavior. ...
... Social disadvantages such as poverty, low income, low homeownership rates are clustered spatially and racially, owing to the history of racial inequality in the United States (Massey & Denton, 1993;Sampson & Wilson, 1995;Sampson et al., 2005). A result of inequality is the clustering of disadvantage; and as Sampson and Neil (2024) describe, it positions White children to be generally born into more advantaged and Black children into more disadvantaged neighborhoods. In these contexts, they are differentially exposed to ecological risks that contribute to child welfare involvement and affect the capacity of families whose children have been removed to reunify or otherwise achieve permanency. ...
... Reference Title S1 [8] Evaluating Fairness in Predictive Policing Using Domain Knowledge S2 [5] Improving fairness in criminal justice algorithmic risk assessments using optimal transport and conformal prediction sets S3 [13] Predictive policing and algorithmic fairness S4 [22] Data augmentation for fairness-aware machine learning: Preventing algorithmic bias in law enforcement systems S5 [27] Is the data fair? An assessment of the data quality of algorithmic policing S6 [20] Cohort bias in predictive risk assessments of future criminal justice system involvement S7 [19] A penalized likelihood method for balancing accuracy and fairness in predictive policing S8 [25] AI For Bias Detection: Investigating the Existence of Racial Bias in Police Killings S9 [30] Accuracy and fairness in a conditional generative adversarial model of crime prediction S10 [12] Artificial fairness? Trust in algorithmic police decision-making S11 [31] Crowdsourcing perceptions of fair predictors for machine learning: A recidivism case study S12 [16] Singular race models: addressing bias and accuracy in predicting prisoner recidivism S13 [17] Algorithmic bias in recidivism prediction: A causal perspective S14 [15] Accuracy, Fairness, and Interpretability of Machine Learning Criminal Recidivism Models S15 [24] Case study: predictive fairness to reduce misdemeanor recidivism through social service interventions S16 [3] A review of predictive policing from the perspective of fairness S17 [7] Fairness, accountability and transparency: notes on algorithmic decision-making in criminal justice S18 [33] Achieving equity with predictive policing algorithms: a social safety net perspective and diversity of the field. ...
... Items were drawn from prior work about youth gun violence exposure which included focus groups and cognitive interviews with youth as young as age 10 (18). These were designed to expand upon previous measures that tend to focus mainly on seeing or being present when someone was shot (30) or census tract numbers of firearm homicides (16). Before these questions, participants were told that we were only asking about things they may have seen or heard about in real life-not things they may have seen on TV, in a movie, on the news or in a video game. ...
... For instance, residential mobility and neighborhood choice are shaped by socio-demographic factors (Foster 2017). Furthermore, neighborhood resources and processes are shaped by the sociodemographic composition of neighborhoods and communities, resulting in 'unequal neighborhood environment by racial and economic status' (Candipan and Sampson 2023). This is seen through decisions such as where to locate new schools, healthy grocery stores, and companies that pollute. ...
... In some cohort studies, such as the Panel Study of Income Dynamics-where respondents are spread out across the entire United States-such an assumption may basically hold. In many multilevel datasets, however, sample members are intentionally selected such that they are collectively representative of the larger units from which they are sampled, e.g., neighborhoods (Sampson et al., 2022). In these cases, neighborhood clustering will loom large. ...
... As the dust settled and narratives crystallized, it became evident that predominantly white zealots aligned with former President Donald Trump, and with deeply entrenched sentiments regarding electoral fraud, led the assault. Existing scholarship has linked insurrectionist support to factors such as racial status threat, racial resentment, and Christian nationalism, but has primarily focused on white attitudes (Barreto et al. 2023; Davis and Wilson 2023;Armaly, Buckley and Enders 2022;Bucci, Kirk and Sampson 2022). We focus our exploration on Latines, an under-explored group in this line of work, but whose support of the insurrection is even more puzzling. ...
... The data used come from an extension of the PHDCN and include 1,057 individuals from four different age cohorts (22). The study began in the mid-1990s with a representative sample of children, ranging from newborn to 18 y old, drawn from a representative sample of Chicago neighborhoods. ...
... The second issue is that Census data estimates marginal demographics of residents who live in each geographic area but does not contain information about the actual population present throughout the day who can interact with police. The racial composition of an area during the day could be significantly different than of the population who live there (Athey et al., 2021;Sampson and Candipan, 2023), which would bias our estimates of this probability and of the CRR estimand. ...
