The U.S. Office of Justice Programs’ National Institute of Justice commissioned a study to advance knowledge and understanding of deaths in jail and prison custody and to develop recommendations that support efforts to prevent and reduce such deaths. This report summarizes key findings and recommendations from the national-level review, case studies conducted with ten criminal justice agencies, and an analysis of the Death in Custody Reporting Act data collected by the U.S. Bureau of Justice Assistance between 2020 and 2023. Key findings include the prevalence of arrest-related deaths increased 17% between 2013 (1,778 deaths) and 2020 (2,100 deaths); in 2019, 1,200 individuals died in local jails, reflecting a 5% increase from the previous year and a 33% increase from 2000; and the number of individuals who died while in the custody of state prisons declined 7% from 2018 to 2019. However, the mortality rate increased between 2001 and 2019, peaking at 347 per 100,000 incarcerated individuals in 2018. The causes of death varied, including officer-involved shootings and vehicle incidents, illness, suicide, and homicides. The report provides several recommendations to address deaths in custody for law enforcement, jails, and state prisons including leveraging alternative and co-response models, implementing de-escalation training, identifying and mitigating facility-based and individual risks, delivering mental health care, implementing contraband detection technologies, enhancing risk assessment and classification processes, and supporting comprehensive healthcare access.

The consensus that disasters do not cause an increase in crime rates is receiving renewed attention. In recent years, research has emerged that challenges this consensus by positing that crime rates and the type of crime may vary depending on the phase of the emergency. To address this, this research utilizes comprehensive crime data from the National Incident-Based Reporting System and hazard event data from the Spatial Hazard Events and Losses Database for the United States. Employing regression discontinuity design principles, swaths of linear regression models across different time scales are fitted, yielding nearly 120 statistically significant coefficients. The findings reveal correlations between certain natural hazard types and changes in crime rates. For instance, a correlation between winter weather hazard events and a subsequent drop in crime rates is observed whereas severe thunderstorms were associated with an increase in crime rates. Additionally, an increase in crime rates following natural hazard events that were observed in the shorter time scales (e.g., hail, tornadoes) did not persist into the longer time scale, where negative treatment effects and a negative change in trend were found.

With foundations in the Work Works model, Bridge House’s Ready to Work program combines three elements—paid work, housing, and supportive services—to help people experiencing homelessness in and around Aurora and Boulder, Colorado, transition into employment and permanent housing, and remain in substance use recovery. This report examines the program’s components and operations, its impacts on housing instability, incarceration, and employment, and implications for policy and practice. It found that graduates were significantly less likely to experience housing instability within 6 to 36 months of graduating compared with people who dropped out and people who were referred but didn’t enroll. In addition, during the program and immediately after graduating, participants who completed the Ready to Work program (and who therefore engaged with the program longer than people who dropped out) experienced improved housing stability and reduced incarceration compared with people who dropped out and people who didn’t enroll. Finally, the report suggests replicating facets of the Ready to Work program, like strong hiring practices and diversified contracts to support employment.

Going to college has consistently conferred a large wage premium. The research team shows that the relative premium received by lower-income Americans has halved since 1960. The research team decomposes this steady rise in collegiate regressivity using dozens of survey and administrative datasets documenting 1900–2020 wage premiums and the composition and value-added of collegiate institutions and majors. Three factors explain 80% of collegiate regressivity's growth. First, the teaching-oriented public universities where lower-income students are concentrated have relatively declined in funding, retention, and economic value since 1960. Second, lower-income students have been disproportionately diverted into community and for-profit colleges since 1980 and 1990, respectively. Third, higher-income students' falling humanities enrollment and rising computer science enrollment since 2000 have increased their degrees' value. Selection into college-going and across four-year universities are second-order. College-going provided equitable returns before 1960, but collegiate regressivity now curtails higher education's potential to reduce inequality and mediates 25% of intergenerational income transmission.

Cross-age tutoring is an educational model where an older tutor is paired with a younger tutee, valued for its economic advantages and capacity to engage participants. This model leads to improvements in both academic performance and behavior, as evidenced by a 2016 meta-analysis, which reported statistically significant positive effects across various educational settings and demographic groups. In this study, the research team aimed to update this previous meta-analysis by systematically examining 32 studies on cross-age tutoring. In the research team’s updated meta-analysis, the research team observed a small to moderate positive effect on academic outcomes for both tutors and tutees. The overall effect size was 0.34, with tutees benefiting at 0.33 and tutors at 0.39. The research team’s moderator analyses revealed no significant differences in impact from the number of sessions, tutor type, tutee risk status, or subject area. These findings highlight the broad applicability and effectiveness of cross-age tutoring, particularly emphasizing the benefits of using older students as tutors in resource-limited settings. Further research is recommended to explore additional influencing factors.

Persistent gender disparities in Science, Technology, Engineering, and Mathematics (STEM) fields, even when young girls perform as well in STEM in school as boys, highlight the potential importance of preconceived views of STEM work in this difference and the potential need for role models to upend these views. In this study, the research team investigated whether female math tutors positively influence girls’ STEM interest, attendance, and math performance. The research team randomly assigned 422 ninth grade students taking Algebra 1 in an urban New England school district to either same-gender or opposite-gender tutors. Girls paired with female tutors reported significantly higher STEM interest (0.73 SD) compared to those assigned to male tutors and were more likely to pass the course with a C- or better (3.9 percentage points). The research team found no evidence that students’ attendance patterns systematically differed based on their tutors’ gender. The effects appear stronger for students working with tutors in-person, as opposed to virtually, and during the school day, as opposed to after school. As the first experimental study of the impact of the tutor-student gender match, the research provides evidence that pairing girls with female tutors in school can enhance girls' STEM self-concept and academic performance.

The share of mothers who worked before their first birth more than doubled to 78% over the past half century, according to a new U.S. Census Bureau report. The share of fathers who took paid leave after the birth of their first child rose in recent decades too. The policy and employment landscape changed in that period, including the introduction of the Family and Medical Leave Act (FMLA) which guarantees eligible employees up to 12 weeks of unpaid, job-protected leave. The Census Bureau report, based on the 2022 Survey of Income and Program Participation (SIPP), explores parental leave and employment patterns among first-time mothers and fathers in the decades leading up to 2022. The share of fathers working before their first child’s birth remained stable (around 76%) for the cohorts whose first born came prior to 1981 until the 2006–2010 timeframe. In contrast, the share of first-time mothers who worked before their child’s birth was as low as 38% for the cohort whose child was born prior to 1981, climbed to 53% from 1981 to 1985 and remained relatively stable at around 60% from 1986 to 2015. But by the 2021–2022 cohort, the share of first-time parents who worked before their first child was born had increased for mothers (78%) and remained stable for fathers (81%).

Wage records are administrative data collected quarterly as part of the federal-state Unemployment Insurance (UI) program, which provides cash benefits to eligible unemployed workers. Each state administers a separate UI program, following the same guidelines established by federal law, including reporting wage records. All wage records contain an employee’s total wages and provide a rich source of labor market data without creating additional reporting burdens for states. Wage records are available for most regular, paid employees — covering about 90% of the labor force. While all states report total wages, few collect more detailed information. Enhancing these records by adding fields like job title or location can help identify jobs most likely to create upward mobility, where they’re located and the training needed to access them. Enhanced wage records allow states to measure job quality thoroughly. Enhanced wage records could include data about: (1) occupation to identify jobs that offer the best economic opportunities; (2) pay rate for more granular information on hourly wages; (3) primary work location to identify regions where occupations are underrepresented or wages are low; (4) worker type to provide better context for reported wage levels (e.g., full-time, apprenticeship, etc.); and (5) worker demographics to help understand how different groups of workers are faring. States that collect enhanced wage records include Alaska, South Carolina, Virginia, and Washington.

Chronic obstructive pulmonary disease (COPD) is a group of diseases of the lung for which there is no cure and that worsens over time. In 2023, COPD was the fifth leading cause of death in the United States, resulting in 141,733 deaths. COPD incurs annual medical costs of $24 billion among adults age 45 and older. The most common COPD types are emphysema and chronic bronchitis. This report uses 2023 National Health Interview Survey data to present age-adjusted estimates of COPD in adults age 18 and older by selected sociodemographic and health characteristics. Key findings include that in 2023, the age-adjusted prevalence of diagnosed COPD in adults age 18 and older was 3.8%, with women more likely to have COPD (4.1%) than men (3.4%). COPD increased with increasing age, from 0.4% in adults ages 18–24 to 10.5% in those age 75 and older. Asian non-Hispanic adults were less likely than adults of all other racial and ethnic groups to have COPD. The prevalence of COPD decreased with increasing family income and varied by region. Adults with fair or poor health were about five times as likely to have COPD (11.5%) than adults with excellent, very good, or good health (2.2%), and the percentage increased with increasing level of difficulties in functioning.

The report highlights the commonality of co-occurring mental health and substance use disorders and negative outcomes in the absence of evidence-based integrated care. It also provides an overview of treatment barriers and potential solutions. According to the 2022 National Survey on Drug Use and Health report, close to 21.5 million adults aged 18 or older had co-occurring mental health issues and substance use disorders in the past year. Of these adults, nearly half did not receive treatment for either condition. Of those adults who did receive treatment, most only received mental health treatment. Potential barriers to treatment may include provider waitlists and referral policies, less generous insurance benefits for substance use treatment compared to services for mental disorders, the lack of integrated care services, and a need to travel long distances to receive integrated care. Individuals with co-occurring mental health issues and substance use disorders are at risk for several negative socioeconomic and health outcomes that may subsequently impact many facets of their lives, including increased risk of self-harm and lower physical and mental health functioning. The U.S. Substance Abuse and Mental Health Services Administration provides six guiding principles for working with individuals with co-occurring mental health issues and substance use disorders including using recovery perspective, developing a holistic viewpoint, developing a phased approach to treatment, addressing specific real-life problems early in treatment, planning to address individual’s cognitive and functioning concerns, and using support systems to maintain treatment effectiveness.