This report is the seventh in a series in response to the Death in Custody Reporting Act of 2013 (P.L. 113–242) that requires federal law enforcement agencies to report information about the death of any person who dies while detained, under arrest, being arrested, or in the custody of federal law enforcement officers. It describes decedent, incident, and facility characteristics for deaths in federal custody and during arrest by federal law enforcement agencies during Fiscal Year 2024. Homicides accounted for the largest portion (39%) of arrest-related deaths, followed by suicides (38%) and accidents (22%). In arrest-related deaths, 94% of decedents were male, 65% were white, and 51% were ages 25 to 44. In 53% of arrest-related deaths, law enforcement officers were serving a warrant when they made initial contact with the decedent. A violent offense was the most serious offense allegedly committed by decedents in 49% of arrest-related deaths.

This report summarizes key findings from a study on the juvenile justice system in South Carolina. Approximately 288,000 youth across the nation are placed on some form of probation every year; less than 30% of these placements are the direct result of an offense that involved another person. Approximately 20% of all youth who are detained and 15% of youth who are incarcerated in state custody are detained or incarcerated due to technical violations of the conditions of their probation. In the school year prior to system involvement, South Carolina’s diverted youth and youth on probation missed an average of 26 and 42 days of school, respectively. On average, the school attendance of youth who were diverted or placed on probation declined in the first year of their system involvement. When controlling for other factors, youth who were 15 years of age or older missed, on average, approximately 7.5 more school days during their first year of system involvement compared to the prior year, as opposed to their peers under the age of 15 who missed approximately 1.5 more school days.

Desistance from crime, or the act or process of ceasing, especially from criminal or antisocial behavior, has become a central concept in contemporary criminology. This article argues that recidivism, although useful as an observable outcome, is an insufficient proxy for the desistance process itself. Drawing upon life-course criminology, developmental criminology, identity theory, social-control perspectives, and contemporary methodological research, the article distinguishes three analytically related but non-identical dimensions of change: behavioral desistance, identity transformation, and relational or structural desistance. It further examines the methodological consequences of treating desistance as a binary state rather than a temporally extended process. The analysis develops a process-oriented framework in which desistance is understood as a dynamic reconfiguration of the relationship between offending behavior, identity, social bonds, opportunity structures, and future orientation. This framework does not treat desistance as synonymous with rehabilitation, nor does it assume that an absence of recorded offending necessarily demonstrates complete behavioral transformation. Criminological research should distinguish the observable endpoint of non-offending from the broader process of becoming progressively less involved in crime. This is essential for theoretically valid research, accurate longitudinal measurement, and the development of criminal-justice interventions that evaluate change without reducing it to recidivism alone.

Participation in registered apprenticeship programs by young people ages 16 to 24 more than doubled between 2015 and 2025, rising from 185,201 registered youth apprentices to 411,371. Despite this expansion, no standard definition of a youth apprenticeship program exists, and previous approaches apprenticeship classification have defined youth apprenticeship by apprentices’ age rather than by program features. This report addresses the gap in youth apprenticeship classification by defining four models of apprenticeship based on delivery partners: (1) high school-aligned youth apprenticeship programs; (2) college-aligned youth apprenticeship programs; (3) career and technical education (CTE) center-aligned youth apprenticeship programs; and (4) other youth-focused programs. The authors used their classification system to describe youth apprentices nationwide, finding that all of the above models of youth apprenticeship serve a higher share of young women than the 12.7% of youth apprentices in adult programs who are women, and all serve a higher share of rural apprentices than the 6.6% of youth in adult programs who live in rural areas. And while models of youth apprenticeship vary widely in the types of occupations they serve, all models have a significantly lower share of youth apprentices training in construction occupations than the 63.8% of youth in adult programs.

Research shows that access to postsecondary education in prison increases the odds that individuals will secure employment after release, supporting economic stability and decreasing the likelihood of recidivism. State policymakers play an important role in ensuring incarcerated learners have access to the programs and support they need to succeed. This 50-state comparison provides information on which state agencies have duties and responsibilities for programs that offer postsecondary credit, degrees, or industry-recognized credentials and on the statewide coalitions, consortia, and task forces that provide advice or support. In Florida, the Department of Corrections is required to enter into agreements with colleges and universities for educational programs. There is not an active statewide prison coalition, consortium, or task force to advocate or coordinate higher education opportunities for prisoners. States that have such a statewide organization include Arizona, Maryland, New Hampshire, North Carolina, Oregon, Pennsylvania, Texas, Utah, Virginia, and Washington.

Students who leave college often lack clarity about their path through college. In a survey of more than 8,000 public two-year college students, 96% of first-year students expressed a strong commitment to their education. However, real and persistent challenges frequently derail this initial motivation. In a recent study of students at four community colleges who stopped out before the start of their second year, financial hardship and academic struggles were the top two reasons reported for leaving. Moreover, a third of respondents cited student-related stress and shifting career goals as contributing factors. To help colleges address associated challenges, the brief describes actionable frameworks like the Ask-Connect-Inspire-Plan (ACIP) onboarding approach, and it reintroduces the Community College Program Mapper—a tool designed to help institutions visualize program enrollment and track student plan adoption. Accompanying this report is a series of case studies from diverse community colleges nationwide (including Florida’s Tallahassee State College). These profiles explore how institutions of varying sizes, regions, and governance structures have successfully redesigned business processes, leveraged technology, and engaged faculty and staff to scale career and educational planning.

A data center is a physical facility that houses computer systems for managing and transmitting data. The federal government has been pursuing policies to promote data center development, particularly as a primary component of artificial intelligence (AI), a "critical and emerging technology" with enormous data storage and processing requirements. To date, data centers have generally secured power from their local electric utilities. Reportedly, some utilities are experiencing infrastructure constraints and cannot accommodate new data center requests until new grid infrastructure—power plants, transmission facilities, and distribution facilities—is built. Typically, building grid infrastructure takes several years longer than building data centers. This report discusses frequently asked questions about data center electricity use, the potential impacts of data center growth on the electric grid, and government actions related to data center development. The report is intended as an introduction to data center electricity issues and does not provide in-depth coverage of all data-center- and electric-grid-related issues. Topics include how data centers use electric power, where data center growth is happening, and how have data centers affected electricity prices.

Many white-collar professions develop expertise through learning by doing, raising the question of whether artificial intelligence (AI) helps or hinders skill development. In a pre-registered three-month randomized controlled trial, the authors gave 133 practicing patent lawyers at eleven U.S. intellectual property law firms access to a custom AI drafting assistant and measured their performance while using AI and their professional judgment afterward without it. Findings include that AI access raised the quality of work delivered on benchmark patent drafting tasks at 10 days and 90 days, with larger gains among junior lawyers. After three months, all subjects redlined an existing patent application without AI, a core task of patent practice requiring expert judgment. Treated lawyers outperformed controls, but this advantage was concentrated entirely among senior lawyers. Junior lawyers showed no average gain; their scores instead dispersed, with sharply fewer mediocre scores offset by more poor and more good ones. The authors conclude that AI helped some junior lawyers acquire skills while hindering skill development for others. Among seniors, AI use produced more good scores and fewer mediocre and poor scores. The authors conclude that foundational expertise may be a prerequisite for extracting durable skill from AI-assisted practice.

In 2024, the ranked order of 9 of the 10 leading causes of death in the U.S. remained unchanged from 2023. The 10 leading causes of death in 2024 in ranked order were: 1) diseases of heart; 2) malignant neoplasms; 3) accidents (unintentional injuries); 4) cerebrovascular diseases; 5) chronic lower respiratory diseases; 6) Alzheimer disease; 7) diabetes mellitus; 8) nephritis, nephrotic syndrome and nephrosis; 9) chronic liver disease and cirrhosis; and 10) intentional self-harm (suicide). These causes accounted for 70.9% of all deaths occurring in the U.S. Rankings are presented by age, race, Hispanic origin, and sex. The 10 leading causes of infant death for 2024 in ranked order were: 1) congenital malformations, deformations and chromosomal abnormalities; 2) disorders related to short gestation and low birth weight, not elsewhere classified; 3) sudden infant death syndrome; 4) accidents (unintentional injuries); 5) newborn affected by maternal complications of pregnancy; 6) bacterial sepsis of newborn; 7) newborn affected by complications of placenta, cord and membranes; 8) respiratory distress of newborn; 9) diseases of the circulatory system; and 10) intrauterine hypoxia and birth asphyxia.

Each year, bad actors defraud Medicare and taxpayers of millions of dollars by billing for durable medical equipment, such as walkers and hospital beds, that enrollees do not need or never receive. This white paper presents actions that the U.S. Centers for Medicare and Medicaid Services (CMS) and others can take to stop bad actors from exploiting Medicare, focusing on three elements that bad actors need to commit fraud: a Medicare-enrolled supplier, a physician order, and an enrollee identification number. For each of these elements, the authors describe how bad actors commit fraud, identify existing vulnerabilities, and present calls to action to address the underlying problems that allow fraud to persist. Calls to action related to Medicare-enrolled suppliers include improving detection of unreported changes of ownership and straw owners and increasing oversight of newly enrolled suppliers. Actions that CMS can take for physicians’ orders include harnessing new technologies and analytical tools to better detect and stop payments for fraudulent claims (e.g., fake physician orders). Finally, to prevent bad actors from stealing or purchasing Medicare enrollee identification numbers, CMS could partner with social media companies to prevent the sale of enrollee numbers on their platforms and change the way Medicare protects enrollee identification numbers.

Adolescent childbearing is associated with significant social, health, and financial risks for teens, their families, and society more broadly. This report describes teen birth rates, which are measured with the number of live births per 1,000 teen mothers and have markedly declined in recent years. For example, teen births peaked in 1957 at 96.3, decreased most years from the 1960s through the 1980s, and reached a historical low of 12.6 in 2024. Other findings include that the teen birth rate has decreased across all racial and ethnic groups in recent years but declined more for certain groups than others, with non-Hispanic Asian teens experiencing the greatest percentage decline from 2016 to 2024 (59.0%) and non-Hispanic Black teen experiencing the least percentage decline (37.3%). In 2024, the state with the lowest teen birth rate was New Hampshire (4.6), while the state with the highest teen birth rate was Mississippi (22.8). In 2024, Florida’s teen birth rate was 11.8, a substantial decrease from 2014 (22.5). The number of second (and additional) births to teen parents has also declined over time, with an 80% total decline from a high of 8.1 in 2007 to a historical low of 1.7 in 2024.