For years there has been a push to reduce incarceration for certain drug offenses under the assumption that addiction is a health problem better served by treatment. Law Enforcement Assisted Diversion (LEAD) is a replicable program that diverts certain offenders away from jail and prosecution and into community-based systems of care. Over the past decade, LEAD has been implemented in more than 20 states based on claims that it reduces recidivism and improves outcomes like employment and homelessness. This report from the Washington Policy Center reviews the public citations of LEAD’s claimed evidence from 2015 to 2026, finding significant factual and statistical errors and almost exclusive reliance on inaccurate and unsupported descriptions. Per the author, for example, one study commonly used as evidence in support of LEAD confuses concepts like odds, probability, and absolute versus relative change as well as substantially overstating findings from original research that showed little or no clear positive effect of LEAD on measures like recidivism. The author concludes that inaccurate reporting on LEAD has permeated LEAD adoption and funding across more than twenty states and been codified into congressional appropriations, government grants, private donations, municipal budgets, and state and local legislation.

Cognitive bias in criminal investigation presents a distinct epistemic and procedural problem. Research on confirmation bias, asymmetrical skepticism, contextual information, and investigative tunnel vision indicates that these influences can affect decision-making even among experienced practitioners, although the magnitude and conditions of the effects vary across tasks and domains. This article examines cognitive bias at the investigative stage, distinguishing it from the separate problem of judicial evaluation of expert evidence. The article proposes a four-part procedural framework: hypothesis plurality, task-specific information management, active disconfirming inquiry, and documented revision. It applies this framework to witness assessment, suspect interviews, crime-scene interpretation, forensic submissions, and organizational review. The analysis does not claim that investigators can become cognitively neutral or that investigative focus is itself improper. Its narrower proposition is that procedural fairness and evidentiary reliability are better protected when an investigation remains capable of testing and revising its own working assumptions. The article concludes that the legitimacy of criminal investigation depends not on the appearance of certainty, but on the capacity of investigative procedures to expose, evaluate, and, when necessary, abandon hypotheses that no longer adequately explain the evidence.

Drawing on survey responses from more than 2,200 incarcerated individuals across 38 states, this report reveals how abuse, neglect, exploitation, and instability often precede youth incarceration. The report centers on adverse childhood experiences (ACEs). The average ACE score among surveyed individuals was 6.31 out of 10—a level that indicates severe trauma and is far higher than rates found in the general population. This suggests that many children prosecuted as adults were victims long before they became defendants, and, per the authors, many of these children encountered punitive systems that criminalized trauma-related behaviors rather than receiving support, treatment, or intervention. The authors characterize children prosecuted as adults using three major categories: youth experiencing severe abuse, neglect and trauma that contributed to behavioral challenges; victims of forced criminality and trafficking who were coerced into criminal acts; and children who committed offenses against abusive adults or exploiters. These pathways often intersect with poverty, unstable housing, family disruption, and systemic inequities.

Debates over whether college is worth it tends to swing between the claim that bachelor’s degrees lead to strong earnings gains and the claim that too many graduates struggle to find good jobs. The authors of this study argue that these contradictory claims exist because the benefits of higher education are usually calculated as average returns or as returns for the median student. This study examines the relationship between bachelor’s degree return on investment (ROI) and geography. Findings include that bachelor’s earners living where there is a greater concentration of college graduates is linked to greater likelihood of positive bachelor’s degree ROI, and the number of college graduates out-earning high school-educated peers has declined in most states over the past decade. Per the authors, areas with the largest percentage of bachelor’s degree holders who experience positive ROI are Washington D.C. (82%), New York (80%), California (78%), Alaska (77%), and Illinois (76%). Areas with the smallest percentage of bachelor’s degree holders who experience positive ROI are Vermont (56%), Wyoming (59%), North Dakota (59%), Rhode Island (61%), and Idaho (61%). In Florida, the share of bachelor’s holders who experience a positive ROI is 74%, a decrease of 9 percentage points from 2013.

School closures are often justified through seemingly neutral criteria such as enrollment or performance, but these metrics can unintentionally deepen educational disparities. This study uses a large urban district’s administrative data to simulate 5,040 closure scenarios, systematically varying seven policy design principles, including proximity, enrollment, seat utilization, building quality, academic performance, demographic disproportionality safeguards, and ordering of schools considered for closure. Key findings include that safeguards explicitly designed to prevent demographic disproportionality improve fairness but reduce the cost savings and seat reductions needed to balance capacity and demand; common criteria like enrollment do little to advance either efficiency or equity; and how schools are ranked for evaluation is a promising policy lever. Per the authors, ranking a composite score comprised of metrics for academics, resource use, family choice and demand, teacher turnover, student enrollment, building condition, number of proximate schools, and number of specialized programs is the only approach to ordering school closures they tested that simultaneously improves demographic proportionality, total savings, and seat removal.

This report updates a two-decade body of research tracking the teacher pay penalty—the pay gap between public school teachers and their college-educated peers. Because public school teachers must attain at least a bachelor’s degree to teach in the U.S., the study compares weekly earnings of public school teachers (elementary, middle, and secondary) with those of college graduates who chose other careers. Findings include that in 2025, the teacher pay penalty stood at approximately 25.2%—meaning teachers earned about a quarter less than comparable college graduates in other professions. Per the author, this a slight improvement from 2024’s record high (26.9%), but still more than four times the penalty recorded in 1996 (6.1%). In addition, inflation-adjusted weekly wages for public school teachers have fallen 6.2% over the last three decades, while wages for other college graduates rose 28.8% over the same period. Across states, the teacher pay penalty in 2025 ranged from 10.4% in Rhode Island to 40.7% in Colorado. Per the study, the teacher pay penalty in Florida was 24.2% in 2025.

The City of Milton, Florida has considered ways to effectively deal with increased wastewater treatment demands resulting from growth and development. These considerations have included constructing a proposed new wastewater treatment facility (WWTF). This operational audit focused on the progress that the city had made in addressing findings from a previous Auditor General report. The audit disclosed that the City had corrected 5 findings from the previous report and applied for several grants. During the period October 2024 through February 2026, the city expended $3.9 million on the proposed new WWTF and effluent disposal projects. City personnel followed the policy and recognized sales tax exemption savings of $10,333 related to the purchase of materials totaling $147,465 for a construction project. In December 2025, the city purchased land related to the proposed new WWTF for $1.8 million. The Auditor General’s examination of city records disclosed that the City Council obtained and considered three independent appraisals with property values ranging from $780,000 to $1.5 million before approving the purchase, by extraordinary vote, in accordance with state law. In February 2025, the City Council approved revisions to its anti-fraud policy that adequately addressed Auditor General recommendations.

The economic burden of natural disasters is shaped by who is exposed, who experiences losses, how those losses alter individual economic trajectories, and how affected communities evolve in the aftermath. For households whose homes are destroyed by wildfires, recovery implies ground-up reconstruction rather than repair, with potential consequences for economic trajectories and the composition of affected neighborhoods. The authors linked property-level damage data from over 100 U.S. wildfires between 2003 and 2022 to administrative tax and Census data to understand individual-level relationships between wildfire damage and economic outcomes. Key findings include that residents of burned areas are wealthier than the U.S. population on average, but within wildfire perimeters the likelihood that a home burns is negatively correlated with pre-fire income. In addition, home destruction imposes costs beyond rebuilding the property: occupants of destroyed homes experience lower income for three years post-fire, producing cumulative income losses equal to 26% of pre-fire annual income. Post-wildfire community recovery is slow and accompanied by changes in composition. For example, four years after a fire, half of destroyed addresses remain unoccupied and reoccupied addresses house residents with substantially higher incomes than before.

Remote work has quadrupled in the half-decade since the onset of the COVID-19 pandemic, rising from 7% of American workers in 2019 to 28% in 2023. This study examines the effect of remote work on self-reported isolation and mental health using a nationally representative sample (n=588,322) of American workers conducted from 2011 to 2024, omitting the peak pandemic years of 2020-2021. The authors compared changes in mental health among people in remotable work jobs—who have experienced a large and persistent rise in remote work since COVID-19—to people in nonremotable jobs, where remote work increased far less. Findings include that remote work increases time spent alone, worsens mental well-being across multiple measures, and increases the use of mental health services and prescriptions, with these effects concentrated among people who live alone. The authors estimate that the rise in remote work explains approximately one third of the increase in isolation and mental distress between 2011-2019 and 2022-2024.

The weed killer Roundup, which contains glyphosate, has been used in the U.S. since the 1970s. Per a nationally representative sample of U.S. adults (n=1,639), two-thirds of Americans favor banning the use of the herbicide glyphosate, though most survey respondents had read, seen, or heard little or nothing about glyphosate in recent months (83%) and many were uncertain about its health effects (46%). Other findings include that banning glyphosate cuts across political and ideological groups, with 72% of people who identify as supporters of the Make America Healthy Again (MAHA) movement and 70% of those who do not support MAHA in favor of banning glyphosate. Per survey findings, support for a ban is also strong among both major parties: 70% of self-identified Republicans and 69% of self-identified Democrats are in favor. When asked which of three statements is most accurate about the use of glyphosate, a third of survey respondents (33%) said it is most accurate that glyphosate increases the U.S. food supply by helping farmers control weeds; nearly half of respondents (48%) said they were unsure; 12% said that glyphosate has no effect on food production, and 6% thought that the use of glyphosate decreases the food supply.

Americans use artificial intelligence chatbots for a host of health-related reasons, from helping make diagnoses on their own to learning more about information they may have gotten from a doctor. This report explores why U.S. adults use chatbots for health using data from a nationally representative sample of U.S. adults (n=3,488). In all, 34% of survey respondents use chatbots for health-related reasons. Among these chatbot users, 28% say that getting health information quickly is a motivation for using chatbots. Other top reasons for using chatbots include understanding the causes of symptoms (25%), getting health information at little to no cost (22%), getting information about treatments (22%), and learning more about a diagnosis (22%). Nearly all chatbot users (95%) find the information at least somewhat helpful. While chatbot health users find these tools helpful, they have varying comfort levels sharing personal health information with chatbots. For example, chatbot health users are about as likely to be extremely or very comfortable (29%) sharing personal health information with a chatbot as they are to be not too or not at all comfortable (26%). The largest share sits in the middle, saying they would be somewhat comfortable sharing personal health information with chatbots (42%).