Generative artificial intelligence (GenAI) models are revolutionizing the way people in many fields do their work. To explore how judges in the U.S. are considering and using GenAI tools, the Thomas Reuters Institute/National Center for State Courts AI Policy Consortium on Law and Courts' Governance and Ethics Working Group embarked on an interview-based research project and summarized its key findings. Findings included that GenAI can support, but not supplant, the essential work of judges as human decision-makers. Every judge who participated in the interviews was using GenAI in their own way. But there was unanimous consensus among the judges that, regardless of how they are using GenAI, judges must always remain "the deciders" who determine the ultimate outcome of any legal decision before them. The judges interviewed were identified as early adopters of GenAI, and they are using that technology in novel and innovative ways. The top benefit the judges identified was increased efficiency and using GenAI to help streamline certain tasks to save time. This could include using it for both administrative tasks (e.g., planning continued legal education trainings) and for judicial work (e.g., summarizing lengthy documents and filings). Early adopters are mindful of a variety of known risks and tailor their usage to responsibly mitigate or eliminate those risks as they understand them. Every judge who participated in the interviews stressed that it is critical that any judge who is considering using GenAI be aware of all of the potential risks of the technology in addition to the potential benefits.

States spend roughly $80 billion annually on corrections. Understanding how many people released from prison return within 3 years is one important way to assess whether that investment is working and where the system is falling short. Yet comprehensive 50-state recidivism data has long been severely limited. Nationally, reincarceration rates have dropped 20% over the past decade, from approximately 35% to 28%. As a result, 72 out of every 100 people released today do not return to prison within 3 years, compared with 65 a decade ago—about 30,000 fewer people returning to prison each year. In Texas alone, nearly 50,000 of the 58,670 people released in 2021 did not return to prison within 3 years. In 9 states, 1 in 5 or fewer people released were reincarcerated within 3 years. Eight states recorded a 30% or greater reduction in reincarceration rates over the past 10 years (California, New York, New Jersey, Colorado, Massachusetts, Michigan, Oregon, and Missouri), demonstrating that meaningful progress is possible across diverse policy environments. Florida’s reincarceration rate declined 16% between 2010 and 2020. Changes in reincarceration rates reflect many factors: changes in who is incarcerated, how people are supervised after release, changes in technical violation policies, and what reentry opportunities exist. No two states are the same, and national trends can obscure important variation.

Transfers from the Florida Lottery to the Educational Enhancement Trust Fund decreased in Fiscal Year 2024-25 to $2.2 billion, $225.5 million less than the prior fiscal year. According to Florida Department of the Lottery (Florida Lottery) officials, this decrease is due to a decline in draw game sales resulting from a lack of large jackpots. However, the department continues to outperform the legislative performance standard for its operating expense rate, which is the third lowest in the nation. In the past fiscal year, Florida Lottery continued to take steps to enhance revenue, including changing the product mix and increasing the number of locations with in-lane QUICKTICKET sales. The department also offered several promotions to increase player engagement and conducted player surveys to gauge interest in new games and identify ways to enhance player satisfaction. Adding new games, expanding product distribution methods, and increasing the retailer network could enhance lottery revenues, but these options have advantages and disadvantages and may have implications related to the Gaming Compact Between the Seminole Tribe of Florida and the State of Florida.

Teacher turnover negatively impacts student achievement, contributes to shortages, strains schools, and undermines efforts to build a well-prepared, stable, and diverse teacher workforce. Findings of this research show that about 1 in 7 public school teachers moved schools or left the profession between the 2020–21 and 2021–22 school years. A large majority of teachers who moved schools stayed within their district (45.1%) or within the state (36.5%), while 17.3% moved to teach in a different state. Among those who left teaching, 38.6% stayed in education in a non-teaching position, while another 31.2% retired. The remaining teachers who left teaching were working outside of education, caring for family, unemployed, or in college. Turnover rates were higher among early-career teachers; uncertified teachers; teachers teaching English as a Second Language, foreign languages, career and technical education, or special education; and teachers in city, charter, or high-poverty schools. After accounting for teacher and school characteristics, results show that teachers with higher salaries, more effective and supportive leadership, or higher job and workplace satisfaction are less likely to leave their schools. Policies to reduce turnover include improving compensation, working conditions, and access to strong preparation and supports for high turnover groups of teachers.

Average grades continue to rise in the United States, raising the question of how grade inflation impacts students. The research team provides comprehensive evidence on how teacher grading practices affect students' long-run success. Using administrative high school data from Los Angeles and from Maryland that is linked to postsecondary and earnings records, the research team developed and validated two teacher-level measures of grade inflation: one measuring average grade inflation and another measuring a teacher's propensity to give a passing grade. These measures of grade inflation are distinct from teacher value-added, with grade inflating teachers having moderately lower cognitive value-added and slightly higher noncognitive value-added. These two measures also differentially impact students' long-term outcomes. Being assigned a higher average grade inflating teacher reduces a student's future test scores, the likelihood of graduating from high school, college enrollment, and ultimately earnings. In contrast, passing grade inflation reduces the likelihood of being held back and increases high school graduation, with limited long-run effects. The cumulative impact is economically significant: a teacher with one standard deviation higher average grade inflation reduces the present discounted value of lifetime earnings of their students by $213,872 per year.

In the fall of 2023, Texas’ community colleges enrolled over 676,000 students, an increase of 4% since 2021 and a return to pre-pandemic levels. However, over 60% of students enrolling in a Texas community college in fall 2023 were deemed not ready to take college-level courses in math, reading, or writing. Currently, many incoming college students in Texas are placed into college-level coursework based only on their scores on the Texas Success Initiative Assessment 2.0, a statewide standard­ized test. However, evidence suggests that test-only placement systems can mistakenly assess substantial numbers of students as “not college-ready”; in other words, they may not reflect a student’s true ability to succeed in college-level coursework. In recent years, multiple measures assessment (MMA) has emerged as a promising alternative. MMA is a placement method that relies on more than one indicator—such as cumulative high school grade point averages, high school course-taking patterns, or standardized test scores—to assess students’ college readiness. This evaluation of MMA placement policies at two community colleges in Texas was designed to provide evidence on the effectiveness of MMA in Texas community colleges, with the potential to inform changes to statewide placement policies. . The initial findings from this study demonstrate that using multiple measures to place students increased first-semester college-level math completion rates by 1.9 percentage points but had no discernable effect on college-level English completion rates. Only one semester of data were available for this report, which makes degree-completion effects impossible to observe at this time. However, college credits earned—an early indicator of progress toward a degree—showed a statistically significant positive effect of 0.4 credits. These findings suggest that MMA can be used successfully in the state of Texas.

The North American Industry Classification System (NAICS) defines chemical manufacturing as the transformation of raw materials by a chemical process into everyday products. Businesses that manufacture chemicals run the gamut from refining oil into margarine to synthesizing willow bark into aspirin. The number of U.S. Chemical Manufacturing establishments increased 10.2% from 13,571 in 2017 to 14,961 in 2022 while the number of Manufacturing sector establishments as a whole decreased 1.7% during the same period, from 291,586 to 286,626. Chemical Manufacturing value of shipments also increased by 22.4% during that period, rising from about $735.9 billion to $901.0 billion (the data are not adjusted for price changes). Despite growth in the Chemical Manufacturing industry, job numbers in industries downstream from these firms were in flux. Between 2017 and 2022, the number of pharmacists, chemical engineers and chemists, and material scientists increased about 28.1%, 54.5% and 24.8%, respectively. Chemical technicians, however, experienced a 2.2% decrease over the same period. The number of Pharmacies and Drug Stores establishments declined along with 5.6% decrease in employment in the industry between 2017 and 2022. Chemical Manufacturing firms and revenues weren’t spread evenly across the nation with a significant portion of the industry’s workforce concentrated in the South and Midwest in 2023.

Congress and the executive branch have taken steps to improve the transparency of information on federal spending and programs. However, the Government Accountability Office (GAO) has found that challenges remain in various areas and has made recommendations to federal agencies and Congress to help address them. Agencies are required by law to report federal spending data to USAspending.gov, the government’s official public source of such data. While progress has been made to improve the data on USAspending.gov, GAO has continued to identify challenges. For example, federal agencies do not consistently report spending data for other transaction agreements—legally binding agreements other than standard contracts or grants that are not subject to certain federal acquisition laws and requirements. The GAO also has identified issues with the completeness and accuracy of data on USAspending.gov describing subawards—awards provided by a recipient to a subrecipient to carry out part of a federal award. Further, the Office of Management and Budget (OMB) is required to develop and update annually an inventory of federal programs on a publicly available website. In recent years, OMB has made progress developing a complete inventory. However, the inventory does not yet include all federal programs—such as acquisitions, defense, or foreign assistance programs—or provide all required information—such as each program’s contribution to its agency’s mission and goals. Improper payments—those that should not have been made or were made in the incorrect amount—have also been a longstanding and persistent issue for the federal government. For Fiscal Year 2025, 15 federal agencies reported an estimated total of $186 billion in improper payments across 64 programs. However, that estimate does not include certain programs that agencies have determined are susceptible to significant improper payments and does not represent the full extent of government-wide improper payments. Improving the transparency of information on federal programs and spending is foundational for increasing the efficiency and effectiveness of the federal government as well as addressing persistent management challenges, such as preventing fraud and reducing improper payments Improving transparency also provides taxpayers with key information on how their tax dollars are spent. However, to realize this promise, agencies need to continue to take steps to improve the transparency of federal programs.

How does fiscal stimulus affect durable goods sales and to what extent does stimulus drive inflation? This research studies this question in the context of how the unprecedented pandemic fiscal stimulus affected household car purchases and auto prices. Using administrative data on vehicle registrations, the research found that the stimulus increased purchases by 5.5 million vehicles (3.2%) during 2020–2022, implying a medium-run (3-year) marginal propensity to spend on autos of 0.19 and a total marginal propensity to consume of 0.47. Despite this substantial demand response, fiscal transfers account for less than 20% of the surge in auto prices. When transfers push households onto the new-car margin, trade-ins expand used supply and limit price increases. This channel weakens when supply is tight or when policy targets borrowing constraints, in which case stimulus manifests more as inflation than output. Non-fiscal factors, including supply constraints, relaxed credit conditions, and preference shifts, explain the majority of the observed inflation.

The Equitable Homeownership Collaborative (EHC) launched the Inspire100 mortgage pilot to catalyze systems change in the mortgage market, connecting the financial pipes to underserved communities and helping families in these neighborhoods achieve homeownership. Launched in 2024, the Inspire100 mortgage pilot works with lending partners to originate low-cost, sustainable mortgages to low-income and low-wealth borrowers, with an innovative strategy to provide community development financial institutions the liquidity they need to do so. In this report, the research team document the Inspire100 mortgage’s theory of change and logic model, the evidence base for the Inspire100 mortgage, and implementation hurdles and solutions the EHC used to address challenges. Of the 217 originated Inspire100 loans, 80% were originated to first-time homebuyers, 77% went to low- and moderate-income households, and 71% went to households of color. Though the pilot is early in implementation, the serious delinquency rates on Inspire100 loans remain manageable. To improve implementation and to expand the Inspire100 pilot, the EHC can increase its lending capacity by onboarding new community development financial institution partners and engaging the traditional secondary market, find sustainable alternatives to philanthropic subsidies, and commission additional research to evaluate the program’s medium- and long-term outcomes.

Florida Statutes direct OPPAGA to biennially review the Agency for Health Care Administration’s (AHCA) efforts to prevent, detect, deter, and recover funds lost to fraud and abuse in Florida’s Medicaid program. OPPAGA found that while measuring prevention activities is considered the best practice and a more effective method for addressing Medicaid fraud than detection and recovery efforts, AHCA primarily uses detection-based performance measures to evaluate Medicaid program integrity (MPI) and managed care organization (MCO) performance. During Fiscal Years 2022-23 and 2023-24, MPI and MCOs had mixed results in meeting fraud detection and prevention performance targets. For example, MPI did not meet the detection-based performance target for overpayments identified but exceeded prevention-based performance targets for overpayments prevented. Similarly, MCOs have not consistently met the detection-based performance target for fraud referrals.

This report presents maternal mortality for 2024 based on data from the National Vital Statistics System. A maternal death is defined as “the death of a woman while pregnant or within 42 days of termination of pregnancy, irrespective of the duration and the site of the pregnancy, from any cause related to or aggravated by the pregnancy or its management, but not from accidental or incidental causes”. Maternal mortality rates—the number of maternal deaths per 100,000 live births—are shown in this report by age group and race and Hispanic origin. In 2024, 649 women died of maternal causes in the United States, compared with 669 in 2023. The maternal mortality rate for 2024, 17.9 deaths per 100,000 live births, was not significantly lower than the rate of 18.6 in 2023. In 2024, the maternal mortality rate for Black non-Hispanic women was 44.8 deaths per 100,000 live births, which was significantly higher than the rates for White non-Hispanic (14.2), Hispanic (12.1), and Asian non-Hispanic (subsequently, Asian) (18.1) women. The observed decreases in 2024 for Black, White, and Hispanic women and the increase for Asian women were not significant. In 2024, maternal mortality rates were 13.7 deaths per 100,000 live births for women younger than age 25, 16.5 for those ages 25–39, and 62.3 for those age 40 and older. The rate for women age 40 and older was five times higher than the rate for women younger than age 25.

The private-pay cost of long-term services and supports (LTSS) has long been unaffordable for most Americans. Although these services became more affordable during the 2010s with the cost of care rising more slowly than income, over the past five years (2019–2024), the affordability trend has sharply reversed for home and community services, which older people and adults with disabilities overwhelmingly prefer over nursing facility care. This report explores the trends of increasing LTSS costs and decreasing affordability of home care, as well as variation in costs among different types of services and among states. The private pay cost LTSS increased dramatically between 2019 and 2024. Cost inflation was particularly high for the most frequently used home and community services (home health and assisted living), increasing by almost 50% over those five years. The cost of adult day services increased by 33%, compared to a 23% to 25% increase in the cost of nursing home care and a 22% increase in the median income of older households headed by someone 65 or older. LTSS affordability (income divided by cost) improved throughout the 2010s, as the cost of care rose more slowly than the income of older households, but the affordability of home and community services plummeted between 2019 and 2024. There is considerable variation in private pay LTSS costs among states; for every type of service, the cost in the most expensive state is at least twice the cost of the least expensive state.