This report is part of a series that began in 1996 and examines the nature and frequency of contact with police as reported by U.S. residents, including demographic characteristics, types of contact, and perceptions of police misconduct, threats of force, or use of nonfatal force. Contact with police includes instances where U.S. residents contacted police, where police approached or stopped residents (police-initiated contacts), and where a traffic accident was involved. This report finds that about 21% (53.8 million) of U.S. residents age 16 or older had contact with police in 2020 and that a smaller share of persons had contact with police in 2020 than in 2018 (24%). In 2020, females (12%) were more likely than males (11%) to initiate contact with police, while males (11%) were more likely than females (9%) to experience police-initiated contact. Additionally, among U.S. residents who initiated their most recent contact with police, almost half (49%) did so to report a possible crime.

This report provides data on full-time sworn officers and civilians employed by local police departments. In 2020 local police departments in the United States employed about 473,000 full-time sworn officers and 126,000 full-time civilian personnel. However, almost half (46%) of all local police departments employed fewer than 10 full-time-equivalent sworn officers. In local police departments serving 250,000 or more residents, about 16% of police chiefs were female. Additionally, about 14% of full-time sworn officers and 11% of first-line supervisors across local police departments were female.

This commemorative guide is intended to provide a better understanding of the history and use of federal education statistics that have been collected and reported by the federal education statistics agency (now the National Center for Education Statistics) since 1868. The statistical profiles in this report use updated historical trend data from 120 Years of American Education: A Statistical Portrait to offer an in-depth look at what each statistic measures, how it has been collected over the years, and what the data reveal about the statistic. Statistics covered in the report include enrollment in elementary and secondary schools; high school graduates and graduation rates, and postsecondary student costs and finances. Notably, over the past century and a half, the number of schools has been regularly reported and broken out by an increasing number of different types and categories of schools to provide an ever more comprehensive picture of the nation’s school systems. The overall number of public elementary schools declined during the 20th century through the 1970s, despite periods of marked increases in the school-age population during this period. The vast majority of public elementary schools that closed during the 20th century were schools with one teacher (sometimes called one-room schoolhouses). The number of such schools declined steadily throughout the 20th century, from about 150,000 in 1929 to fewer than 1,000 by the 1980s, and there are nearly 200 of these schools today. However, since the late 1980s, increases in the overall number of public elementary schools have largely followed increases in the school-age population.

This research examines how the racial wealth gap interacts with financial aid in American higher education to generate a disparate impact on college access and outcomes. Retirement savings and home equity are excluded from the formula used to estimate the amount a family can afford to pay. All else equal, omitting those assets mechanically increases the financial aid available to families that hold them. White families are more likely to own those assets and in larger amounts. The authors document this issue and explore its relationship with observed differences in college attendance, types of institutions attended, degrees attained, and education debt using data from the Survey of Consumer Finances (SCF), the National Postsecondary Student Aid Study (NPSAS), and the Panel Study of Income Dynamics (PSID). The results show that this treatment of assets provides an implicit subsidy worth thousands of dollars annually to students from families with above-median incomes. White students receive larger subsidies relative to Black students and Hispanic students with similar family incomes, and this gap in subsidies is associated with disadvantages in educational advancement and student loan levels. This result may explain 10% to 15% of white students’ advantage in these outcomes relative to Black students and Hispanic students.

This study explores whether the Trade Adjustment Assistance Community College and Career Training (TAACCCT) Program, the largest federal investment in community colleges in this nation’s history, expanded and improved vocational training programs. The authors find that, on average, the completion of credentials in career-technical fields increased at institutions receiving a TAACCCT in the first wave of the program, compared with other public, 2-year colleges. In particular, credentials in business, health care, and information technology (IT)-related fields increased, and the growth is concentrated in certificates. These findings support previous literature examining the relationship between college funding and student outcomes, and suggest that additional funding enables public 2-year colleges to expand and improve technical education programs, despite some of the unique challenges facing these programs.

Before the 2007-2009 Great Recession, homeownership among households headed by adults ages 25-34 spiked as risky lending practices greatly expanded access to mortgage credit. Soon after, the nation experienced a housing market crash and economic downturn. Since then, lending options tightened and home prices soared. This limited young householders’ ability to own a home, driving down ownership rates which had yet to return to pre-recession levels by 2019. Homeownership among households headed by adults ages 25 to 34 was highest from 2003 to 2007. Rates for those years, just prior to the Great Recession, ranged from 46.4% to 47.0%, according to data from the 2000-2019 American Community Survey 1-year estimates. By 2010, right after the recession ended, homeownership among young heads of household had declined to 41.3% and continued to decline to 36.8% in 2015 and has not meaningfully increased through 2019. The education gap in homeownership – the difference between householders with less than a high school education and those with a bachelor’s degree or higher – still exists but narrowed after the recession. The education gap in young household homeownership was 16.6 percentage points in 2000. It increased before and during the recession, plateauing at 28.9 percentage points in 2009. It declined to 21.0 percentage points in 2019.

Approximately 27 million passengers with disabilities traveled by air in 2019, according to the U.S. Department of Transportation (DOT). Without accommodations, such as appropriate assistance and communication, passengers with disabilities may face challenges when flying. The U.S. Government Accountability Office (GAO) testified about these challenges and actions that airports, airlines, and the DOT are taking to address them. Past GAO work has highlighted a range of barriers to accessible air travel that passengers with disabilities face. This report discusses: (1) barriers to accessible air travel; (2) steps that airports and airlines have taken to reduce those barriers; and, (3) the status of actions taken by DOT to respond to accessibility-related provisions in the Federal Aviation Administration (FAA) Reauthorization Act of 2018 and enforce accessibility-related regulations. Based on GAO's review of regulations and information obtained from officials with the DOT and FAA, GAO found that the DOT has taken steps to implement the relevant accessibility-related provisions of the act.

Machine learning (ML) can have a significant impact on public policy by modeling complex relationships and augmenting human decision making. However, overconfidence in results and incorrectly interpreted algorithms can lead to peril, such as the perpetuation of structural inequities. The authors give an overview of ML, algorithms that can automatically detect patterns in data and use this information to predict future data or other outcomes, and discuss the importance of its interpretability. The authors also describe methods that aid interpretation and the characteristics of effective explanations. In addition, they offer the following recommendations, which will help policymakers develop trustworthy, transparent, and accountable information that leads to more-objective and more-equitable policy decisions: (1) improve data through coordinated investments; (2) approach ML expecting interpretability, and be critical when assessing data quality, modeling, assumptions, and other contextual factors that lead to the results; and (3) leverage interpretable ML to understand policy values and predict policy impacts.

This is a report on trends from 2000 through 2017 in death rates among adults in the United States described increasing rates starting around 2012 for adults aged 25–44 in the three largest race and ethnicity groups. Most of the adults in this age range in 2020 are part of the millennial generation (born from 1981 to 1996) and have recently become the largest generation group in the United States. This report presents trends for the three leading causes of death among people aged 25–44 in 2020 (unintentional injuries, heart disease, and suicide) for non-Hispanic Black, non-Hispanic White, and Hispanic people. Key findings from the report include that death rates for non-Hispanic White, non-Hispanic Black, and Hispanic people aged 25–44 increased from 2013 to 2019, followed by large 1-year increases from 2019 through 2020. Death rates for unintentional injuries began to increase before 2019 for the three largest race and ethnicity groups but experienced the greatest annual increases from 2019 to 2020, ranging from 23% to 47%. People aged 25–44 in the three largest race and ethnicity groups experienced recent increases in death rates due to heart disease, with the largest increases occurring between 2019 and 2020. Suicide rates did not change significantly from 2017 (23.3 deaths per 100,000 people) to 2020 (22.7) for non-Hispanic White people aged 25–44, but continued to increase through 2020 for non-Hispanic Black (12.7) and Hispanic (11.3) people.

This report describes changes between 2020 and 2021 in the percentage of home births by month, race and Hispanic origin, and state of residence of the mother, and makes comparisons with changes occurring between 2019 and 2020. The percentage of U.S. home births rose from 1.26% (45,646) in 2020 to 1.41% (51,642) in 2021, an increase of 12% and the highest level since at least 1990. Increases ranging from 10% to 21% were seen for the three largest race and Hispanic-origin groups. The percentage of home births for all women increased between 2020 and 2021 for most months, peaking in January 2021 at 1.51%. Patterns by month differed somewhat by race and Hispanic origin, with more consistent monthly increases seen for non-Hispanic White women. Home births increased in 30 states, including Florida, (with non-significant increases for 11 additional states) and declined in 2 states (with non-significant declines for 7 additional states and D.C.). The 12% increase in home births from 2020 to 2021 follows a 22% increase from 2019 to 2020, with increases by maternal race and Hispanic origin ranging from 21% to 36%.

In recent years, the number of publicly available tools and indices assessing social determinants of health (SDOH) has grown exponentially. While many of these indices have been developed to assist researchers and practitioners with identifying vulnerable communities, it is difficult to determine the most appropriate measure, index, or combination of indices to use given the research question of interest. This paper presents an overview of the most commonly included indices, highlights commonalities, and identifies some differences in what they measure. The authors also discuss challenges with using these measures, including the use of state level data to examine local level issues and how the use of atheoretical indices challenges the application of SDOH measurement. The authors found that publicly available social determinants of health (SDOH) indices are not designed to be universal measures or compendia of measures that can be readily entered into statistical models. Additionally, these indices draw from multiple data sources, such as the American Community Survey and Behavioral Risk Factor Surveillance System, and use different measures to conceptualize SDOH domains, such as economic stability, education access and quality, health care access and quality, neighborhood and built environment, and social and community context. The authors also note that it is important to consider that many of the data available today predate the pandemic and may not capture the current state of key SDOH factors. These findings are intended to provide researchers and practitioners with information about SDOH data available through these common indices to inform how they are applied based on the needs of their work.