An expanding body of research highlights the rising importance of class in arrests and incarceration in the United States. This research shouldn’t be read to suggest that ethnoracial disparities have closed or that racial domination is no longer a major driver of hyperincarceration in the United States. However, it should lead to a sharpened focus on the role of class in criminal justice, welfare, and other policy areas.

Here are some key findings from this research:

There have been significant declines in drug arrests among Black individuals but not in other groups, and there are more substantial declines in drug-related imprisonment for Black individuals compared to other racial and ethnic groups. Beckett and Brydolf-Horwitz (2020) document that while racial disparities in drug arrests and incarceration remain considerable, “the main change that has occurred [between 2007 and 2018] is a decline in the Black share of arrests and imprisonments.” They conclude that support for drug policy reform has grown much more in urban areas — including among Black leaders who initially supported the war on drugs, as well as prosecutors and judges — than in white rural areas.

Black men’s lifetime risk of incarceration has decreased. In an update to previous research on the lifetime risk of incarceration, Robey, Massoglia, and Light (2023) highlight that the risk of incarceration for Black men decreased by 44 percent between 1999 and 2019, with declines noted in every state. Through a life table analysis, they estimate that 1 in 5 Black men born in 2001 will be imprisoned, compared to 1 in 3 for the 1981 birth cohort. Although they are careful to mention that these declines could stall or reverse, they also emphasize that rates fell the fastest for young Black men, which may suggest further declines in the future.

Black–White inequality in federal sentencing has narrowed. Light (2021) finds that the average sentencing difference between black and white defendants in US federal courts fell from nearly three years in 2009 to less than six months in 2018. For defendants convicted of drug offenses, the difference fell nearly four years to zero over the same period. He concludes the declines were due to increases in sentences for white defendants, decreases for black defendants, and changes in how prosecutors used mandatory minimum sentences.

There have been considerable declines in racial inequality in the US prison admission rate and increases in class inequality. Compared to the incarceration rate, the prison admission rate — the number of people per 100,000 admitted to prison each year — provides a better measure of recent changes in incarceration. Muller and Roehrkasse (2022) find that black-white inequality in the prison admission rate peaked in 2000 and then steadily declined. Over the same period, using education (college vs. no college) as a proxy for class, they find that class inequality in admission has surged among both blacks and whites. In Muller and Roehrkasse (2025), they update their analysis through 2019, and look at differences for three types of offenses (drug, property, and violence). They find substantial declines in the black prison admissions rates for drug and property crimes but not for violent crimes. Among whites, they find substantial increases in prison admissions rates for all three offense types.

Adverse shocks to local labor markets due to the rise in Chinese exports to the United States led to significant increases in total incarceration rates for both black and white people. Past research has generally not found a positive relationship between unemployment and incarceration. Clegg and Usmani (2024) argue this research may be flawed because of the lack of substate data, inadequate measures of unemployment, and other methodological problems. Expanding on the approach used by Autor, Dorn, and Hanson (2013), they use exposure to the rise in Chinese exports at the commuting zone level as an instrument for local job loss and find that had the 1990s employment gains been preserved in the 2000s, incarceration rates would have been more moderate.

In a piece published in Catalyst, Clegg and Usmani (2019) propose a broader and more contested historical and economic theory regarding the rise of incarceration, one that partially disputes the notion that the carceral system in the United States functions mainly as a race-based social control mechanism. (See also Clegg, Spitz, Wolcke, and Usami (2024)). They argue that the punitive turn in crime policy was driven by a complex set of factors, including:

• Rising Violent Crime Rates: Crime rates, particularly violent crime, rose dramatically in the 1960s, creating a public panic and demand for more punitive measures, including among African Americans. (See also Forman 2012).
• Economic Restructuring and Urban Decline: The decline of American cities, particularly historically Black areas, due to economic shifts, suburbanization, and middle-class flight, contributed to social strain and rising crime rates.
• An Incomplete Social State: The United States spends considerably more on its penal system than other rich capitalist countries while spending less on social programs. Consequently, social problems were substantially addressed through punitive means rather than social policy.

To some extent, Clegg and Usmani are building on the influential work of sociologist Löic Wacquent, who studied under sociologists Pierre Bourdieu and William Julius Wilson and collaborated with both of them on some of their key work in the 1980s and 1990s (see, e.g., Wacquant and Wilson (1989) and Bourdieu and Wacquant (1992)). Wacquant (2010) has argued that hyperincarceration — a term he views as a better descriptor than “mass” incarceration — is due to fear among whites in reaction to urban riots and related racial upheavals in the 1960s, a post-industrial economic transition shifted employment from manufacturing to services, which made African American workers redundant and contributed to the decline of African-American ghettos that began in the 1970s, and the atrophy of the social state and the emergence of punitive welfare “reforms” that attribute joblessness to personal irresponsibility. To this, I’d suggest at least two other potential factors: military downsizing (see Han 2018) and the decline of federal expenditures on employment and training services (see Mullin 2022).

A note about “race” and “racial” categories. Race isn’t an essence or even a thing that exists apart from racism. It’s a social practice and ideology with colonial origins. Powerful people and nations took ethnicities and racialized them for various purposes, including domination and exploitation. (Fields and Fields 2022; Williams 2022). Our understanding of “race” in the United States today, including as something that stands apart from ethnicity/ancestry, is shaped by history and laws but also by current federal statistical practices and elite conventions.

So, for example, one can use the Census Bureau’s categories and data to produce varying estimates of the ethnoracial composition of incarcerated people. Still, regardless of the method used, there is little question that people racialized as Black remain a wildly disproportionate share of the incarcerated compared to their share of the overall population. The reverse is true of people with non-Hispanic/Latino ethnicities who have been racialized as White. (See, for example, Census Table S2603). It’s also worth noting that a substantial and growing number of people identify as belonging to two or more “races” (about 13 percent in the 2023 ACS) and, among people who identify as belonging to only one race, about 8.5 percent say it is “some other race” than the conventional ones listed on the ACS form.

Finally, the research I’ve highlighted generally covers the pre-pandemic period. The crime rate, jail incarceration, and prison incarceration rates are lower today than in 2019. The crime rate fell in 2023 and 2024, and drug arrests have fallen sharply since 2019 (Sawyer and Wagner 2025). The jail incarceration rate — which is mainly made up of unconvicted people — has fallen from 224 per 100,000 people in 2019 to 198 per 100,000 people in 2023. However, the percentage of the jail population classified as non-Hispanic white (about 47 percent) has changed little during this period (Bureau of Justice Statistics 2024). A recent descriptive analysis of state data concludes that incarceration has decreased less since 2019 in the Midwest and the South and that local jail incarceration rates in rural areas are higher today than in 2010 (Kang-Brown and Zhang 2024). Still, much more research will be needed to determine whether the significance of class has continued to increase and whether and how recent political events have changed incarceration trends.