William J. Sabol and Thaddeus L. Johnson of the Council on Criminal Justice have released a new report breaking down the black–white gap in incarceration since 2000—focusing on state prisons, which house the vast majority of prisoners in the U.S., and excluding Hispanics from both the black and white populations.
They find that the disparity has markedly decreased, a little-noticed trend. In 2000, fewer than 300 of every 100,000 white adults were in state prisons. For blacks, the number was about eight times higher—nearly 2,500. Over the next two decades, the black rate fell by nearly half, while the white rate first edged up and then dipped somewhat, bringing the disparity under five-to-one in 2020.
But disparities will persist, they report, unless at least one of three things changes. One, the racial gap in violent offending could decline; two, since black offenders tend to have longer criminal histories than whites, states could base incarceration less on criminal history; and three, states could otherwise reduce the difference in “time served” between prisoners from the two racial groups.
Where did the decline come from? The report breaks the numbers down numerous ways, though key aspects of the picture remain highly unclear.
State prisons have never been packed full of drug offenders—as of 2000, less than 30 percent of imprisoned blacks and 20 percent of imprisoned whites were there for drugs—but the shifting pattern is most dramatic in that category. The per capita disparity fell from 15-to-one to less than four-to-one, as the black imprisonment rate for drugs fell by more than two-thirds. (Drugs certainly haven’t become less serious of a problem in the past 20 years, but the U.S. has been incarcerating far fewer drug offenders, the racial balance of the issue has evolved as focus shifted to opioids, marijuana has been legalized in many states, and activists have had success in labeling drug enforcement racist.) The disparities for violent, property, and public-order offenses also fell, but much less dramatically.
Prison admissions, not time served, drove the changes. In fact, the disparity in time served increased slightly, though the authors do not calculate whether a shifting composition of offense severity, differences in criminal histories, or something else was the culprit. Throughout the report, owing to data limitations, they can’t break crimes down further than the major categories of violent, property, public order, and drugs.
But the authors do address the elephant in the room: racial differences in rates of offending. This is hard to measure for many types of offenses. Estimates of drug use typically rely on individuals to confess their illegal behavior in surveys, for example, and no one can record the race of a burglar who isn’t caught. But for nonfatal crimes of interpersonal violence such as robbery and assault, surveys exist of the general public in which victims report the race of the person who attacked them. Analysts can then compare those numbers with arrest and incarceration statistics.
Analyses have shown over the years that the racial breakdown of offenders in these surveys roughly matches the racial breakdown of arrests. (See, for example, the late James Q. Wilson here and a more recent Bureau of Justice Statistics report here.) The new report confirms a violent offending gap of about three-to-one to four-to-one that is, if anything, rising over time. And, comparing arrests with survey data on violent crimes that victims say they reported to police, the authors find that a similar proportion of white and black offenses seem to result in arrests these days, though one of their two arrest measures does suggest a disparity back in the early 2000s.
Beyond the arrest stage, differences in offending rates by themselves now explain more than half of the white–black incarceration gap for violent crimes. The remaining unexplained gap is not necessarily bias, but it also includes differences in criminal histories among whites and blacks arrested for the same offense, as well as differences in offense severity, both of which can result in longer sentences. As the report notes, for example, murderers make up about 30 percent of violent offenders kept in state prisons—and black “individuals were convicted in more than half (56 percent) of homicides in cases with a known assailant.”
As of 2019, almost two-thirds of black state prisoners and half of white state prisoners were in for violence, with fewer than one-fifth of each group doing time for drugs. Disparities in parole violations are apparently a thing of the past, too. By the end of the period, whites “faced a somewhat higher rate of prison admission on a technical violation than did” blacks. All this suggests future incarceration and disparity reductions might be harder to come by without jeopardizing public safety. Of the three possibilities noted toward the beginning of this article, a decline in actual offending would clearly be something to celebrate, but the wisdom of reducing incarceration by other means depends on the details.
Other variables are also worth considering. For example, the authors note that the black adult population grew by 38 percent over the 20-year period, versus 8 percent for whites. It would be worth breaking down how age structure and immigrant status shifted as well to see if they could be contributing to or slowing the decline in black incarceration. (Immigrants tend to have low incarceration rates; young adults, high ones.) From my quick analysis of some census data—which differs a bit from the population estimates the report employs but has the advantage of including birthplace information—about one-fifth of the total black adult population increase took place among those born outside the 50 states and D.C. (versus just 5 percent of the smaller increase for whites). And the percentage of black adults under 40 fell from 47 to 43 between 2001 and 2020, with a similar drop among the older-on-average white population, from 37 to 33.
The black incarceration rate has fallen and the disparity between whites and blacks has shrunk. Precisely why that happened, and whether it might continue, remain open questions.