Crime rates are falling, the mainstream press reports. Despite what it calls the “widespread national perception that law-breaking and violence are on the rise,” NBC News maintains that the opposite is true: “Americans believe crime rates are worsening, but they are mistaken.” A Washington Post column by Philip Bump echoes that assertion and opines, “This is not the narrative that has dominated on the right.” Bump adds, “Fox News coverage has consistently focused on crime in urban areas,” in part because the network’s “audience (understandably) associates cities with Democratic leadership, because cities are more heavily non-White.”

The best available statistics, however, indicate that crime has risen dramatically in America’s urban areas, and little reliable evidence suggests that this crime wave has started to recede. NBC, the Washington Post, and others are resting their claims on preliminary FBI data. As I have explained in highlighting the widening crime gap between California and Florida on their respective governors’ watches, the FBI’s statistics aren’t wholly reliable. This is even truer of the FBI’s preliminary statistics.

In 2021, the FBI stopped allowing “summary” reporting from states and localities and started mandating “incident-based” reporting, which provides more detail. Many law-enforcement agencies, however, didn’t make that switch on the FBI’s timeline, leaving the bureau with very incomplete national data. For example, in 2022, the most recent full year for which the FBI has released statistics, 17 percent of law-enforcement agencies didn’t report to the bureau. Among those not reporting were the New York, Los Angeles, and San Francisco Police Departments.

Preliminary FBI statistics are even less reliable. They are not final, they don’t cover the whole year, and 26 percent of agencies didn’t report. That amounts to more than 5,000 agencies not providing data, including some of the nation’s largest. Yet based on these preliminary and partial-year numbers, with more than a quarter of agencies missing, the mainstream media tout an 8 percent drop in violent crime from 2022 to 2023. That’s far from a definitive number.

Even if the FBI’s statistics were based on far more thorough reporting from police departments and sheriff’s offices around the country, many crimes aren’t reported to law enforcement. According to the 2022 National Crime Victimization Survey (NCVS), 58.5 percent of violent crimes—and 68.2 percent of property crimes—were not reported to police. That’s according to the victims of those crimes.

The NCVS, which dates to the Nixon administration, is the nation’s largest crime survey and one of the largest federal surveys on any topic. It’s designed to register crimes that people, for whatever reason, chose not to report to police—as well as to unearth more information about any crimes that are committed. The NCVS is run by the Bureau of Justice Statistics (BJS), which I directed from 2017 to 2021.

According to the most recent NCVS statistics, the number of violent victimizations rose from 5.8 million in 2019 to 6.6 million in 2022—an increase of 14 percent. Excluding simple assault—the least serious of the violent crime categories (and the one least likely to be prosecuted as a felony)—the rise in violent victimizations was 37 percent. After adjusting for population growth, the increases were 12 percent (overall) and 34 percent (excluding simple assault). Meantime, the number of unique victims of violent crime rose from 3.1 million in 2019 to 3.5 million in 2022—an increase of 15 percent. Excluding simple assault, the number of unique victims went up 26 percent.

The rise in crime from 2020 (or 2021) to 2022 was even starker than from 2019 to 2022. The NCVS, however, was not carried out in its usual fashion in 2020 because of the Census Bureau’s decision, perhaps forced upon it from higher up in the administration, not to do door-to-door surveys during much of that year. As with the post-2020 FBI figures, then, one must take statistics from the NCVS in 2020 (and in 2021, as the effects of 2020 carried over) with more than a grain of salt. Comparisons between the 2019 and 2022 versions of the NCVS are much more reliable.

One of the most striking things in comparing crime rates from 2019 and 2022 is seeing where the increase in violent crimes occurred—and here the American people’s common sense appears to be spot-on. The NCVS classifies whether crimes happened in urban, suburban, or rural areas. By far the largest crime spikes were in urban areas.

In suburban areas, the violent crime rate went from 22.3 violent victimizations per 1,000 residents ages 12 and over in 2019 to 23.9 per 1,000 in 2022, an observed 7 percent increase that was apparently not statistically significant at the 95 percent confidence level. In rural areas, the rate went from 16.3 to 15.4 per 1,000, an observed 6 percent decrease that was also not statistically significant. In urban areas, however, the rate per 1,000 rose from 21.1 in 2019 to 33.4 in 2022—a whopping 58 percent increase that was quite statistically significant.

The property crime rate also stayed relatively constant in suburban and rural areas while rising 15 percent in urban areas—going from 153.0 to 176.1 victimizations per 1,000 households (statistically significant at least at the 90 percent confidence level).

It used to be that the NCVS’s urban, suburban, and rural crime statistics didn’t reveal much. They were based on a classification system that BJS had inherited decades earlier from the Census Bureau, and those classifications often left much to be desired. For example, Union City, New Jersey, located just across the Hudson River from Midtown Manhattan, was classified as “suburban,” despite its having a population density (51,918 people per square mile) almost three times that of San Francisco (17,180). At the same time, Yuma, Arizona, with a population density less than 2 percent as high as Union City’s, was classified as “urban.” When I was BJS director, we improved the NCVS’s classifications—you can read about that undertaking, as well as about which places in the U.S. are most urban—and now the urban, suburban, and rural statistics are far more meaningful.

In short, violent crime rates rose dramatically in urban areas from 2019 to 2022, and we have no clear indication yet whether they have risen or fallen since. We’ll know somewhat more when the official 2023 FBI figures are released early next fall, and far more when the 2023 NCVS comes out around that same time. If crime in the cities has indeed dropped, it seems a safe bet that it hasn’t fallen back anywhere close to pre-Covid levels, at least when taking into account crimes not reported to police (but captured in the NCVS). The media may want to believe that our cities have become safe again overnight, but the most reliable statistics, as well as Americans’ daily experiences, say otherwise.

Photo by Spencer Platt/Getty Images


City Journal is a publication of the Manhattan Institute for Policy Research (MI), a leading free-market think tank. Are you interested in supporting the magazine? As a 501(c)(3) nonprofit, donations in support of MI and City Journal are fully tax-deductible as provided by law (EIN #13-2912529).

Further Reading

Up Next