Crime statistics are mushier than partisans on either side admit

In June, Attorney General Merrick Garland greeted the latest quarterly update of available FBI statistics like a conquering hero: “This data makes clear that last year’s historic decline in violent crime is continuing. In the first three months of this year, violent crime has decreased by over 15 percent compared to same period last year, and murder has decreased by over 26 percent.”

During the presidential debate with Kamala Harris, Trump argued that “the FBI — they were defrauding statements. They didn’t include the worst cities. They didn’t include the cities with the worst crime. It was a fraud.”

Trump was partially correct; there’s no reason to think that the FBI statistics are deliberately fraudulent, but large gaps and inconsistencies in the collection of data mean the numbers aren’t offering a complete picture.

As the Marshall Project explains, in 2021, “in an effort to fully modernize the system, the FBI stopped taking data from the old summary system and only accepted data through the new system. Thousands of police agencies fell through the cracks because they didn’t catch up with the changes on time.” That year, Miami-Dade, New York City and Los Angeles did not submit their data. Philadelphia reported nine months’ worth of data; Chicago reported seven months, and Phoenix reported only one month. For 2021, the FBI noted, “crime estimates will fill in the gaps where data is not available.”

In 2022, Miami-Dade, Philadelphia and Chicago got the data in, but Phoenix didn’t, and neither did New York or Los Angeles. Crime statistics that don’t include information from the two largest U.S. cities would seem to guarantee an incomplete understanding about the state of crime in America.

The stumbling transition to the new FBI reporting system meant big variations in the number of law enforcement agencies that participate year by year. In 2020, 16,572 of 18,641 participated (88 percent). The following year, when the new system was introduced, saw a big drop in participation: just 13,344 of 18,939 (70 percent). The next year, 2022, brought a rebound, with 16,100 of 18,930 participating (85 percent).

But even for 2020, the FBI was missing crime reports from about 2,000 jurisdictions. The most recent complete year in the FBI Crime Data Explorer is 2022, but the numbers in the quarter that Garland cited included data from 13,719 of 19,268 law enforcement agencies in the country — 71 percent.

After reviewing the figures that Garland touted in June and comparing them to publicly available crime data from localities, crime and data analyst Jeff Asher concluded, “My impression is that the trend direction shown in the FBI data through the first quarter is likely correct but that the overall percent changes are almost certainly overstated by a good bit.”

He also observed: “I found places where violent crime was basically spot on compared to publicly available data (like Phoenix, Philadelphia, Houston, Seattle, San Antonio, and Denver), places (like Washington, DC, San Diego and Long Beach) where the FBI’s data is understating declines, and places (like Baltimore, Dallas, and New York City) where things are clearly quite wrong in the FBI data (violent crime in Baltimore and Dallas is down but nowhere near as much as suggested by the FBI’s data, and NYPD data showed a small uptick in violent crime in Q1 2024).”

There’s another complication to keep in mind when discussing crime statistics: The FBI can count only crimes that are reported to the police. Quite a few victims of violent crimes never call the cops. Some victims might believe that there’s no point in reporting the crime to the police because the perpetrators are unlikely to be apprehended and punished. Victims of spousal, relationship or child abuse might not want the perpetrator arrested. The “snitches get stitches” threat and other efforts at intimidation likely ensure that some victims of violent crime never fill out a police report.

And undocumented immigrants might make particularly tempting targets for violent criminals, as people who are in the country illegally “are fearful of admitting that they have been a victim of a crime in part because they believe they will be removed (deported) from the United States if they report the crime,” according to the Department of Homeland Security.

There’s an alternative to measuring crime with police reports: the National Crime Victimization Survey. Each year since 2001, the Justice Department’s Bureau of Justice Statistics has collected a nationally representative sample of about 240,000 people in about 150,000 households.

Survey respondents provide information about themselves — age, sex, race, marital status, education level — and whether they’ve been a victim of a violent crime in the past year. Self-reported information is hardly definitive, but it can be useful to consider. The NCVS collects information about the type of crime, whether the crime was reported to police, reasons the crime was or was not reported, and victim experiences with the criminal justice system.

The NCVS data indicates that in 2023, the rate of nonfatal violent victimization in the United States was 22.5 victimizations per 1,000 people age 12 or older, which was similar to the 2022 rate of 23.5 violent victimizations per 1,000 people age 12 or older. Violent victimization includes rape or sexual assault, robbery, aggravated assault and simple assault.

That is technically a decline from 2022 to 2023, but it’s a decline of one-tenth of a percentage point, so … not much. We should also note property crimes increased slightly, from 101.9 incidents per 1,000 households to 102.2 incidents.

Despite widely cited FBI statistics about a surging homicide rate during the pandemic years of 2020 and 2021, the NCVS data also indicates that violent victimizations overall were actually lower than in the following years — 16.4 per 1,000 people in 2020, and 16.5 per 1,000 people in 2021. That might seem unlikely, given news coverage of that period, but it makes sense: Millions of people were stuck at home and worked from home, putting fewer people on the street to commit a crime or be the victim of one.

The NCVS — which has collected data only through December 2023 — paints a portrait of a country grappling with a distinctly post-pandemic crime rise, as crimes reported to the police slight increased: In 2023, 44.7 percent of violent victimizations were reported to the police, an increase from 41.5 percent in the previous year.

Perhaps most troublingly, “a smaller percentage of robbery victimizations that occurred in 2023 (42 percentage) than in 2022 (64 percentage) were reported to police, and the percentage of motor vehicle thefts reported to police decreased from 81 percentage in 2022 to 72 percentage in 2023,” the Bureau of Justice Statistics reported.

All of which is to say that getting a grip on what’s happening with crime in the United States over the past few years is considerably more difficult than politicians, government officials (and journalists) tend to acknowledge. Has the rate of violent crime declined a little in quite a few corners of America? Probably so. That’s good news, but it might prove illusory if an increasing number of Americans don’t see any point in telling the police that they’ve been the victim of a crime.