Donald Trump wrong that crime stats exclude 30% of cities, including the “biggest and most violent"

Former President Donald Trump has long focused on high crime rates as a campaign issue, but this position has been complicated by news of declining violent crime, including murders. So, Trump and his allies have tried a new approach — calling into question whether the FBI numbers are incomplete and, thus, wrong.

In a June 15 speech at "The People's Convention" in Detroit, Trump said, "Joe Biden is out the other day trying to claim that crime is down. But actually, crime is way up. They don’t report crime properly anymore, as you probably have heard. Crooked Joe is citing statistics that no longer include data from 30% of the country, including the biggest and most violent cities."

Asked for evidence of Trump’s missing 30% claim, Republican National Committee spokesperson Anna Kelly did not directly address the statistic, saying instead that "in most cities, homicide rates remain higher than pre-pandemic levels" and that "63% of Americans believe that crime is a serious issue."

Trump’s comments follow the FBI’s June 10 release of preliminary data showing that during the first three months of 2024, violent crime fell 15.2% compared with the same period in 2023. Within that category, murders declined 26.4%, reported rapes decreased 25.7%, aggravated assaults declined by 12.5% and robberies fell 17.8%. 

The report followed other data releases showing crime’s downward trend. An independent analysis by crime data analysis company AH Datalytics projected that murders will have dropped by 10% from 2022 — the last year for which full FBI data is available — to 2023. The company projected an additional 18% decline by the end of 2024, based on current trends.

Trump’s point echoed claims by allies on social media that FBI crime statistics are meaningless because many police departments, including those in New York and Los Angeles, did not submit data, leaving their crimes uncounted.

However, experts in crime statistics say that such claims are wrong on several levels.

Although the FBI did finalize a methodological change in 2021 that left that year’s data with an unusually wide margin of uncertainty, those problems were fixed beginning in 2022.

Also, the FBI estimates crime patterns for any missing cities using data from cities of similar size — and it did that in 2021 as well; no city was truly "missing" from the data. Finally, the general patterns for the FBI’s 2021 figures align with data that has been collected independently of the FBI.  

This argument is a "convenient boogeyman, but it's not an accurate one," said Jeff Asher, AH Datalytics’ co-founder.

What happened in 2021?

The FBI changed the way police departments report crime, requiring that they shift to the new National Incident-Based Reporting System for 2021 data collection.

When the 2021 data-collection process was complete, the information submitted to the FBI reflected data covering about 65% of the nation’s population. This meant that about 35% of the population was not covered that year, including such major cities as New York City and Los Angeles. 

To make up for the absence of that 35%, the FBI followed its standard procedure: It replaced the missing data with estimates, using data from comparable cities. This estimation made the 2021 data as complete as it could be under the circumstances, but experts caution that was not an ideal fix.

In a typical year, the FBI collects from 90% to 99% of the data it needs, Asher said. So, most of the time, estimating the missing data "isn't a big deal. But with the coverage suddenly dropping to 65%," he said, "it mattered in 2021."

Featured Fact-check

Barbara Ann Radnofsky

stated on October 22, 2009 in in a Web site video

"The attorney general requires that rape victims pay for the rape kit."

For 2021, the range of uncertainty — essentially the margin of error — for the murder rate was -7% to +17%, and for violent crime overall, it was between -12% and +11%, Asher said. That compares with close to zero uncertainty for a typical year other than 2021.

As a result, Asher and other crime statistics experts caution against reading too much into the 2021 data. 

"It’s true that the 2021 numbers were particularly underwhelming because of the low response rate," said James Alan Fox, a Northeastern University criminologist.

What’s happened since 2021

The issue of data being compromised because of low response rates as it was in 2021 hasn’t occurred since.

For the 2022 data — the last full year of FBI data released — the agency allowed police departments to report either using the new or old system for filing data. That got coverage rates back up to about 94% of the nation’s population, in line with all recent years other than 2021. And the indications for 2023 and 2024 look like the coverage rates for those years will be similarly high as well, experts said.

So, the problems that hampered the FBI data in 2021 are a one-time issue that hasn’t been repeated.

New York and Los Angeles were among cities that submitted data in 2022, he said. For 2023, New York submitted data using the new National Incident-Based Reporting System and Los Angeles is expected to submit through the old system as it continues to transition its processes.

Even if New York hadn’t been reflected in FBI data, Trump would be wrong to lump it in with the nation’s "most violent cities," said Ernesto Lopez, a research specialist with the Council on Criminal Justice. New York "tends to have a lower homicide rate compared to other major cities, so that is not likely an issue," Lopez said.

The FBI reports "have always been incomplete" to some degree, "but that does not mean they are misleading as to trends," added Candace McCoy, a professor emerita at John Jay College of the City University of New York. The FBI report, she said, remains "the best measure available of crime nationwide."

Also, the FBI data’s trend lines have generally tracked the patterns in data compiled by other groups — groups that were not affected by the FBI’s methodological switch. "There are lots of data sources telling the same story about what is happening with murder and violent crime," Asher wrote. These include Asher’s own murder dashboard and violent crime calculations by the Major Cities Chiefs Association and the Council on Criminal Justice. All use data collected directly by each group from a range of police departments.

Our ruling

Trump said that crime statistics "no longer include data from 30% of the country including the biggest and most violent cities."

This was never quite accurate, but it’s been wrong since 2022 and going forward.

In 2021, the FBI required police departments to report data using a new system. That year, data for about 65% of the U.S. population was covered in the FBI’s annual report, rather than the typical 95%. This technically omitted about 35% of the U.S. population. However, as is standard FBI policy, the missing data was estimated using comparable jurisdictions.

More to the point, the FBI’s 2022 data coverage returned to 94% of the population, which is in line with other years excluding 2021, and preliminary numbers for 2023 and 2024 indicate that the coverage level should continue.

And, the FBI data’s trend lines have generally tracked the patterns in data compiled by other groups — groups that were not affected by the FBI’s methodological switch.

We rate the statement False.