Common myths about the new FBI crime statistics, debunked
This week, the FBI released its report on crime in the United States during 2023. The report aggregates data from over 16,000 law enforcement agencies that collectively cover over 94% of U.S. residents. It found a large decline in murders compared to 2022 (-11.3%), and moderate but still significant declines in violent crime overall (-3.0%) and property crime (-2.4%).
Former President Donald Trump and his Republican allies, however, insist that America is in the midst of an unprecedented "crime wave." They argue that the FBI data collected through the Uniform Crime Reporting Program is incomplete or fraudulent. In a September 12 post on X, Trump cited the Department of Justice's National Crime Victimization Survey to claim that there has been a 40% increase in violent crime since he left office. "Kamala Crime is destroying America, and gangs are taking over!” Trump insisted.
To sort through the confusion, Popular Information spoke with Jeff Asher, co-founder of AH Datalytics, creator of the Real-Time Crime Index, and one of the nation's leading experts in crime statistics.
This interview has been edited for length and clarity.
Popular Information: Can you explain what, in your view, are the most important takeaways from the recent FBI report on crime in the United States during 2023?
Jeff Asher: The murder trend—the decline—is the largest we've ever seen, based on the data we have, which goes back to 1960. While it’s not the full history of the nation, it’s the largest percentage change recorded over that period, in terms of fewer victims from one year to the next. That’s significant because murder is the most serious offense, with the highest societal cost, and it's the one we have the most confidence in. Everything else tends to be underreported to some degree, but we believe the murder count is fairly accurate.
For violent crime, a 3% decline met expectations. It’s nothing unusual and aligns with what the Bureau of Justice Statistics (BJS) baseline shows. What’s more important than looking at a one-year change is to compare where we are now to where we were five, ten, or 35 years ago. In 2023, the violent crime rate was virtually equal to 2014’s level, which was the lowest since 1970. The country has a lower level of violent crime than it had 35 years ago, and while the numbers fluctuate, there’s no dramatic increase or decrease. A 3% decline is good, but it’s more "steady as she goes" than anything drastic.
Looking ahead to 2024, the data so far shows an even larger decline in murder. By the end of the year, the murder rate could be at or below the 2019 level, heading closer to the historic lows of 2013-2014, rather than the spike we saw in 2020.
PI: Trump is using the National Crime Victimization Survey (NCVS) to argue that violent crime has increased dramatically since he left office in 2020. Is this a valid use of the data?
JA: No, I would strongly caution against using percent changes with NCVS. First, murder is what spiked in 2020, and NCVS doesn’t measure murder because it’s a survey, and murder victims can’t respond. So you’re already excluding the most notable crime trend of the last four years. Second, since NCVS is a survey, there’s a margin of error that gets ignored when you focus on percent changes. Violent crime might have been up between 2020 and 2023, but it could also have been up just a little—or not at all—depending on how you interpret the confidence intervals. When you compare 2023 with 2019, BJS found that violent crime wasn’t statistically different. So focusing solely on 2020 can give a distorted view of the trend.
If you dig deeper into specific offenses, like rape, NCVS might show a 42% increase, but the lower and upper bounds overlap. So it might have been up 42%, or it might have been down. This is how margins of error work. It’s like polling: if a poll shows Trump up five points, but the margin of error is 4.4%, he could be up 14 points or down nine. Ignoring the margin of error leads to misleading conclusions.
PI: Another argument being made by Trump, and others online, is that the FBI data isn’t accurate because it doesn’t include all the cities. Is that true?
JA: The data is incomplete in the sense that it only covers 94.3% of the population, but it’s always been incomplete. That’s not unique to 2023. Historically, 5-6% of the population isn’t covered, and the FBI estimates that portion. But it’s mostly smaller cities that don’t report, and these cities tend to have less crime. Only two out of 90 cities with populations over 250,000 didn’t submit data in 2023. So the FBI is only estimating a small portion of the crime, and those estimates are consistent with what we see from other sources like the CDC and the Gun Violence Archive, particularly for murder and gun violence.
PI: Trump has called the FBI data "fraudulent," suggesting it’s being manipulated by the Biden administration. Is that possible?
JA: Having worked in the federal government, I don’t think a secret like that could be kept. We also have independent sources. We recently launched the Real-Time Crime Index, which gathers data from over 300 agencies, and it shows nearly identical trends to the FBI’s data. Other external sources, like the Gun Violence Archive and the CDC, also align with the FBI’s findings. If the FBI had reported something like a 4% increase in murder last year, I would have questioned it. But the data was exactly what we expected, based on other sources. That consistency gives us confidence in the accuracy of the FBI’s numbers.
PI: Some argue that the data doesn’t mean much because people have stopped reporting crimes, believing they won’t be prosecuted. Is that a valid concern?
JA: I doubt many prosecutors are refusing to prosecute murder. Also, we have external sources like the CDC and the Gun Violence Archive that back up the trends in the FBI data. NCVS actually measures whether people are reporting crimes, and over the last five years, there hasn’t been a significant change in the rate at which people say they report crimes to the police.
PI: Is there any evidence that undocumented immigrants are fueling a crime wave?
JA: I don’t see any evidence of that. And if it’s happening, it’s within the context of an overall decline in reported crime. It’s also not showing up in the places you would expect to see it if there were a surge in crime driven by undocumented immigrants.
The GOP is without shame and has sunk to a level that I don’t think they can come back from. We do need to have alternate political views to keep our politics from becoming one sided. The problem is that the GOP is o longer an actual political party. Even if they loose this election it doesn’t seem that they will purge themselves of this malign and incompetent and backward thinking
The reality based community may finally be winning. In part because real journalists are working overtime to spread the truth and in part because the right is flailing in their efforts to create their own reality. They can no longer create a narrative out of whole cloth and get it amplified and perpetuated by a willing and complicit corporate media. The alternative realities of the republican right have gotten so insane that corp media can no longer aid and abet, not because they no longer want to but because they can’t. The lies are just so transparent, so preposterous, so far fetched that corp media can’t both sides them or reshape them to make them seem reasonable. At least to sane, sensible people who are willing to listen.