The Weaponization of Crime Data Disinformation is Exploding
Is violent crime skyrocketing or plummeting? This question has become a fixture of the 2024 presidential race, with dueling claims from both campaigns.
On October 16th, the headline at Real Clear Investigations, an online media site, touted what it called a bombshell report to support the narrative that crime is skyrocketing. However, the report was written by discredited former academic John Lott. This report was swiftly shared by Fox News, the New York Post, the Washington Examiner, and a host of other conservative outlets. Elon Musk’s PAC to help elect former president Donald Trump even boosted the claims.
In the report, Lott alleges that the FBI quietly revised 2022 violent crime numbers from a 2.1% decrease to a 4.5% increase with no public acknowledgment. He further contends that the FBI’s revision reveals that they underreported 80,029 violent crimes and 1,699 murders in 2022.
However, examining the same FBI data Lott cites provides a dramatically different story than one of violent crime skyrocketing during President Joe Biden’s administration.
In fact, the FBI revision indicates that 31,543 fewer violent crimes occurred during Biden’s time in office than it had previously recorded, along with 449 fewer murders. And this is before addressing the substantial decreases in violent crime and murders that all crime-related data sources, including the NCVS, agree occurred in 2023, and preliminary data sources show continued in 2024.
So how does Lott’s report show an additional 80,029 violent crimes and 1,699 murders in 2022 when the data shows something completely different? By omitting key information combined with statistical malfeasance.
Lott focuses on the FBI’s 2022 revision, which did show an increase in the number of violent crimes and murders for the year, but not to the degree Lott states. Crucially, Lott doesn’t inform readers that the FBI also revised their 2021 numbers downwards, finding fewer violent crimes and murders. This means that while 2022 had slightly more violent crime and murder than originally reported, 2021 had substantially less.
Lott then turns this overall reduction in the amount of reported violent crime and murders on its head. He compares the lower 2021 to the higher 2022 crime numbers while never mentioning that the 2021 numbers were revised downward. He then claims the widening of the gap between these two years are all violent crimes and murders the FBI missed. But the increased difference in crimes between 2021 and 2022 due to revisions does not mean the FBI was missing additional crimes, particularly because there are fewer crimes overall after the revision.
In short, Lott uses the widened gap between 2021 and 2022 to manufacture a surge in violent crime, while failing to tell readers that the overall number of violent crimes and murder during those two years was revised down. This is blatant statistical malpractice in service of a political agenda, and is not the first time Lott has been caught manipulating or even fabricating data.
It is important to note that the FBI is not a perfect data source by any stretch, and it should be used in context with other sources to provide a broader picture. Reporting to the FBI’s system is voluntary for law enforcement agencies, and in 2021 the FBI changed its reporting system, causing even more uncertainty in that year’s numbers. This makes it even more manipulative to build a narrative solely around the difference between 2021 and 2022 crime rates. Further, unlike what Lott alleges in his report, the FBI does regularly revise earlier years of data, so such changes are not unexpected or evidence of partisanship.
Additionally, Lott’s preferred NCVS data source also underwent substantial changes to its data collection efforts due to the pandemic preventing in-person surveys. The last time the NCVS changed its practices, it resulted in a purported crime spike in 2006 that no other source showed, to the point where the NCVS often erases that year from data charts. Using a single year’s upward spike in a survey estimate with a substantial margin of error during a time when the survey was adapting to the pandemic is also statistically irresponsible, especially when no other sources show a similar trend during that short timeframe. The NCVS itself does not present its data in the way Lott does, and instead focuses on longer term trends rather than yearly fluctuations. And just like FBI data, NCVS should not be the only data source to rely on.
Most importantly, the year is 2024, not 2022. Using data that is two years or older to claim that crime is increasing now is deeply dishonest.
With just weeks left before Election Day, disinformation about crime and gun violence has the potential to sway the electoral balance. Approximately 75% of registered voters view crime as an extremely or very important issue, with 74% stating the same about gun policy. Republicans in particular rate crime as their 4th most important issue overall.
Concern about crime is appropriate. Violence in the United States is unacceptably high, and an epidemic of gun violence steals far too many lives every day. However, that justified concern and fear for our safety must be tempered by accurate data. All of the data currently available shows violent crime and murders falling in 2023 (murders down 11.6%), and again in 2024 (murders down between 16% and 23%) at a record rate, driving near or even dipping below pre-pandemic lows.
More work must be done, but that work must not be manipulated by a deliberate disinformation campaign.
Devin Hughes is the founder of GVPedia, a gun violence prevention nonprofit that provides access to research about gun violence and counters disinformation.
_______________________________________________________________________________________
We are grateful and humbled to be able to do this work and to support efforts to bring attention to some of the most serious components of gun violence. If you’d like for us to speak to your group or organization further about this topic or others, please reach out to us.