The news media relies almost exclusively on FBI data to report on changes in crime rates. But there is strong evidence that FBI data are less reliable than in the past.
The U.S. uses two different measures of crime. The FBI Uniform Crime Reporting program counts the number of crimes reported to police annually. The Bureau of Justice Statistics, by contrast, uses its annual National Crime Victimization Survey (NCVS) to ask 240,000 people a year whether they have been victims of a crime. The NCVS is used to estimate total crime (reported and unreported). The survey indicates that only 42% of violent crimes and 32% of property crimes were reported in 2022.
I have pointed out before that since 2020, the FBI’s number of reported crimes and the NCVS’s number of total crimes have gone in opposite directions. For example, between 2021 and 2022, the FBI showed a 2.1% drop in violent crime, but the NCVS showed an increase of 29.3%.
But there is a more fundamental problem. The FBI and NCVS’s estimates of reported crimes have also gone in opposite directions since 2020. From 2008 to 2019, the FBI and NCVS measures of reported violent crimes generally tended to move up and down together. But from 2020 to 2022, these two numbers were almost perfectly negatively related to each other. Each time one measure of reported violent crimes rose, the other measure fell.
For example, while the FBI’s number of reported violent crimes fell by 2% in 2021 and 2.1% in 2022, the NCVS’s measure showed increases of 13.6% and 29.3%, respectively.
The fact that measures of reported and total crimes don’t match isn’t that puzzling. But when even these two measures of the same thing – reported crime – are going in opposite directions it raises concerns with the FBI data.
A frequently discussed concern with the FBI data is the decline in police departments reporting crime, after a new reporting system was used. In 2022, 31% of police departments nationwide, including Los Angeles and New York, didn’t report crime data to the FBI. That is better than 2021 but still much worse than the 97% of agencies covering most of the U.S. reported in 2020. In addition, in cities from Baltimore to Nashville, the FBI is undercounting crimes those jurisdictions reported.
But there are other issues. Police departments downgrading crimes may also explain the drop in the FBI numbers. Classifying an aggravated assault as a simple assault means that it will be excluded from FBI violent crime data, which doesn’t include simple assaults. The difference often involves whether the criminal used a weapon in committing an assault, but many radically left-leaning DAs are refusing to include weapons charges against defendants. That could explain the difference between the two measures of reported crime because the NCVS will ask victims if the assault involved a weapon, even if the police reports ignore that characteristic of the crime.
Soros-backed District Attorneys nationwide, from New York to Chicago to Los Angeles, are downgrading felonies to misdemeanors. New numbers show progressive Manhattan DA Alvin Bragg downgraded felonies to lesser charges 60% of the time, while 89% of the time they were downgraded to misdemeanors. This isn’t a new problem. In the past, Chicago has intentionally misclassified murders, instead labeling them as subject to non-criminal “death investigations.” But the problem may be increasing and police may also be responding to the decisions by prosecutors.
Over the last few years, as the number of police have been reduced to cuts in budgets and a slew of retirements, police departments nationwide from Charlottesville and Henrico County, Virginia, to Chicago, Illinois, to Olympia, Washington, stopped responding to non-emergency 911 calls. Instead of police coming out, people can still go down to the police station. There is the possibility that people think that calling up 911 reports a crime, but a crime isn’t officially counted until police make out a report.
Initial estimates cited by some news organizations show murder rates dropping 13% between 2022 and 2023. Murders usually are reported, though as cities such as Chicago have demonstrated, even these numbers can be played with, but they are considered more reliable than for other violent crimes. Yet last year’s projected murder rate was still 5.51 per 100,000 people, or 7% above its 2019 level. The NCVS doesn’t measure murders, which make up about 1% of violent crimes.
The news media relies on reported crime numbers without considering unreported crime. But the gap between the two measures of reported crimes provides strong doubts about the accuracy of the FBI’s reported crime data. Americans believe that crime is increasing as law enforcement is collapsing. They also say that they are reporting more crimes to the police, but that isn't showing up in the FBI reports.
Mr. Lott is president of the Crime Prevention Research Center. He served as senior adviser for research and statistics in the Office of Justice Programs and the Office of Legal Policy at the Justice Department.
Join the conversation as a VIP Member