Lagging US crime data fuels misperceptions and thwarts prevention, study finds

When the FBI next releases nationwide crime data this fall, the numbers will be up to 18 months old.

That’s a problem for policymakers and for the American public, according to a report issued Tuesday from the Council on Criminal Justice that calls on police and federal agencies to provide more information, faster.

More recent crime data currently comes from outside groups that gather it piecemeal from local police websites, which the reports’ authors and crime analysts in interviews said impedes both law enforcement’s responsiveness and the public’s true understanding of crime in the moment.

“As a democratic society,” said John Roman, director of the Center on Public Safety and Justice at the University of Chicago, “we need to know what the facts are, so elected leaders can solve the problem as it exists, not as it’s messaged.” Homicides in America have plunged recently to levels not seen since the 1960s, said Roman, who chairs the working group that issued the report, but “the public reaction is, ‘How are we going to fight this violence epidemic?’, when there is no epidemic.”

The council’s report urges the federal Bureau of Justice Statistics to begin publishing monthly crime statistics from a nationally representative sample of cities and counties, similar to how the Bureau of Labor Statistics gathers and releases employment data every month. It also recommends that the bureau begin including crimes typically excluded from conventional crime data, such as cybercrime, environmental crime and white-collar fraud.

The FBI issues a national crime report once a year based on submissions from police departments, a practice that began in 1930 with the goal of providing annual trends in major crimes. At the urging of policing experts, crime victims and politicians, the agency in 2021 switched to a much more detailed set of statistics that many police departments have been slow to adopt — an evolution that has also further slowed the pace of reporting.

“The public perception of crime and the reality of crime rarely connect,” said Jeff Asher, a criminologist with AH Datalytics who pulls current data from local police, and is a member of the working group at the council, a nonpartisan think tank. “We’ve basically created a system where in the minds of many people, crime can only go up. So even when it’s going down, you’re never able to learn from and appreciate your successes. It makes people scared, and it leads to bad policy outcomes.”

In the summer of 2021, for example, when local police departments began noticing that a social media trend appeared to be fueling the theft of large numbers of Kia and Hyundai vehicles, it was not immediately clear whether they were seeing a long-term trend or a short-term spike — patterns that elicit a different response. Police didn’t know, because the most recent statistics they had on auto theft were from 2019 — data that was two years old.

The council cited shoplifting as a prime example. Retailers reported a significant spike in theft starting in 2020, locking up merchandise and increasing security, but police data showed that shoplifting had risen only to pre-pandemic levels, the council said. Facing that discrepancy, local governments decided to act, with 24 states creating retail theft task forces, while others enacted tougher sentences or rolled back previous reforms. Congress is considering creating an Organized Retail Crime Coordination Center.

“Perhaps these measures are appropriately designed and scaled,” the council wrote. “Without timely, accurate data, however, we just don’t know.”

“There can be a tendency to overreact to small spikes,” said Jerry Ratcliffe, a criminal justice professor at Temple University. “Sometimes it’s easy to pay too much attention to short-term spikes, and ignore long-term trends. At that point, we don’t have intelligence-led policing, we have media-led policing.”

Bill Brooks, the recently retired police chief of Norwood, Mass., and a committee chair for the International Association of Chiefs of Police, said police chiefs want to know about national crime trends that might be coming their way, such as the spread of fentanyl overdoses or the rise in shoplifting. But he said he sensed that “most agencies have a fairly good handle on what’s happening in their jurisdiction.” More local politicians “should probably listen more to their police chiefs about what they need,” rather than wait for outdated statistics, he said.

Another problem that has bedeviled crime statisticians is the slow switch nationally from the Uniform Crime Reports tallied up by the FBI since 1930, to the National Incident-Based Reporting System (NIBRS) instituted in January 2021. The Uniform Crime Reports covered seven major crimes: murder, rape, robbery, aggravated assault, burglary, larceny and auto theft. For decades, police and crime reporters used the statistics from those seven crimes to mark how a jurisdiction was faring.

But many crimes simply weren’t accounted for, whether it was securities fraud, wage theft or cyberstalking. Crime analysts applauded the expansion of data in NIBRS, which tracks 62 different offenses, which then-FBI Director James B. Comey said in 2015 would be the new national standard.

But many police departments were slow to make the transition, including New York City and Los Angeles. Only 57% of the U.S. population was covered by the new statistics in 2020, whereas it had been closer to 98% under the old format. In 2022, according to the Bureau of Justice Statistics, fewer than 10% of police agencies in Pennsylvania and Florida were sending their stats in NIBRS format, and New York’s statistics were covering less than 25% of the population.

The numbers are improving — Los Angeles and New York are both now compliant — and the council predicts that 83% of the population will be covered this year. But it is calling for greatly increased funding to the bureau to train police departments in gathering the more detailed data, which includes not only expanded crime types but also profiles of victims and offenders, types and value of property lost, and whether a case was cleared by the police.

“I think the transition to and rollout for NIBRS was much slower than anticipated,” said Alexis Piquero, a former director of the Bureau of Justice Statistics and now a professor of sociology and criminology at the University of Miami. “This involved getting 18,000-plus law enforcement agencies to change their coding, purchase new record management systems, train officers, hire staff, etc.,” he said. “That is not like updating one’s iPhone overnight. This is years in the making.” Piquero was also on the working group for the council’s report.

Brooks said many police departments were stuck with old records systems that couldn’t easily transition to the new system. Officers filling out their initial incident reports typically provide most of the required data, but “if the records management system isn’t asking them (for details about a case), that information isn’t going to come across.”

The council also called on the Centers for Disease Control and Prevention to improve its systems that collect data on violent deaths and firearms injuries, and for all federal agencies tracking crime to begin publishing a single, comprehensive annual report on crime and victimization data. The report also suggests that Congress fund a National Justice Data Analysis Center, both to analyze data and to help agencies collect and use their own data.

“Even small discrepancies can have big consequences in the policy and political environment,” the council wrote. “The potential fallout can be huge — in terms of crime tactics and strategies, whether the police chief stays on the job or gets fired, even whether the mayor or prosecutor is re-elected or driven from office. At stake are billions in government spending on public safety and criminal justice, the viability of businesses and urban centers, and the extent to which Americans are safe, and feel safe, in their homes and communities.”