The program was required to obtain data representing 60 percent of law enforcement officers, to meet a standard of quality set by the Office of Management and Budget, or else stop the effort by the end of 2022. In 2019, the data covered 44 percent of local, state, federal and tribal officers, and last year the total increased to 55 percent, according to the program’s website. So far this year, the data represents 57 percent of all officers, the FBI said Wednesday.
“Due to insufficient participation from law enforcement agencies,” the GAO wrote, “the FBI faces risks that it may not meet the participation thresholds” established by OMB, “and therefore may never publish use of force incident data.”
The Justice Department said in its response to the report that “the FBI believes the agreed upon thresholds will be met to allow the data collection to continue, and is taking steps to increase participation in data collection efforts.” The response by Assistant Attorney General Lee J. Lofthus also said that Justice “sent a letter to federal law enforcement agencies encouraging their participation.”
On Wednesday, the FBI said in an emailed response to questions that “each day is a new snapshot in time,” and that as of Oct. 18, the data represented 54 percent of officers. But by Wednesday, the “participation rates are at 57.15 percent for 2021,” the FBI said.
“I’d be surprised if they didn’t make 60 percent,” said Bill Brooks, chief of the Norwood, Mass., police and a member of the International Association of Chiefs of Police board of directors. He said a key problem is that many agencies that have no force incidents are failing to input “zero reports” each month, so the agency is counted as not participating. The IACP has long supported the data collection, and low participation numbers “make us look like we’re hiding something, when in reality I don’t think that’s the case.”
As of Sept. 30, 81 percent of federal officers were represented in the data, even though only 43 of 114 federal agencies, or about 38 percent, had participated by then, according to the FBI’s website. Two of the largest federal agencies, the Department of Homeland Security’s Customs and Border Protection and Immigration and Customs Enforcement, have been sending in data this year but the Justice Department’s largest agency, the Bureau of Prisons, had not.
The GAO report also says the Justice Department has largely ignored a requirement included in the 1994 crime bill to “acquire data about the use of excessive force by law enforcement officers” and “shall publish an annual summary of the data acquired.” No such summary has been published in at least the past five years, the GAO found. Justice Department officials suggested to the GAO that the national use-of-force program could provide that data, but the program does not differentiate between incidents involving reasonable force and those involving excessive force.
The impetus for the use-of-force data program was the fact that no government agency was tracking how often police killed citizens. Law enforcement officials, criminologists and other policing experts, said solid data was needed to know just how often police used force, and whether high-profile incidents such as the killing of Eric Garner in New York, Laquan McDonald in Chicago and Tamir Rice in Cleveland, all in 2014, were aberrations or the norm.
“Transparency and police data are what lead to accountability,” said Nancy La Vigne, executive director of the Council on Criminal Justice’s Task Force on Policing, last summer. “When you don’t know what use-of-force cases are happening, it’s difficult to know if you’re making improvements.”
The Washington Post began tracking fatal police shootings in 2015 through media reports and information from police. The Post has found roughly 1,000 fatal shootings per year, more than twice what was being reported annually to the FBI through its Uniform Crime Reporting system.
In 2016, the FBI declared its intent to start capturing its own data. Then-FBI Director James B. Comey said Americans “actually have no idea whether the number of black people or brown people or white people being shot by police” has gone up or down, or if any group is more likely to be shot by police, given the incomplete data available.
The FBI conducted a pilot program to collect data in 2017, and opened it up to all law enforcement agencies in 2019. The request for data is not minimal: the FBI wants the location and circumstances of every force incident, and detailed information on both the subject and the officers involved.
The FBI has said it will not publicly report data from any specific agency or incident, only by state. The OMB has said no data can be released if less than 40 percent of all officers are covered. If up to 59 percent of all officers are covered, the FBI “may publish limited information,” the OMB said, “such as the injuries an individual received in the use of force incident, and the type of force that the law enforcement officer used.” If more than 60 percent of officers are covered by the data, the FBI “may publish the most frequently reported responses to questions, expressed in either ratios, percentages or in a list format.” At 80 percent of officers, “the FBI may unconditionally publish collected data,” the OMB said.
Although the response rate covered 44 percent of officers in 2019 and 55 percent last year, the FBI has not released any use-of-force information so far. The names of participating agencies and the number of participants per state are available on the FBI website.
The number of agencies participating has steadily risen, the FBI’s website shows, from 27 percent of state and local police departments in 2019 to nearly 41 percent of departments so far this year. The FBI estimates there are 18,514 state, local, tribal and federal police agencies in the United States, with a total of 860,000 sworn police employees.
In prior years, the FBI accepted data long after the conclusion of the calendar year, and participation rose in those months. On Wednesday, the bureau said a deadline for this year’s data would be sometime in the first quarter of 2022.
But 60 percent is the magic number, the GAO’s report says, noting “OMB set these participation thresholds because a high response rate is an important indicator of data quality.” The report says “OMB officials also stated that use of force data are highly influential and, therefore, warrant a high standard of quality.”
Police officials have told the FBI that inputting the data can be time-consuming or difficult, the bureau reported after its pilot program. Art Acevedo, the former police chief of Houston and Miami, said that Houston police estimated they would need “three full-time employees to go through everything to pull all the data,” and that probably contributed to departments deciding not to participate. The FBI estimated it took about 38 minutes to enter the information for each incident.
Last year, after the police killing of George Floyd in Minneapolis, President Donald Trump issued an executive order calling for a database “concerning instances of excessive use of force related to law enforcement matters,” even though the FBI program was underway, and also said federal funds should be withheld from agencies that fail to contribute. The order was not addressed in the GAO report, but the idea of tying participation to federal funding was endorsed by many police officials.
But a report issued in March by the Congressional Research Service noted that many smaller jurisdictions don’t receive federal funds, and for larger agencies the federal funding was not big enough to compensate for the time spent compiling the data.
FBI officials told GAO investigators that they knew the 60 percent threshold “presented a risk to the program … because law enforcement agencies may be reluctant to share sensitive data on use of force incidents,” the report states. But the FBI reported it was focused on “a strong recruiting strategy to meet the OMB participation thresholds.”
Mark Berman contributed to this report.