Every year since 1930, the FBI publishes annual crime statistics based on data gathered from police departments across the country.
The FBI had always relied on voluntary submission of this data by police departments through the Summary Reporting System (SRS), a sub-branch of the FBI’s Uniform Crime Reporting (UCR), until January 2021, when the URC retired SRS as the channel of gathering data from law enforcement departments across the country.
Currently, police departments submit their data through the National Incident-Based Reporting System (NIBRS), which is more detailed and effective.
Lower Data Submissions
Since transitioning to the NIBRS, the FBI now releases crime data more frequently and is working with the Bureau of Justice Statistics to encourage police departments to submit data through the new system. Also, the two bodies are engaged in providing training and grants to police departments to encourage them to embrace reporting using the new approach.
“Submitting data to the NIBRS is purely voluntary for police departments, so not all police departments submit their data,” says attorney Teresa DiNardi. Even so, estimates are solely calculated based on the data submitted to the NIBRS. In 2021, approximately 66 percent of all police departments submitted their data to the NIBRS, a little lower than the 75 percent submitted the previous year through the SRS.
Unlike the older system, the NIBRS system requires detailed reporting, which can sometimes be challenging for some law enforcement departments, explaining the lower reporting rates.
Challenges to Estimates and How They Are Addressed
One of the challenges facing the new system is missing information. Missing information can be in the form of having a specific crime type reported correctly but having bits of critical information missing.
For example, an agency could submit data for a specific incident but have its rate per specific demographic, such as age, missing. To address the challenge of missing information, the NIBRS applies a process called amputations, which utilizes data from similar fields to create an estimate of the missing information.
The other challenge with the new system is partial reporting. Partial reporting occurs when an agency does some reporting but does not cover all the months in a year. Under such circumstances, the FBI and other agencies use the reported months to create a straightforward amputation ratio. For example, if an agency reported six months in the year, the figure can be doubled to get yearly estimates.
For non-reporting agencies, estimates are made by applying a statistical weight. A statistical weight involves taking data from another agency with similar characteristics as the non-reporting agency in terms of size and the type of agency and using the values as a representative figure for the missing agency.
Estimates Reliability
Due to the large number of agencies that failed to submit their data due to the change in systems, the statistical measure of confidence in the estimates deduced from the data is significantly lower. When the uncertainty of an estimate is too high, the data is deemed unreliable. NIBRS considers the reliability of estimates before determining whether the estimates meet the threshold for publication.
When a specific state’s data fails to meet the threshold, the NIBRS flags it as too unreliable, thus striking it off the final publication. The rates for data sharing between agencies and the NIBRS can vary widely. While the recording rates have seen a significant decline in agency reporting, experts believe that the uptake will improve with time.
