JRSA Forum. March 2015. Volume 33, Number 1.

Crime and Public Safety: Working Toward a More Accurate Assessment of Criminal Incidents

Jeffrey Sedgwick, Ph.D., Executive Director, Justice Research and Statistics Association

We as a nation are in a considerable intellectual ferment and creative thinking at the grassroots level when it comes to public safety. In our nation's capital alone, three significant groups are working on parallel tracks.

The White House's Taskforce on 21st Century Policing aims at enhancing trust between law enforcement agencies and the people they protect and serve since this is deemed essential to the stability of our communities, the integrity of the criminal justice system, and the safe and effective delivery of policing services.1 A principal way of enhancing trust is increasing transparency and promoting shared understanding of crime, policing, and public safety.

Meanwhile, the Crime Indicators Working Group, sponsored by the Bureau of Justice Statistics (BJS) and hosted by the International Association of Chiefs of Police (IACP), is working to create more detailed, accurate, and actionable measures of crime, criminality, and justice system performance. The group is charged with providing guidance in the development of crime indicators using police administrative record information and other data sources to supply a better picture of the crime problem in local jurisdictions and throughout the nation.2

Finally, across town at the National Academy of Sciences, the National Research Council's Panel on Modernizing the Nation's Crime Statistics is charged with assessing and making recommendations for the development of a modern set of crime measures in the United States and the best means for obtaining them. For example, better information is needed on certain crime types, such as crimes against businesses or organizations as well as personal identity theft; also needed is a greater ability to associate attributes such as firearms or drug involvement with crime types, and more complete adoption of electronic reporting, data capture, and system interoperability.3

Emerging from these discussions is a consensus view that the nation needs more than simple annual summary counts of offenses in order to think more strategically about crime, to enhance trust between communities and their law enforcement agencies, and to promote a sense of community wellness and safety. In short, attention is shifting from the FBI's traditional Uniform Crime Reports (UCR) of Crime in the United States to more detailed collections of crime incidents and their characteristics. The National Incident-Based Reporting System, or NIBRS, is such a collection.

First proposed in 1982 in a study sponsored by BJS and the FBI, NIBRS was designed to collect detailed information on the attributes of each crime incident known to law enforcement, including: the date, time and location of the incident; a detailed list of all offenses that occurred in the incident, not just the most serious offense; demographic information on each victim and offender involved in the incident; the relationships between each of the victims and the offenders; other details of the incident, including victim injury, type of weapon involved, alcohol or drug involvement, property loss, and drugs seized; clearance information, including both arrest and clearances by exceptional means; and date of arrest and arrestee demographics.4

Thirty years after its introduction, about a third of the nation's law enforcement agencies reported their UCR crime statistics via NIBRS. In 2012, NIBRS-contributing agencies served approximately 30% of the U.S. population and accounted for 28% of all crime reported to the UCR Program. Since many of the police departments representing large metropolitan areas throughout the nation do not yet contribute data to the system, incident-based law enforcement data are still unable to address the current information needs of policy makers, researchers, the media, and the public.

The intriguing question is why adoption of NIBRS is proceeding at a slow pace when the need for transparency and better data is so widely acknowledged. Some insight into this question can be gained by examining a recent Inspector General (IG) report5of a federal incident-based reporting system that does not share its data with the FBI and thus goes unreported to the public: the Defense Incident-Based Reporting System of the United States Department of Defense (DIBRS). The DoD's experience with its own incident-based crime reporting system illuminates a number of the pitfalls and challenges law enforcement agencies face as they transition from summary reporting to incident-based reporting.

In its investigation, the IG found that "DoD is not reporting criminal incident data to the Federal Bureau of Investigation (FBI) for inclusion in the annual Uniform Crime Reports to the President, the Congress, State governments, and officials of localities and institutions participating in the Uniform Crime Report Program, as required by Federal law."6 The relevant Federal law is the Uniform Federal Crime Reporting Act of 1988 (28 U.S.C 534 note), supplemented by additional data collection and reporting requirements contained in the Victims' Rights and Restitution Act of 1990 (49 U.S.C. 10601 note) and the Brady Handgun Violence Prevention Act (18 U.S.C. 921 note).

At the root of this failure to report incident-based criminal data to the FBI is the fact that the DIBRS system has never been certified by the FBI. As the IG noted, "For an agency to submit criminal incident data to NIBRS for inclusion in the annual uniform crime reports, the FBI's Criminal Justice Information System (CJIS) Division, Crime Statistics Management Unit, Uniform Crime Reporting Office, must first certify the agency's data system."7 System certification, in turn, depends on three criteria: the agency's reporting system must be compatible with the FBI's UCR system; the agency must demonstrate its ability to update submissions, meet deadlines, respond to FBI queries and requests, and correct errors received from the FBI UCR Program in a timely manner; and the agency must achieve a sustained error rate of 4% or less for three separate data submissions.

The challenges that the DoD's incident-based reporting system faces in terms of certification begin with its inability to consistently report data within 15 days of the end of each month as called for in the DoD's own guidelines for the operation of DIBRS. Indeed, the IG noted that the "Air Force Office of Special Investigations (AFOSI) was not reporting DIBRS information during the period of August 2012 through January 2013. The DIBRS reporting lapse was due to the AFOSI DIBRS program administrators departing before replacements were trained on their responsibilities."8 In addition, the Defense Criminal Investigative Service (DCIS) does not report any incident data to DIBRS because "DCIS could fund only its primary mission requirements and could not meet unfunded mandates such as DIBRS reporting."9 So lack of funding and lack of training are hampering DoD's ability to abide by federal legal requirements to report incident-based data.

Another problem with implementing incident-based reporting is the lack of a consistent method of correcting identified errors in data submitted to DIBRS. For example, the Army's Criminal Investigation Command (CID) does not correct identified errors in incident records; rather, it simply deletes the incorrect information and submits a null entry for the data element in question. According to Army spokespersons, "stripping the errors" was a more efficient way to manage the timeliness of DIBRS submissions because CID does not have a process to obtain corrected data from the originating investigative field unit.10 The Navy's Criminal Investigative Service (NCIS) reported that it submits only error-free incident records to DIBRS since "NCIS does not have the time, money, or resources to correct the errors."11 To get an idea of the magnitude of the resulting loss of information about criminal incidents in the Navy, consider the following: In February 2013, NCIS had a total of 2,097 incidents that qualified for DIBRS submission. Of these, 121 incident records were error free and submitted to DIBRS; hence, NCIS is reporting 5.8% of the criminal incidents it investigates. And, of course, since DIBRS is not FBI certified, this error-free incident information never reaches the NIBRS database for inclusion in the FBI's Uniform Crime Report.

The DoD's experience with operating an incident-based reporting system for criminal offenses is instructive as the nation transitions from a summary count system of recording crime to an incident-based reporting system with much richer content and potential to inform policy makers, the public, and public safety officials. To be successful, this transition must forthrightly address technology issues in terms of system interoperability; temporal issues in terms of timely submission of initial data as well as error corrections and record updates; issues related to training of system operators; and data quality issues, including consistent methods of error correction (instead of data or record deletion). The benefits of an incident-based reporting system are manifest; the path to their realization deserves to be carefully charted.

1 See "Fact Sheet: Task Force on 21st Century Policing" (December 18, 2014).
2 See David J. Roberts, "Technology Talk: Advances in Law Enforcement Information Technology Will Enable More Accurate, Actionable Analysis," The Police Chief: The Professional Voice of Law Enforcement (February 2015).
3 See "Modernizing the Nation's Crime Statistics: Project Scope."
4 See "Data Collection: National Incident-Based Reporting System (NIBRS)."
5 Inspector General, U.S. Department of Defense, "Evaluation of the Defense Criminal Investigative Organizations' Defense Incident-Based Reporting System Reporting and Reporting Accuracy." (October 29, 2014).
6 Ibid., p. i.
7 Ibid., p. 4.
8 Ibid., p. 6.
9 Ibid., p. 7.
10 Ibid.
11 Ibid., p. 8.