In our recent posts, we have explored the use of criminal justice data to improve policing and police-community relations. A number of officer-involved shootings (including the deaths of Michael Brown in Ferguson, Mo., Eric Garner in Staten Island, N.Y., Walter Scott in North Charleston, S.C., and Freddie Gray in Baltimore, Md.) have led to an increased demand of criminal justice data to hold public officials accountable and to improve the way our criminal justice system functions. But police data is only one type of criminal justice data that can be used to improve our criminal justice system and outcomes for the public.

According to Joan Petersilia, faculty co-director of the Stanford Criminal Justice Center, “police accountability data is good because it acts as a crossover issue … [it] helps people begin to understand” criminal justice data. As Petersilia also points out, there are a number of other transaction points in the criminal justice system that create data and can be used for accountability purposes. Ultimately, the goal that Petersilia and many others share is that we be able to take data from all points along the criminal justice system in order to assess the impact of what we do with the criminal justice system, allowing us to find the programs and practices that create the best possible public outcomes.

And we do know that the U.S. collectively holds a massive amount of criminal justice data. In our developing inventory, we are collecting and centralizing much of this data from law enforcement offices, departments of corrections and state court systems – as well as many other related organizations – from all across the country. However, several federal agencies offer critical data sources that are, of what publicly exists, the most nationally comprehensive and the most widely used of all criminal justice datasets. We will review these key sources of data, and then consider how current practitioners envision improving our criminal justice data systems beyond current practice.

The FBI’s contribution to the world of public criminal justice data: The UCR and NIBRS

The most well-known criminal justice data set is the Uniform Crime Report (UCR), collected by the FBI since 1930. The UCR includes statistics on seven crimes classified as either violent crime or property crime: murder, rape, assault and robbery fall under the violent crimes category, while arson, burglary, larceny-theft and motor vehicle theft fall under property crimes.

According to the FBI, the UCR serves many purposes and has many benefits for the people who access it. The UCR is described as benefiting law enforcement in all aspects of operation, from actual policing to budget formulation and to help local community development efforts. Criminal justice researchers use the UCR to study crime trends and find ways to improve the criminal justice system. The news media use the UCR to report trends to their audience and add context to stories.  

As an alternative to the Uniform Crime Report, many police departments report crime statistics to the FBI using a system called the National Incident Based Reporting System (NIBRS). The NIBRS dataset is lauded by many, as it provides a more comprehensive view of crimes that have been committed. The NIBRS system collects information about the nature and types of specific offenses that occurred during the crime, characteristics of the victims and offenders, types and value of property stolen and recovered, and characteristics of people arrested in connection with a crime that occurred.

Additionally, the Task Force on 21st Century Policing encouraged police agency participation in NIBRS in its interim report, saying, “greater acceptance of the Federal Bureau of Investigation’s (FBI) National Incident-Based Reporting System could also benefit policing practice and research endeavors.” Police agencies can generally choose between participating in NIBRS or the UCR program, and because the UCR has been around for decades longer with many agencies’ reporting systems built around it, changing to NIBRS is not always an easy option.

Below is a description of benefits the FBI says police agencies and society gain from participating in the NIBRS system:

The NIBRS system can furnish information on nearly every major criminal justice issue facing law enforcement today, including terrorism, white collar crime, weapons offenses, missing children where criminality is involved, drug/narcotics offenses, drug involvement in all offenses, hate crimes, spousal abuse, abuse of the elderly, child abuse, domestic violence, juvenile crime/gangs, parental abduction, organized crime, pornography/child pornography, driving under the influence, and alcohol-related offenses.

• Using the NIBRS, legislators, municipal planners/administrators, academicians, sociologists, and the public will have access to more comprehensive crime information than the traditional summary reporting system can provide.

• NIBRS produces more detailed, accurate, and meaningful data than the summary reporting system. Armed with such information, law enforcement can better make a case to acquire the resources needed to fight crime.

• NIBRS enables agencies to find similarities in crime-fighting problems so that agencies can work together to develop solutions or discover strategies for addressing the issues.

• Full participation in NIBRS provides statistics to enable a law enforcement agency to provide a full accounting of the status of public safety within the jurisdiction to the police commissioner, chief, sheriff, or director.




Even with all the benefits identified by the FBI that come from using this system, resource issues and legacy data systems already in place have kept the majority of police agencies from migrating to the NIBRS system since its pilot program in 1987. However, it seems that it was always understood that complete migration to NIBRS would be difficult. Because of that, at least one part of the original plan to implement the NIBRS system was to create a representative sample of data that could be applied to the whole country so the benefits described above could be felt by everyone, not just adopting agencies. But the plan to create a sample was never implemented and the only option became to have all 19,000 police departments adopt the program.

National Crime Statistics Exchange: The Bureau of Justice Statistics

Funding and other resource issues appear to stand in the way of the effort to successfully enroll all 19,000 police departments in NIBRS, but the FBI remains committed to achieving the goal. As the FBI acknowledges, however, though “participation [in NIBRS] grows steadily, data is still not pervasive enough to make broad generalizations about crime in the United States.”

In order to circumvent the problem of inaccessible local data, the FBI collaborated with the Bureau of Justice Statistics to create a valid sample of police departments currently using NIBRS, whose data and outcomes can be applied generally across the country. What’s come of the partnership is the National Crime Statistics Exchange (NSC-X). The Bureau of Justice Statistics describes it as “an effort aimed at generating nationally-representative, incident-based data on crimes reported to law enforcement agencies.”

NCS-X will use the existing data created by NIBRS by recruiting a sample of law enforcement agencies to join the more than 6,000 agencies currently participating in the NIBRS program. The hope is that the data collected through this new effort will make policing across the country more effective.

In a paper describing NCS-X, Howard Snyder, the Deputy Director of the Bureau of Justice Statistics said, “NCS-X also needs the support of advocacy groups, the media, and the research community at the local level. Local law enforcement agencies need to hear from persons in their own communities how important these data are to others.”

Snyder acknowledges that even with such a promising system for the future of incident data, a culture unfamiliar with the benefits of open data might inhibit the project’s success:

Even with the support NCS-X will provide to the sampled agencies, some agencies will need other encouragements to participate in the NCS-X initiative; for some, helping to produce detailed national crime statistics will not be enough of an incentive on its own. For these agencies, evidence of the value of NIBRS data to those outside of law enforcement community may tip the scales.

National Crime Victimization Survey

Sometimes thought of as a complementary dataset to the UCR – although it is produced by the Bureau of Justice Statistics, not the FBI – is the National Crime Victimization Survey (NCVS). The NCVS surveys a sample of households to understand the consequences of violent crimes and the likelihood of victimization. The list of violent crimes looked at in the survey are classified by the UCR, including rape and sexual assault, robbery, aggravated assault and simple assault. The survey also looks at victimization resultant of property crimes, including household burglary and motor vehicle theft. Similar to what NIBRS attempts to accomplish, but from the victims’ perspective, the NCVS provides an opportunity for people to understand the impact of crime and how likely people are to victimized.

According to the Bureau of Justice Statistics, “[The National Crime Victimization Survey] is the nation’s primary source of information on criminal victimization.” Annually, the survey samples 90,000 households comprising nearly 160,000 persons, asking about the frequency, characteristics and consequences of being victimized in the United States. The survey makes it possible for the Bureau of Justice Statistics to estimate:

… [the] likelihood of victimization by rape or sexual assault, robbery, aggravated and simple assault, theft, household burglary, and motor vehicle theft for the population as a whole as well as for segments of the population such as women, the elderly, members of various racial or ethnic groups, city dwellers, and other groups. The NCVS provides the largest national forum for victims to describe the impact of crime and characteristics of violent offenders.

When comparing the UCR and the NCVS, the FBI concludes that the two serve important complementary functions. According to the FBI:

By understanding the strengths and limitations of each program, it is possible to use the UCR and NCVS to achieve a greater understanding of crime trends and the nature of crime in the United States. For example, changes in police procedures, shifting attitudes towards crime and police, and other societal changes can affect the extent to which people report and law enforcement agencies record crime. NCVS and UCR data can be used in concert to explore why trends in reported and police recorded crime may differ.

The challenges of making improvements with insufficient data

However, even when considering the efforts made by the FBI and the Bureau of Justice Statistics to create good crime data that can make the country safer, observers often say the data isn’t sufficient. This becomes especially clear as we look at parts of the system beyond policing.

The Colson Task Force on Federal Corrections provides a useful example. Created by congressional mandate to address the issues faced by the Bureau of Prisons, the Colson Task Force is focused mainly on prison overcrowding and how that affects the prisons and the prisoners. The questions the group will attempt to answer are who’s in prison, how they got there and what problems will work to stop growth in the future.

Unsurprisingly, much of the group’s work will rely on federal data in order to make evidence-based decisions, including data collected by the Bureau of Prisons and the United States Sentencing Commission. However, some experts note that what’s missing for the Colson Task Force as it tries to answer its questions and make recommendations is data that explains why some prisoners are put in jail in the first place.

Criminal justice researchers point out that in order to assess why people are in prison, accessing data about how and why prosecutors decide to prosecute some offenders is very important. The argument is that it’s difficult to see how decisions in each step affect other actors in other branches of the government, including how prosecutorial decision making affects sentences and affects the prison population over time.

Joan Petersilia agrees that data that could be used to improve our criminal justice system and evaluate its effectiveness is missing, but she goes a step further to say that data driving the decision-making processes are oftentimes flat-out hidden. Where data about decision-making processes can’t be surfaced and accessed, it is impossible to assess whether our assumptions about “what works” are correct.

She demonstrates the problem through the lens of understanding “risk assessment tools” used in the criminal justice system. Risk assessment tools — instruments designed to tell whether someone is likely to commit crime in the future — are “ubiquitous” throughout the system, as Petersilia describes them, and are in essence a series of questions that are asked to determine if a person is prosecuted or released from prison on parole.

Specifically referring to risk assessment tools for parole boards, Petersilia said, “it’s impossible to uncover the factors being considered to let people out of jail.” The tools also present transparency issues, as Petersilia notes, with prisoners themselves not allowed to know what factors were used to determine whether they are kept in jail or released, making the decisions very hard to appeal.

In sum, when criminal justice data is of good quality and made available, good things can come of it. There are examples of different uses of criminal justice data that demonstrate how varied the benefits can be — such as the case in California, where emergency room doctor and gun violence prevention researcher Garen Wintemute uses data provided to him by the state of California to study ways to prevent gun violence. In New York, reporters used crime data to identify a political battle between Mayor Bill DeBlasio and the New York Police Department, which was an unexpected benefit of accessing data. In that case, reporters were able to access weekly police activity statistics and see that tickets had stopped being issued and arrests had stopped being made, indicating that an informal or unofficial strike had been organized.

It is helpful to consider the example of these effective major data sources and these use cases as we contemplate the remaining challenge. Many obstacles – from funding and focus, to politics and staffing issues – still prevent important data qualities, like accessibility and interoperability, from being the norm for criminal justice data in this county. However, as technology advances and as interest grows, we anticipate continued improvement. In the best case scenario, a virtuous cycle will emerge, and we’ll witness improved availability of criminal justice data and improvements to the criminal justice system based on analysis of that data going hand in hand.