“The more open we are, reality is, the more legitimate we are going to be perceived by the public,” said John Kapinos, a strategic planner with the Fairfax County Police Department, who has been in policing for 35 years. “I think you lose legitimacy when you hide data that doesn’t need to be hidden. It sets up the concept of trust – we work with the community, and as part of trying to establish that trust, we’re going to open the books and show you what we do.”

As yesterday’s post concluded, while data-based predictive policing may yet have a way to go before we can conclude it statistically improves on traditional policing methods for directly reducing crime, new methods of policing can create real benefits by helping improve community contact and community relations. Improving the public transparency of police practice through data release is proving to be a major public benefit arising from the use of criminal justice data.

This is an essential finding. In recent months, and in response to a number of officer-involved shootings that have received a great deal of national attention, activists have increasingly called for access to police data in order to help shed light on problematic police practices and to help find ways to improve relationships between the police and the communities they serve.

Challenges and opportunities

The types of data most frequently discussed in connection with police-community relations relate to the use of force, pedestrian or traffic stops, complaints and video data. What activists are often looking for is a way to measure how frequently police use forceful or inappropriately aggressive tactics when dealing with the public, and to be able to learn if those tactics are disproportionately used on certain segments of society, such as racial minorities or juveniles.

However, releasing data has been a longstanding challenge within the policing community, as various departments differ on their approach to making information available to the public. Kapinos explained that there are ingrained difficulties to releasing information. A “traditional culture” aims to keep data close to the vest (within the investigative unit, for example), and the technology available isn’t conducive to information sharing. But, he said, “there are tremendous benefits to much more openness and transparency,” suggesting that there are open opportunities to shift the conversation around data access.

“Just like corporations that share data with their shareholders, I think that that’s the same philosophy that police agencies and really all government agencies need to take,” Kapinos said. “Really, the public is our shareholders and they’re entitled to certain bits of information.”

Complicating the task further is that it has so far been difficult to quantify the benefits of releasing data for the purpose of improved police-community relations. “Better community relations” have been loosely observed, with unstandardized, qualitative measurements not fitting neatly into a metrics report.

That said, after a number of recent high-profile, officer-involved shootings situated themselves in public discourse (including the deaths of Michael Brown in Ferguson, Mo., Eric Garner in Staten Island, N.Y., Tamir Rice in Cleveland, Ohio, and Freddie Gray in Baltimore, Md.), the improvement of relations between police and their communities is a new national priority.

The opportunity to collect data relevant to this task has also been strengthened by White House mandate. In December, President Obama signed an executive order to create what’s called the Task Force on 21st Century Policing. The mission of the group is to “strengthen community policing and strengthen trust among law enforcement officers and the communities they serve.” At the hearings conducted by the task force, panelists frequently reinforced the role of data in achieving the group’s objectives. A great deal of the Task Force’s findings and recommendations released in an interim report in March 2015 focused on data collection by police departments and noted the lack of information available to evaluate the state of policing in America:

Data collection, supervision, and accountability are also part of a comprehensive systemic approach to keeping everyone safe and protecting the rights of all involved during police encounters. Members of the Division of Policing of the American Society of Criminology recently wrote, “While the United States presently employs a broad array of social and economic indicators in order to gauge the overall ‘health’ of the nation, it has a much more limited set of indicators concerning the behavior of the police and the quality of law enforcement.”

The task force also highlighted that police use-of-force data — a dataset the media has frequently reported as nonexistent on a national scale, but which has been asked for regularly since Michael Brown died last July — is actually required by a 20-year-old law to be collected by the Department of Justice, but isn’t due to a lack of resources:

[The American Society of Criminology] noted that Section 210402 of the Violent Crime Control and Law Enforcement Act of 1994 requires the U.S. Attorney General to “acquire data about the use of excessive force by law enforcement officers” and to “publish an annual summary of the data acquired under this section.” But the U.S. Department of Justice (DOJ) has never been allocated the funds necessary to undertake the serious and sustained program of research and development to fulfill this mandate.

The interim report went on to recommend a series of collaborative actions for two Justice Department offices, the Community Oriented Policing Services (COPS) office — the group housing the task force — and the Bureau of Justice Statistics (BJS).

According to the report, the BJS and the COPs office can:

• Establish a central repository for data concerning police use of force resulting in death, as well as in-custody deaths, and disseminate this data for use by both community and police;
• provide local agencies with technical assistance and a template to conduct local citizen satisfaction surveys;
• compile annual citizen satisfaction surveys based on the submission of voluntary local surveys, develop a national level survey as well as surveys for use by local agencies and by small geographic units, and develop questions to be added to the National Crime Victimization Survey relating to citizen satisfaction with police agencies and public trust.

The broad effort to improve local police data release is also beginning to take form. In early April, a White House initiative related to the task force brought together over a dozen police chiefs from cities across the country to discuss how the police can open data to improve relationships with their communities. The project aims to support police departments who are ready to lead in this area — departments like Dallas, which openly and proactively releases officer-involved shooting data online.

These developments are also being facilitated by other police-support organizations. The Police Foundation recently released a public information sheet on open data in policing to communicate the key principles and rationales for the proactive, online release of accessible data by police departments. In January, the International Association of Chiefs of Police released the report from their National Policy Summit on Community-Police Relations, buttressing the case for opening police data by observing, “Transparency in all areas is key. Open, accessible reporting of statistics, arrest information, and any other law enforcement data is expected, even when the information provided does not paint the best picture.” The Police Executive Research Forum, meanwhile, supported a pledge made by police chiefs attending its September 2014 meeting to “quickly release details of [officer-involved] shootings, including names of officers involved in jurisdictions where it is legal to do so,” with the goal of greater information transparency.

The use of data to improve police-community relations

Stop-and-frisk

While the current movement to encourage police to release data broadly is in its early stages, earlier efforts have already paid off by improving outcomes that communities perceived as unfair. The case of released stop-and-frisk data provides an important example of this: New York’s public release of granular pedestrian stop data, and the analysis it permitted, led to the discovery that almost nine out every 10 people stopped were entirely innocent, and that nine out of every 10 people stopped were nonwhite.

The analysis then powered policy change. The stop-and-frisk data show the NYPD has dramatically reduced its stop and frisk practice since 2011 when the department conducted 685,724 stops in a single year. In 2014, the New York Police Department conducted 46,235 stops – a reduction likely due to the public release of data and awareness of misconduct.

The example demonstrates the great significance of collecting and releasing data for the observations of problems and solutions in policing. As the NYPD example shows, collecting and releasing data indicated that there were issues with how the department carried out stop and frisk practices. Had the department not collected this information, it would be impossible to assess the practice’s efficacy. As the task force report points out: “Expanded research and data collection are also necessary to knowing what works and what does not work, which policing practices are effective and which ones have unintended consequences.”

Police body-worn cameras

Another data source that has been the subject of substantial recent attention is the video data from police body-worn cameras. The controversial officer-involved shootings of the past year have generated demonstrable public support for police departments to adopt body worn cameras as a regular practice, and the practice has also been supported by a number of law enforcement actors as well.

Decisions about how to legislate the recording and release of data, however, are still very much in flux. According to the National Conference of State Legislatures, as of Feb. 20, 2015, 30 states had introduced legislation regarding the use of body cameras by law enforcement.

While little research has been done so far on the effectiveness of body cameras in improving police practices and holding officers accountable, there has been some evidence that suggests that the cameras can in fact create positive outcomes and even help to de-escalate heated interactions. A report by the Data and Society Research Institute summarized a good deal of the research done to date on body camera, including a case study of Rialto, Calif., that appeared to be a harbinger for positive outcomes for police departments that adopt the technology. The Rialto Police Department outfitted half of its officers with body-worn cameras and had those officers film all of their encounters over 12 months. Compared to their control group, the group wearing cameras showed a 50 percent reduction in use-of-force incidents and achieved an 88 percent reduction in complaints against officers. In addition, “anecdotal evidence reported that the public tended to be more polite when they knew cameras were rolling, and individuals were more likely to retract fraudulent complaints.”

Meanwhile, as quickly as people started demanding police officers start wearing cameras, issues surrounding their use have been brought up starting useful discussions about how to use the cameras and what to do with the data they produce. Both the ACLU and chiefs of police from around the country have cited a person’s expectation of privacy as something to consider as the technology is being implemented. Commonly referenced is the frequent occurrence of an officer entering a private home in response to domestic dispute, where the cameras can either intimidate witnesses, exploit a vulnerable victim or both. People are also debating how cameras can be used most effectively; specifically, when they should be switched on and how much of an interaction they should record. The Seattle Police Department has led the charge in finding solutions to the problem of redacting data captured by the videos, with the overarching goal of figuring out how to effectively blur faces and other sensitive material. One major challenge lies in automating that redaction.

Many further questions remain to be determined in figuring out the details of body-worn camera data access. How long should a police department hold on to a video before destroying it, if ever? Under what circumstances and when can it be released to the public? Under what conditions might the videos be entirely withheld from the public? As these questions are being worked out, the legal terrain is rapidly shifting. Washington, D.C., for example, is currently answering this question through a series of low-profile administrative decisions, from proposed city budgets to new internal police policies. But the issue is still in the midst of serious and active debate.

Meanwhile, we hope that these questions of police transparency, in which the public is so directly and passionately involved, represent only the start of broad public attention to criminal justice data. In our next post, we will describe some of the major current sources of national-level criminal justice data, how they are improving and how the practitioners who make use of them say they can still be improved.