How do we improve open data for police accountability?

Line drawing of handcuffed hands
Image credit: Wikimedia Commons

This is a challenging time for people who worry about the fairness of American governmental institutions. In quick succession, grand juries declined to indict two police officers accused of killing black men. In the case of Ferguson, Mo. officer Darren Wilson’s killing of Michael Brown, the grand jury’s decision appeared to center on uncertainty about whether Wilson’s action was legal and whether he killed under threat. In the case of New York City police officer Daniel Pantaleo’s killing of Eric Garner, however, a bystander recorded and made public a video of the police officer causing Garner’s death through an illegal chokehold. In Pantaleo’s case, the availability of video data has made the question about institutional fairness even more urgent, as people can see for themselves the context in which the officer exercised power. The data has given us a common set of facts to use in judging police behavior.

We grant law enforcement and corrections departments the right to exercise more physical power over the public than we do to any other part of our government. But do we generally have the data we need to evaluate how they’re using it?

It’s a critical time to think about this because the question of whether police are held accountable for their use of state power is urgent and immediate. In addition to the pressing concern about police accountability for violence against civilians, US senators, among others, have expressed surprise and dismay at another trend that exacerbates these issues: the stealth militarization of local law enforcement, federal programs that have allowed police departments to procure military armaments without even informing local governments or the public. (Even after months of media attention, police in 13 states still have not provided information about what military gear they’ve obtained through these programs, though recent events have just led the Pentagon to release more data about their equipment transfers.) A shocking Washington Post investigation into the police practice of civil asset forfeiture — documenting that police, nationally, have taken $2.5 billion from people over the last thirteen years — has helped increase national awareness of the legal latitude police enjoy to take property without charging anyone with a crime.

The time to find good solutions to these problems is now. Responding to widespread frustration, President Obama has just announced a three-part initiative to “strengthen community policing”: an increased focus on transparency and oversight for federal-to-local transfers of military equipment, a proposal to provide matching funding to local police departments to buy body cameras, and a “Task Force on 21st Century Policing” that will make recommendations for how to implement community-oriented policing practices.

While each element of Obama’s initiative corresponds to a distinct set of concerns about policing, one element they share in common is the need to increase access to information about police work. Each of the three approaches will rely on mechanisms to increase the flow of public information about what police officers are doing in their official roles and how they are doing it. How are police officers going about fulfilling their responsibility to ensure public safety? Are they working in ways that appropriately respect individual rights? Are they responsive to public concerns, when concerns are raised?

By encouraging the collection and publication of more data about how government is working, Obama’s initiative has the potential to support precisely the kind of increase in data availability that can transform public outcomes. When applied with the intent to improve transparency and accountability and to increase public engagement, open data — and the civic tech that uses this data — can bridge the often too-large gap between the public and government.

However, because Obama’s initiatives depend on the effective collection, publication, and communication of information, open data advocates have a particular contribution to make. It’s important to think about what lessons we can apply from our experiences with open data — and with data collected and used for police accountability — in order to ensure that this initiative has the greatest possible impact. As an open data and open government community, can we make recommendations that can help improve the data we’re collecting for police transparency and accountability?

I’m going to begin a list, but it’s just a beginning – I am certain that you have many more recommendations to make. I’ll categorize them first by Obama’s “Strengthening Community Policing” initiatives and then keep thinking about what additional data is needed. Please think along with me about what kind of datasets we will need, what potential issues with data availability and quality we’re likely to see, what kind of laws may need to be changed to improve access to the data necessary for police accountability, then make your recommendations in the Google Doc embedded at the end of this post. If you’ve seen any great projects you’ve seen which improve police transparency and accountability, be sure to share those as well.

“Strengthening Community Policing”: Improving body camera data

With the support of Obama’s proposed funding to support police purchase of body-worn cameras, we’re witnessing an important moment for the likelihood that police departments will adopt requirements that their officers wear them. It is therefore also an important moment for thinking about how to improve the collection and use of body-cam data. Body-worn cameras have been shown to reduce police use of force and public complaints, they were the main policy initiative communicated out of the Ferguson protests and they are now the subject of a national campaign spearheaded by Michael Brown’s parents.

However, while experiences to date suggest that increasing the adoption of police video is going to improve police/public interactions, it will also require some hard thinking about how we mandate the effective collection, storage and accessibility of this data. Perhaps most importantly, we must think about how we want this information to be used after it is collected. In all cases, privacy concerns will need to be balanced against the social benefits of publication. The following is the beginning of a list of considerations.

Changes needed to improve access to body camera data:

  • Widespread adoption of body cameras as a way to ensure that we have the data we need to achieve transparency in policing (and, in this way, increasing public confidence in police accountability).
  • Meaningful penalties for violating requirements to both wear cameras and keep them recording during the entire period of active duty. Where these do not exist, officers have chosen not to have their actions recorded. Similarly, ensuring these recordings are secure, both from inadvertent release and deletion, is critical.
  • Improving public access to police video data. This will require changes to state and local public records laws in order to ensure public access to police video, with appropriate restrictions included to protect individual privacy. The ACLU provides a useful starting point for thinking about how to achieve this balance with their list of recommendations for police body-mounted cameras.
  • Improving restricted access to police video data. Criminal justice researchers who have been trained in human subjects protection and whose work can be bound and verified through a data-sharing agreement can help improve police practices when they gain broader access to police-held data. The widespread use of video for police interaction has the potential to revolutionize policing, if this data can be studied.
  • Financial appropriations to ensure the effective storage, maintenance, appropriate redaction and accessibility of police video for a reasonable period of time.

“Strengthening Community Policing”: Improving data about equipment acquisition

As part of the “Strengthening Community Policing” initiative, Obama released the report that the White House had commissioned in order to review “Federal Support for Local Law Enforcement Equipment Acquisition.” This review was ordered in response to public surprise at the degree to which local law agencies were getting access to military equipment — much of it purchased originally for use in war zones — to use in the course of their local policing work.

Perhaps as surprising as the amount of military materiel currently in use by local police departments is the degree to which police departments have been able to independently acquire that equipment without local governmental or community oversight. The entire program, finds the report, demonstrates “a lack of consistency in how federal programs are structured, implemented and audited, and informed by conversations with stakeholders.” The report contains the following initial recommendations report for“improvements agencies might implement as a result of further collaborative review…:

  • Develop a consistent list of controlled property allowable for acquisition by LEAs [Law Enforcement Agencies].
  • Require local civilian (non-police) review of and authorization for LEAs to request or acquire controlled equipment.
  • Mandate that LEAs which participate in federal equipment programs receive necessary training and have policies in place that address appropriate use and employment of controlled equipment, as well as protection of civil rights and civil liberties…
  • Require after-action analysis reports for significant incidents involving federally provided or federally-funded equipment.
  • Harmonize Federal programs so that they have consistent and transparent policies.
  • Develop a database that includes information about controlled equipment purchased or acquired through Federal programs.”

Changes needed to improve access to equipment acquisition data:

The improvements recommended in the report are all useful, but they will only serve the purpose of ensuring public transparency and accountability when the data from their implementation is made freely and publicly accessible. Though some data is made available through individual programs (like the Defense Logistics Agency’s 1033 purchase listings) the current method for obtaining complete data about federal equipment grants to local departments depends on the willingness of volunteers and journalists to submit FOIA requests for it. It’s a parody of governmental transparency on a topic of extraordinary local relevance. Since the equipment in question is all donated or purchased through federal programs, the federal government has an excellent opportunity to make this happen. All of the data emanating from the implementation of these program reforms can be made available in centralized, standardized federal databases. The DATA Act provides one valuable template that could be used to determine how information about applications, awards, and grants from diverse federal agencies to a variety of local actors might be standardized and published.

“Strengthening Community Policing”: Improving data for police-community relations

The new task force that Obama proposes is charged with developing a report that builds on the work of the DOJ’s Community Oriented Policing Services (COPS) office. The COPS office provides local police departments with funding and resources to help them improve “trust and mutual respect between police and the communities they serve.” In support of those objectives, police departments around the country have modeled a number of useful ways to demonstrate their attention to community concerns through the transparent handling of challenging police-community interactions.

Data-driven approaches for improving public trust and mutual respect:

  • The act of giving people access to a public portal to lodge concerns about treatment by police can help all parties learn about problem areas in police-community relations. It is equally important to make public reports on complaints submitted to police and how they have been handled, such as those published by the San Francisco Police Department, in order to ensure that public concerns are properly addressed.
  • Every interaction between an officer and a person that involves the use of force should be logged and made public to increase review of these incidents. The Los Angeles Police Department’s use of force reports are a good example of how requiring detail in how these incidents are tracked can establish trust between law enforcement and the community.
  • Many larger police departments already have independent police auditors to keep checks on officers, but all could benefit from the objective review of potentially questionable actions they facilitate. The relatively small Bay Area Rapid Transit Police Department — who, under AB 1586, established their own review authority following the tragic killing of Oscar Grant by a BART police officer — have made auditor reports available to the public on a monthly basis.
  • False arrests can be incredibly damaging to innocent people and should be made transparent to keep law enforcement accountable when detaining potential suspects. A St. Louis Post-Dispatch investigation found that over seven years, about 100 people were wrongfully arrested for over 2,000 combined days by the St. Louis Police Department, due in large part to clerical errors.

Other important data for police accountability

  • Of the important issues in police accountability with heightened visibility this year, one that Obama’s initiative pointedly did not address was the issue of asset forfeiture. As documented this fall by the Washington Post, police have liberally used their legal latitude to take people’s property. The authorization police enjoy is so ripe for abuse, the Heritage Foundation points out that even the original architects of the civil asset forfeiture program now oppose it. Since it financially benefits both local and federal law enforcement, the program is clearly a tough target — advocates have been working to reform it for decades. However, at the very least, we can achieve improved data collection and publication about police stops and searches which include asset forfeiture. Regular public reporting (and auditing of these reports) will let the public better know how much money their local police force is collecting from civilians without judicial review. (Also, if you haven’t watched the John Oliver segment on this topic, you probably should.)
  • Police stop data is extremely valuable for understanding how police are interacting with their communities. Following public criticism of its controversial “Stop and Frisk” program, the New York City police department began collecting and making more data available about police stops. As a result of this data, advocates were able to demonstrate that the policy resulted in racially biased stops that did not lead to increased arrests or convictions. The data also demonstrated that the program, broadly speaking, was not effective at reducing crime. Today, the data the NYPD makes available on this aspect of police work may be the best in the nation. The “NYPD Stop Question Frisk database file specifications” reveals the list of questions police officers must fill in completing a report on each stop, providing data of sufficient quality to answer specific questions about program effectiveness. (Less detailed information is less useful, as users of the Newark, NJ  and Chicago, Ill. stop and frisk data have learned.)
  • While America’s common law history has broadly ensured public access to court records, some parts of the criminal justice process can remain quite impenetrable, depending on jurisdiction. Access to police records can be nearly non-existent in some states. For example, in Kansas, police records for many years have been sealed entirely, including after investigations are closed. While everyone can appreciate the value of protecting the integrity of ongoing investigations, this justification for withholding records becomes increasingly suspect when it is attached to closed files or when it serves the purpose of protecting administrative decisions. States should strike a more careful balance to ensure that people have the information they need about police job performance, just as they should about all other public servants.

The open data community, used to thinking about how to resolve political problems through use of the collection, communication and analytic efficiencies of new technology, can provide useful support to the broader effort to increase governmental transparency and accountability. This is particularly necessary work when considering such a vital governmental function as policing. Let’s think together about what data we need and how to make it accessible in order to provide greater oversight for and understanding of police action.

Please contribute your ideas here!