The Sandra Bland case arrived in national media first with the news of her death in police custody. Bland, an African-American woman who was starting a positive new chapter of her life as she drove to Prairie View, Texas, to start a new job, now plays a role in a far more terrible, ongoing national narrative: the story of police officers committing excessive and often unaccountable violence against civilians. At a still larger level of analysis, Bland’s case also provides more support for the critical role that video plays in our work to improve police-civilian interactions.

Video is the unquestionable, key playing field of evidence in the nation’s work to understand what police are doing, to whom and why. Meanwhile, one of the things we have learned from the past year of public conversation about police accountability is that videos of incidents are insufficiently available unless public policy protects our rights to them. To continue to push for better laws ensuring police accountability, we must also have policy that protects our right to gather video evidence of the need for greater accountability.

The policies that support governmental accountability — including police accountability — are created through political processes. Policy change happens as a result of marshaling clear evidence of a need for change. As a result of publicly available evidence, policies requiring police to be held strictly accountable for use of force are beginning to come into being. It’s at a pace that feels too slow, but it is movement. It’s also only happening as a result of the clear evidence that video provides.

Since the videotaping of the Los Angeles police beating Rodney King in 1991, access to video has been crucial to understanding violent police-civilian interactions. Video gives voice to people who — given the cruel catch-22 that society mistrusts people who are arrested — are automatically rendered voiceless and suspect. While police have traditionally enjoyed great social trust, we now have, in mind’s eye, many recent memorable images of police violence against individuals. We’ve read stories about retaliation against officers who break ranks to acknowledge police wrongdoing. Although arrestees are a far less trusted class than police officers, we still know law enforcement accounts of incidents may require additional corroboration.

Sandra Bland’s case, as it unfolds after her death, points to many of the places we need to build or preserve policy.

• The public right to record. As video increasingly becomes the most trusted form of evidence about police-civilian interactions, it is important that we protect the right of people to provide it. In Sandra Bland’s case, as well as in the police-caused deaths of Walter Scott in North Charleston and Eric Garner in New York, video of the incident was recorded by bystanders with their own devices. The existence of video in these cases helps demonstrate why the public right to record police interaction is so critical. First, when the interaction occurs away from a law enforcement camera, bystander video may be the only video evidence. Second, where law enforcement video does exist, the presence of alternate sources of video verifies the authenticity of the official video. While there has generally been a basic right to record police in public in the U.S., this right is under attack in a number of jurisdictions.

• The public right to access official recordings. The last year has witnessed a massive wave of adoption of body-worn cameras by police departments, a development we have generally cheered. However, there is an increasingly urgent need to couple these new public requirements to record with public rights to access the official recordings. The release of the official video documenting Bland’s arrest has provided essential documentation of a problematic arrest. This official evidence, shared between police and the public, provides a way to learn about what the Waller County police believe constitute reasonable police practice, since this is the record the department has for their internal examination as well.

  Meanwhile, access to police records is an especially complicated area of public records law, as documented by the Reporters’ Committee for Freedom of the Press (RCFP) in their Open Government Guide; while many states permit public records requests for a number of police records, other states do not, and most states limit records that are classified as belonging to ongoing investigations. These questions of access are equally–if not more–present in the area of police video data, whether from body-worn cameras or police dash-cams. RCFP has also been documenting the evolving policy landscape for accessing police body-worn cameras. As their map demonstrates, there are aggressive policies developing in many places that remove the public right to access these videos, making them effectively useless as tools for public accountability (though they undoubtedly remain useful for internal accountability.) The D.C. Open Government Coalition has also just prepared a 50-state review of existing policy on the public right to access body-worn camera and police dashcam video. While there are legitimate questions to figure out in order to appropriately protect privacy while maintaining public right of access, these questions must not result in a decision to broadly foreclose public access.

• The public right to an accountable evidentiary process. Problems with the official video of Bland’s arrest that was released to the public underscore the serious need for strong protective rules for this public evidence. As RCFP’s review of current policy demonstrates, evidentiary rules for police video are still under evolution in many jurisdictions. Public video requires a clear chain of custody, a clear requirement to maintain the records for public review, and strong penalties for violation of those rules.

In the highly-charged and inevitably biased interaction between police and civilian, video alone provides a raw view of the encounter. Video has become our trusted third party to the dispute, providing some degree of objectivity in an intensely subjective moment. For that reason, in order to improve policy regarding law enforcement-civilian interaction, we need to ensure that we hold on to this right to see into these events, since they help us understand what needs to change.

As we continue to learn how best to bring our policies for ensuring police accountability into line with public needs, we will need to have good evidence about what currently exists in the world of police-civilian interactions. If we care about fixing the parts of our system that are broken, it is incumbent on us to protect our right to see the problems.