About

Use of Force

Use of force, including any shootings by officers, is necessary to assess whether force is more likely to be applied in minority communities when other factors are adjusted for. It is also crucial to publish whether internal investigations on the use of force deemed the action justified to understand whether there is internal accountability.

Officer Complaints

The public availability of complaints against police officers is essential because it is key to understanding what critical complaints against officers are and how they are being resolved internally. This data access would also theoretically prevent officers who are repeat offenders from continuing on the police force without repercussion or evaluation.

Police Demographics

Police force demographics are necessary for communities to understand if the police force truly reflects their community. If a police force fails to retain women and people of color, this could be an indicator of issues with the internal culture and processes of the force.

Stop and frisk

Stop and frisk and Traffic Stop datasets are both necessary for communities to evaluate whether their police force is more likely to stop certain populations and if there is bias in the level of reaction to an incident.

Traffic Stop

All these datasets being openly available would empower communities to hold police forces accountable and build trust (Wardell and Ross, 2020).



We conducted interviews with various stakeholders and experts of police data publishing to gain a qualitative understanding of barriers in publishing police data. The qualitative results are below.

Systemic Reasons for Reluctance Surrounding Publishing

From data collection, exporting, and publishing, to the people available to accomplish all of those processes, there are several systemic reasons for why there is a lack of police accountability data available to the public. Data collection, especially surrounding the previously identified five datasets, is largely dependent on officer reports of incidents. This means that fields in the data, such as race, are entirely dependent on the police officer’s best judgement of the civilian they are interacting with. This data is almost never checked with self-reported data from the civilian, and is also often missing other important factors of identity that may explain bias, such as sexuality or gender presentation. This could lead to inaccurate data being published, and there would need to be considerable effort to communicate that the data published is entirely subjective.

Other reasons include hesitancy to accept accountability in general, and difficulty in knowing how to communicate the intricacies of certain interactions. With any published piece of data, there are parts of the interaction that will be obscured for privacy or resource related reasons, but these parts of the interaction may include important context that is useful to understand the outcome of the published incident. Other difficulties in communication come from certain incidents, such as that in use of force data, being hard to normalize or communicate in a concerte, objective way.

There are many aspects of the police data publishing process that lack adequate resources to make the publishing process smooth. One of these aspects is the actual records management process itself. There are large varieties in record management processes, with some smaller police departments still using paper-based methods, and others using privately generated Record Management Systems (RMS) that only allow for certain fields of data to be inputted. From here, for open data to be achieved, the data needs to be normalized and privatized before publishing, which often requires a trained data management employee or staff, depending on the size of the department. The antiquated systems, along with a lack of staff not trained in working with data that needs to be made available for the public, and lack of funding for data publishing, are all resource related reasons that police departments currently steer away from publishing data. Knowing this, implementing incremental changes into the recording of data with current staff so that data is more readily available and publishable will be key to making data available without significant changes in the resources available to departments.

Lack of Understanding About Benefits of Publishing Data

Finally, there seems to be a lack of understanding and belief in the benefits of publishing police data, either at all or beyond the current modes in which data is currently being published. Beyond vague incentives such as greater transparency and public trust, it is difficult for police departments to understand why they should invest more resources into publishing data when publishing data has the possibility of inviting more scrutiny into their day-to-day operations. Furthermore, there is the possibility that people may not even trust the data that’s published unless it's verified by a third party, either press reports of incidents or an external auditor of data, which indicates that even more effort would need to be made to publish data. However, to address this lack of belief in the benefits of open data, incremental improvements will be key to building belief on the police side of publishing data. By publishing small amounts of data at a time, NGOs, public actors, and the general public themselves will be able to engage with data in a meaningful way, and trust in the community and benefits of open data can grow. This, of course, can all be sped up or encouraged if the federal government were to encourage and standardize data from departments, as they have with the Crime Data Explorer. Federal powers also have the ability to potentially mandate reporting of certain data fields, bypassing many of the obstructions detailed above.