Data is the New Sheriff in Town, but is it Biased?

Today, cities and agencies face critical issues, which are the catalyst for developing solutions filtered through dual lenses: law enforcement and the communities they serve. The process of consolidating and making data available to the public to demonstrate transparency in hopes of improving community policing is fast becoming a trend. But how do we interpret this data?

Data is data. Data doesn’t lie. While these statements are true, data doesn’t tell a story by itself. In my work at Slalom with the City of Oakland, this premise is especially relevant. Our approach includes working closely with public safety and law enforcement agencies, detailed data gathering and interpretation, and using supporting research like Stanford University’s ground-breaking SPARQ (Social Psychological Answers to Real-world Questions). All of these components are factors in being able to draw a more meaningful and objective analysis to tell a holistic story. We applaud law enforcement agencies that have taken proactive steps to answer tough questions and maintained the discipline to provide the complete picture for communities and believe even more can be done in this space.

Communities Gain Immediate and Ongoing Transparency with Data Visualization

From routine stops to more complicated officer interactions, law enforcement complexities require visibility to answer questions such as:

• Do aggregate incidents reflect patterns of biases by age, race, or precinct?
• What types of incidents parallel officers’ tenure or department training?
• What information helps promote safety and keep citizens informed?
• What information can we provide to policing agencies to empower them with data to make informed decisions for the public and measure the impact of policy changes?

On the surface, a question about racial bias, as an example, may seem as straightforward as looking at the racial breakdown percentages by the type of incident. However, the complete story may need additional data for context, such as the incident area’s crime rate percentages compared to other areas or the demographics of the incident area. A perception of bias may look different within the context of additional relevant data.

One of the most critical aspects of police transparency is the timing and availability of information. With dashboards, supervisors and command centers can see daily updates on all situations instead of requesting reports that often take weeks to create for lack of data standardization. Dashboards reflect unbiased communications when segmented data includes the relevance and transparency of all correlating data.

High Impact Solutions for Law Enforcement

Slalom has engaged with multiple law enforcement agencies across the country and in these engagements, there are patterns we’ve observed in developing impactful data solutions. We’ve identified 3 factors in what’s important for the success of these projects.

• Understanding your organization’s current data maturity level
• Availability and accessibility of data and systems
• Deeper understandings of the problems and impact you are solving for

Most often, law enforcement agencies have varied methods of collecting and storing data. Some organizations are mature in their data collection process, while other departments are just getting started. Having this accurate view of your current data landscape gives you a starting point on what goals to set.

An easy and typical way of looking at arrests is to analyze how many arrests by race or ethnic group happened in different areas of the city. By adding in additional data points, like training or academy data from a disparate system, we can gain insight into the type of officers making arrests and start to analyze potential patterns between training and arrests. This simple example highlights the importance of defining baseline metrics and then looking for additional data sources for further analysis.

In addition to data set discovery, our approach is to seek understanding of an agency or city’s unique problems through deeper conversations with law enforcement. The goal is to help departments immediately visualize the potential to get the answers they need. In many cities facing budget shortfalls due to the COVID-19 pandemic or redirecting budgets to safety programs, dashboards can spotlight staffing and training levels, as examples, and help make data-driven funding allocation decisions.

Here are three example dashboards, built by a Slalom colleague Jeff Fowler, that show how integrated data can be used to address personnel needs, training budgets, and community awareness. These solutions are designed to include the available public data of each police department.

Chandler, Arizona Police Department Arrest Summary

The Chandler Police Department makes its Arrest History (2013–2020) data publicly available to promote and encourage joint problem solving, enhanced understanding, and accountability between communities and enforcement agencies. The public has the same level of access to details officers provided in incidents reports, measuring the number of arrests by factors such as suspect gender, age, race, ethnicity, and officer details such as gender, race, age group, and tenure. Also prominent are historical arrests, which aggregate a total number by month and the ability to drill down for details.

Phoenix, Arizona Officer Involved Shootings Summary

In Phoenix, the Officer Involved Shootings data covers almost four years of incidents, showing total incidents, historical incidents, fatal and non-fatal incidents, and incident specifics such as gender, age group, and race by officer and suspect. All detailed incidents are provided, too.

San Francisco, California Stop Summary

The San Francisco Police Department Stop History (January 2007-June 2016) is a public dataset available to fulfill its commitment to treating all people with dignity, fairness, and respect and eliminating any perception that policing appears biased. This dashboard provides a historical view of stops by several demographic data points, including age group, race, the reason for the stop, stop by district, and more.

By allowing law enforcement and the public to visualize incident actions and impact, these solutions assist with:

• Reduced implicit bias and racial disparity in interactions
• Improved police accountability at an officer level
• Increased transparency and understanding of police interactions
• Improved community relations by empowering supervisors with data tools
• Ability to measure department-wide policy and its impact

If you want to build confidence with data-driven law enforcement and police transparency with the expertise of integrating the appropriate correlating data, contact Leslie Wan to have a conversation about how we can assist your agencies and community.

“I think there’s a unique way that Slalom engages with police departments to help them come up with solutions, rather than just implementing a predetermined product.”

- Virginia Gleason, Bureau of Services Director, Oakland Police Department

Disclaimer: Slalom is a partner of Tableau.

Slalom is a modern consulting firm focused on strategy, technology, and business transformation. Leslie Wan is a data visualization and data architect. She is a Senior Delivery Principal with Slalom for the Bay Area and an expert in the possibilities of using the Tableau data visualization platform. Reach her at leslie.wan@slalom.com