Bob Scales, chief executive of Police Strategies LLC, told the Community Police Oversight Board on Jan. 14 that racial disparities in Dallas arrest and use-of-force data are real but that simple comparisons to city population don't identify their cause.

Scales, who said he built the force-analysis system used by Dallas Police Department, told the board that analysts must consider context: whether contacts began from calls for service or officer-initiated stops, suspect behavior, age and gender, and whether arrests are linked to crimes reported by community members. He recommended using risk-adjusted benchmarks such as reported crime suspects, not raw census population, to test for evidence of bias.

That approach, Scales said, changes interpretation. "If 60% of reported crime suspects in a jurisdiction are a given group, you should expect arrests to reflect that distribution," he said. "Reported crime suspects is a much better benchmark to look at than census data."

Scales summarized several findings from his review of Dallas PD data (2020'2024): about two-thirds of use-of-force incidents began with calls for service rather than officer-initiated contacts; roughly 4% of arrests result in a use of force; and a relatively small group of officers account for a disproportionate share of arrests and force incidents. He also showed dashboard maps that identify two ZIP codes, 75216 and 75220, as arrest hotspots, with differing demographic profiles in each ZIP code.

Board members pushed on implications. Board member Jonathan Maples said the monthly arrest-by-district report "disturbs me" and asked whether the city's high arrest totals for some groups could reflect overpolicing. Scales responded that disparities in society'income, housing, and victimization'often drive disparities in policing data, and he urged the board to dig into root causes before drawing conclusions about police bias.

Other board members asked technical and policy questions. Brandon Friedman and others noted that arrest-level conclusions depend on additional data, such as whether blood or toxicology testing confirmed intoxication in individual DWI arrests, and asked for access to the dashboards and underlying datasets. Scales said the dashboards and the underlying incident-level extractions exist and offered to provide the board with copies and follow-up analyses.

Why it matters: The board has for months requested clearer explanations for high arrest rates by race in some council districts. Scales's presentation provided a roadmap for deeper analyses the board and department can pursue to assess whether disparities reflect policing choices or broader community patterns.

What's next: Scales said his team can run tailored follow-up analyses if the board supplies specific questions. The board asked the Office of Community Police Oversight and Dallas PD to collaborate with Scales on data-sharing and to make the dashboard available to the board for continued review.

Scales' presentation and the ensuing discussion highlighted the limits of population-based comparisons and the need for incident-level, risk-adjusted analysis before assigning cause or recommending policy changes.