Pursuit Policy Standards,
Start, Continue, Terminate

Pursuit is among the most dangerous discretionary decisions in policing, and the risk falls on everyone near it: uninvolved drivers and passengers, pedestrians, the officers, and the person fleeing. National policy responded by channeling the discretion into three named decision points, each with stated criteria, so that a pursuit is never a single choice made once and ridden to its end.

Initiation

The initiation question is what offense and what risk. Policies weigh the seriousness of the suspected offense against the danger the pursuit itself will create, and they weigh it before the pursuit exists. Conditions count from the first second: speeds, traffic density, weather, lighting, and where the pursuit will run. Where the driver is identified and can be apprehended later, that fact weighs against starting at all, because apprehension deferred is risk avoided.

Continuation

Continuation is not a decision made once. Policy requires the risk-versus-need balance to be run continuously as conditions change: speeds climb, traffic thickens, the route approaches pedestrians, a school zone or an intersection sequence that raises the stakes. A pursuit that was defensible at initiation can become indefensible three minutes later without anyone making a new mistake; the conditions changed, and the policy question is whether anyone re-weighed them.

Termination

Termination has two forms: who may call it, and when policy requires it. The pursuing officer may always terminate. Supervisors are empowered to order termination, and policies define conditions in which termination is mandatory rather than discretionary. Termination is also a practice, not just a word on the radio: shutting down emergency equipment, visibly breaking off, and reporting position, so the record shows the pursuit actually ended when the order was given.

The three-decision structure is what makes a pursuit reviewable at all. Each decision has an owner, a moment, and stated criteria, which means each can be located in the record and measured against policy. A pursuit case is rarely about whether pursuing is ever permissible; it is about whether these three decisions were made, by the people the policy assigns them to, on the facts available at the time.

Pursuits are among the best-documented events in policing, which is what makes them tractable for practices analysis. The dispatch system logs units and times. The radio carries the pursuit as it happened, including what the supervisor said and did not say. Dash and airborne video carry speeds, spacing and traffic in frame. And the agency’s own after-action review states what the agency itself concluded.

• Computer-aided dispatch entries fixing units, times and status changes
• Radio traffic, including supervisory acknowledgments, questions and orders
• Dash and airborne video, with speeds, spacing and traffic conditions in frame
• Body-worn video from the termination point forward
• The involved officers’ pursuit reports and any supervisor’s report
• The agency’s own pursuit review and its findings
• The pursuit policy in force that day, and training records on it

Scott v. Harris, 550 U.S. 372 (2007) placed recorded video at the center of pursuit litigation, and the practices method matches that reality: where the record contains video, the analysis is anchored to it. Speeds, spacing, traffic and intervals are taken from the frames and the clocks, and the narrative reports are then measured against the footage rather than the other way around.

The agency’s own review matters twice. Its findings are evidence of what the agency concluded when it looked at its own pursuit, and its silences are evidence of what the agency chose not to examine. An after-action review that never addresses the continuation decisions or the supervisor’s role says something about supervision and training before anyone offers an opinion.

Reading that record is detail work. Speeds are taken where the record states them or shows them, not estimated to fill gaps. Intervals between radio transmissions are measured, because the space between a question and its answer is often where a continuation decision lived. Where the record is genuinely silent on a condition the policy made relevant, the analysis says the record is silent rather than supplying an assumption.

The record request drives the engagement. The dispatch data, the full radio traffic, every camera, the pursuit reports, the policy in force on the date, the training records on that policy, and the agency’s own review, in whatever form it took, are the working file. Gaps in that set are themselves findings: a pursuit with no supervisor’s report, or a review that was never conducted, is a fact about the agency’s practice.

Professor Gilbertson has consulted in more than 220 civil rights and criminal cases across 42 states and the District of Columbia, retained by plaintiff’s counsel, by the criminal defense bar, and by counsel defending officers and agencies. On a pursuit question the opinion addresses practices: what the policy required at each decision point, what the record shows was done, and where they diverge. Whether the pursuit gives rise to liability is a legal conclusion, and legal conclusions remain counsel’s.

For a first inquiry, the posture and what the record already holds are enough: whether the dispatch data, radio and video have been produced, and whether the agency’s own pursuit review exists. Nothing privileged is needed to scope the engagement.