Use of Force: From Continuum to Reasonableness

---

Table of Contents

1. Why This Guide Is Framed Around a Question
2. The Use-of-Force Continuum and Why It Is Fading
3. Graham v. Connor and the Reasonableness Standard
4. Tennessee v. Garner and Deadly Force
5. Beyond the Constitutional Floor: Policy, Proportionality, and the Sanctity of Life
6. The Critical Decision-Making Model
7. Officer Training Implications
8. Documentation: Articulation, Not Conclusions
9. Common Evaluator Mistakes
10. Quick Reference Summary
11. Bibliography

---

1. Why This Guide Is Framed Around a Question

For decades, the central training question about use of force was structural: "where on the continuum am I, and what level am I authorized to move to?" That question is being replaced by a different one: "is the force I am about to use reasonable, necessary, and proportional in this specific situation, and can I articulate why?"

That shift is the subject of this guide. It is not a cosmetic change in vocabulary. It changes what officers are trained to do, how force is documented, and how reviewers evaluate it. An officer who learned policing through the continuum and an officer who learned it through the reasonableness standard will think differently in the moment and write differently afterward.

This guide explains the older continuum model and why the field is moving away from it, walks through the two Supreme Court decisions that anchor American use-of-force law, explains how modern policy builds above that legal floor, and then turns to the practical consequences for training, documentation, and review. It is written for working officers and for the training coordinators and reviewers who evaluate them. The throughline is articulation: in the modern model, the officer's ability to explain the decision is inseparable from the decision itself.

Pull quote. "The 'reasonableness' of a particular use of force must be judged from the perspective of a reasonable officer on the scene, rather than with the 20/20 vision of hindsight." Graham v. Connor, 490 U.S. 386 (1989).

2. The Use-of-Force Continuum and Why It Is Fading

2.1 What the continuum was

The use-of-force continuum, in its many agency-specific versions, was a model that arranged force options along a scale, typically from officer presence and verbal direction, through soft and hard empty-hand control, less-lethal tools such as chemical agents and conducted-energy weapons, up to deadly force. It was usually paired with a parallel scale of subject resistance, from compliant, through passive resistance, active resistance, and assaultive behavior, up to behavior posing a risk of death or serious injury. The officer was taught to match their force level to the subject's resistance level, often "one level above" to maintain control.

The continuum was not foolish. It was a genuine attempt to give officers a teachable structure and to give reviewers a consistent yardstick. For a long time it was the dominant model in American policing, and many agencies still use a version of it.

2.2 Why the field is moving away from it

Several criticisms accumulated, and the leading professional bodies came to share them.

It implies a staircase that real encounters do not follow. A continuum suggests force escalates and de-escalates one step at a time. Real encounters jump. A compliant subject can become a deadly threat instantly, and a violent subject can comply instantly. An officer trained to "go up one level" can be both too slow when a situation collapses and too committed to escalation when it does not.

It can read as permission rather than constraint. A continuum that pairs resistance levels with authorized force levels can be understood, and used in reports, as "the subject did X, therefore I was authorized to do Y." That framing skips the actual legal question, which is whether the force was reasonable in the totality of the circumstances, not whether the subject's conduct unlocked a tier.

It underweights time, distance, communication, and de-escalation. Classic continuums describe what an officer may do once force is contemplated. They are quieter about everything an officer can do earlier to make force unnecessary. The modern emphasis on de-escalation does not fit cleanly on a force ladder.

It does not match the legal standard. The constitutional test, set out in Section 3, is reasonableness under the totality of the circumstances. It is not a tiered matrix. A model that trains officers to think in tiers trains them in a structure the law does not actually use.

For these reasons, organizations including the Police Executive Research Forum have recommended moving away from rigid continuums toward decision-making models centered on reasonableness, necessity, and proportionality (PERF, Guiding Principles on Use of Force, 2016).

2.3 What is replacing it

The continuum is being replaced not by a different ladder but by a different kind of model: a decision-making process. Instead of "what tier am I on," the officer is trained to continuously ask what is happening, what is the threat, what are my options, and what is reasonable, necessary, and proportional right now. The Critical Decision-Making Model in Section 6 is the most widely adopted version. The point of the shift is to train judgment, not lookup.

Quick reference: continuum vs. decision model. A continuum asks "what level of force is the subject's resistance worth?" A decision model asks "given everything I know right now, what response is reasonable, necessary, and proportional, and what can I do to change the situation?" The first is a matrix. The second is a process. The law uses the second.

A caution for training coordinators: many agencies are mid-transition, with continuum language still in policy, in field training, and in the muscle memory of senior officers, while newer material uses decision-model language. Officers should be told plainly which model their agency's policy uses, because policy, not a guide, governs.

3. Graham v. Connor and the Reasonableness Standard

Almost every use-of-force conversation in American policing eventually reaches Graham v. Connor. Officers should understand it well enough to think with it, not just recite it.

3.1 The case

Dethorne Graham was a man with diabetes who, during a developing insulin reaction, asked a friend to drive him to a store for orange juice. His behavior at and around the store led to an investigative stop and a physical encounter with officers, during which he sustained injuries. He sued, alleging excessive force. The case reached the Supreme Court, which decided it in 1989. The legal importance of Graham is not its facts but the standard the Court announced for judging police use of force (Graham v. Connor, 490 U.S. 386 (1989)).

3.2 The Fourth Amendment and objective reasonableness

The Court held that claims of excessive force during an arrest, investigatory stop, or other seizure of a free person are properly analyzed under the Fourth Amendment's "objective reasonableness" standard. The key word is objective. The question is not whether the officer had good intentions or bad intentions. As the Court put it, an officer's evil intentions will not make a constitutional violation out of an objectively reasonable use of force, and an officer's good intentions will not make a constitutional violation reasonable. The officer's underlying motivation, good or bad, is not the test.

3.3 The perspective and the totality

Two further holdings shape how the standard is applied.

First, reasonableness is judged from the perspective of a reasonable officer on the scene, not with the 20/20 vision of hindsight. The Court explicitly recognized that officers are often forced to make split-second judgments, in circumstances that are tense, uncertain, and rapidly evolving, about the amount of force necessary in a particular situation.

Second, reasonableness is assessed under the totality of the circumstances. There is no mechanical formula. The Court offered factors, discussed next, but it framed them as considerations within a totality analysis, not as a checklist that resolves the question.

Pull quote. "The calculus of reasonableness must embody allowance for the fact that police officers are often forced to make split-second judgments, in circumstances that are tense, uncertain, and rapidly evolving, about the amount of force that is necessary in a particular situation." Graham v. Connor, 1989.

3.4 The Graham factors

The Court identified factors relevant to the reasonableness inquiry, commonly summarized as:

1. The severity of the crime at issue.
2. Whether the suspect poses an immediate threat to the safety of the officers or others.
3. Whether the suspect is actively resisting arrest or attempting to evade arrest by flight.

Three points officers and evaluators consistently get wrong about these factors:

• They are non-exclusive. Graham did not say these are the only things that matter. Courts since have considered many additional circumstances. The factors are a starting framework, not a closed list.
• The second factor usually dominates. The immediate-threat factor is, in most analyses, the heaviest. A minor crime with a genuine immediate threat is a very different situation from a serious crime with no current threat.
• They describe a moment, and the moment can change. Reasonableness is assessed at the time force is used. A use of force that was reasonable at one instant can become unreasonable seconds later when the threat ends, and continuing force after a threat has ended is a recurring problem in review.

3.5 What Graham is and is not

Graham sets a constitutional floor. It is the minimum standard below which force becomes a constitutional violation. It is not a description of good policing, and it is not the ceiling. Agency policy, state law, and professional standards can and frequently do require more restraint than Graham alone would permit. An officer whose only mental model is "was it reasonable under Graham" is using the floor as if it were the whole building. Section 5 addresses what sits above the floor.

CodeBlu's scenario library is designed to develop the judgment Graham assumes officers will exercise; see for example the traffic stop with escalation risk scenario, which places officers in exactly the kind of evolving, uncertain situation the reasonableness standard contemplates.

3.6 The Relevant Timeline After Barnes v. Felix (2025)

For decades, one question about the reasonableness inquiry went unresolved: when a court judges a use of force, how far back does it look? Only the final instant, the split second when the officer perceived a threat and acted? Or the whole encounter, including the officer's own earlier choices that helped shape that final instant?

Several federal circuits, including the Fifth Circuit, had adopted a "moment of threat" rule. Under that rule, the reasonableness analysis was confined to the narrow window in which the threat was immediately present. Whatever the officer did before that window, however questionable, simply was not part of the legal picture.

The Supreme Court rejected that rule in Barnes v. Felix, 605 U.S. ___, 145 S. Ct. 1353 (2025), decided May 15, 2025. The case arose from a traffic stop for suspected toll violations. The officer stepped onto the doorsill of the car as it began to move and, within roughly two seconds, fired into the vehicle, killing the driver. The lower courts, applying the moment-of-threat rule, examined only those final two seconds and found the force reasonable on that basis.

A unanimous Court, in an opinion by Justice Kagan, vacated that decision. It held that the Fourth Amendment's reasonableness inquiry cannot be cabined to a single slice of time. Reasonableness is judged on the totality of the circumstances, and a court cannot review that totality, as the Court put it, if it has "put on chronological blinders." The events and decisions leading up to the use of force are part of the picture a court must consider.

Pull quote. A court cannot review "the totality of the circumstances" if it has "put on chronological blinders." Barnes v. Felix, 2025.

It is important to be precise about what Barnes did and did not do. It did not hold that earlier officer conduct automatically makes a later use of force unlawful, and it did not adopt any particular "officer-created jeopardy" rule. It held something narrower: a court may not exclude the earlier conduct from view. How much weight that conduct carries remains a totality question, decided case by case, and a separate concurrence flagged that related questions about pre-seizure conduct are still unsettled.

Barnes does not change the Graham standard. It clarifies how widely that standard's lens must open. For officers, the practical lesson reinforces what sound training already teaches: the decisions made in the minutes before force, positioning, pace, communication, the choice to create or close distance, are not legally invisible. They are part of the same encounter a court will later review. This is why the full-timeline emphasis in Sections 7 and 8, and the "collapsing the timeline" evaluator error in Section 9, are now aligned with constitutional law, not merely with best practice. For a fuller treatment, see the companion case-law deep dive on Graham v. Connor (Graham v. Connor), which discusses Barnes in the context of the Graham framework.

4. Tennessee v. Garner and Deadly Force

Where Graham governs force generally, Tennessee v. Garner governs the most serious category: deadly force.

4.1 The case and the holding

Garner arose from the police shooting of a fleeing, unarmed teenager suspected of a residential burglary. At the time, many jurisdictions followed a "fleeing felon" rule, allowing deadly force to stop the escape of any felony suspect. In 1985, the Supreme Court held that rule unconstitutional as applied (Tennessee v. Garner, 471 U.S. 1 (1985)).

The Court held that deadly force may not be used to prevent the escape of a fleeing suspect unless it is necessary to prevent the escape and the officer has probable cause to believe that the suspect poses a significant threat of death or serious physical injury to the officer or others. The Court also indicated that, where feasible, some warning should be given before deadly force is used.

Pull quote. "Where the officer has probable cause to believe that the suspect poses a threat of serious physical harm, either to the officer or to others, it is not constitutionally unreasonable to prevent escape by using deadly force." Tennessee v. Garner, 1985.

4.2 What Garner changed

Garner moved the legal center of deadly force from the crime to the threat. After Garner, the question is not "is this person a felon who is getting away," it is "does this person pose a significant threat of death or serious physical injury." A fleeing suspect who poses no such threat may not be stopped with deadly force, however serious the underlying offense.

It is worth being precise about the relationship between the two cases. Garner and Graham are read together. Garner is generally understood as an application of the same Fourth Amendment reasonableness analysis to the deadly-force situation, not as a separate, rigid rule that displaces the totality-of-circumstances inquiry. Courts analyze deadly-force cases under Graham's reasonableness framework, informed by Garner's recognition that the threat of serious harm is what justifies deadly force.

4.3 The officer-level takeaway

For an officer, the durable lesson of Garner is short: deadly force is about stopping an imminent threat of death or serious physical injury, not about stopping a person or punishing a crime. The escape of a dangerous person is a serious problem, and it is not, by itself, a deadly-force justification. As with Graham, agency policy and state law frequently impose additional restrictions, and several states have enacted use-of-force statutes that are more restrictive than the constitutional minimum.

5. Beyond the Constitutional Floor: Policy, Proportionality, and the Sanctity of Life

A persistent training error is to treat Graham and Garner as the whole of use-of-force doctrine. They are the constitutional floor. Modern policy builds above it, and officers are evaluated against the building, not the floor.

5.1 Why the floor is not enough

"Was it constitutional" and "was it good policing" are different questions. A use of force can clear the constitutional minimum and still violate agency policy, still reflect poor tactics, still have been avoidable, and still damage community trust. Leading professional bodies have argued for years that agencies should hold officers to standards above the constitutional minimum (PERF, 2016; IACP National Consensus Policy). The reasonableness standard tells a court whether to impose liability. It does not tell an agency what to expect of its officers.

5.2 Necessity and proportionality

Two concepts feature heavily in modern policy.

Necessity asks whether the force was needed at all, and whether a lesser response, or no force, would have achieved a lawful objective. A response can be "reasonable" in the Graham sense and still not have been necessary if a better option was available and feasible.

Proportionality asks whether the force used was proportionate to the threat and the seriousness of the situation, considering the totality of the circumstances. PERF has framed proportionality partly as a question officers should be able to answer afterward: would this use of force look proportionate to the public, and can the officer explain why it was.

5.3 The sanctity of human life

Modern model policies often open with a sanctity-of-life principle: that the protection of human life, including the life of the subject, is a primary value, and that force should be used only when necessary and only to the degree necessary. This is not sentiment. It reframes the officer's objective from "gain compliance" to "resolve this safely for everyone, including the person I am dealing with." That reframing changes behavior upstream of any force decision.

5.4 The duty to intervene and the duty to render aid

Two duties now appear in most modern use-of-force policies and in many state laws:

• The duty to intervene: an officer who observes another officer using force that is clearly beyond what is reasonable or lawful has an affirmative obligation to intervene and to report. Programs such as the Georgetown Law-based ABLE Project train the bystandership skills this duty requires.
• The duty to render aid: once a scene is safe, officers have an obligation to provide or summon appropriate medical aid for an injured subject.

5.5 De-escalation as a policy expectation

Finally, modern policy increasingly treats de-escalation as an expectation that is itself reviewable: did the officer, when it was safe and feasible, use time, distance, communication, and tactical positioning to try to make force unnecessary. This is the link between this guide and Guide 1. De-escalation is not separate from use of force. It is the part of the use-of-force timeline that happens before force, and a force review now routinely examines it.

Quick reference: the floor and the building. Graham and Garner are the constitutional floor: reasonableness and, for deadly force, an imminent threat of death or serious injury. Above the floor sit necessity, proportionality, sanctity of life, the duties to intervene and render aid, and a de-escalation expectation. Officers are trained and reviewed against the whole building. Agency policy and state law define its height.

6. The Critical Decision-Making Model

If the continuum is being retired, something has to take its place as the teachable structure. The most widely adopted replacement is the Critical Decision-Making Model (CDM), associated with PERF and built into the ICAT training guide.

6.1 The model

The CDM is a cyclical, repeating process rather than a ladder. Its steps are usually described as:

1. Collect information. What is actually happening? What do I know, what do I not know, what am I assuming?
2. Assess the situation, threats, and risks. How serious and how immediate is the threat? To whom?
3. Consider police authority and agency policy. What am I lawfully and procedurally permitted and required to do here?
4. Identify options and determine the best course of action. What are my choices, including doing nothing yet, repositioning, or waiting for resources?
5. Act, review, and reassess. Take the action, then return to step one, because the situation has changed.

At the center of most depictions of the model sits a core value, typically the sanctity of human life, which every step is meant to serve.

6.2 Why a cycle beats a ladder

The CDM's strength is that it matches how encounters actually behave. Encounters are not monotonic. They change, often fast, and the cycle structure trains the officer to keep re-collecting information and reassessing rather than committing to a path chosen ten seconds and one reality ago. It also explicitly creates room for the options that a ladder hides: waiting, repositioning, calling for resources, and trying again with communication. Those are choices, and the model names them as choices.

6.3 The honest limits

The CDM is a thinking tool, and it cannot run at full deliberative speed in a genuine split-second event. Its defenders would say that is the point: most encounters are not genuine split-second events, and the model exists to keep officers from compressing slow situations into fast ones. But a guide should be candid: the model is most powerful in the large middle range of encounters that have time in them. In a true ambush, training and tactics, not a five-step cycle, carry the moment.

Pull quote. "Slowing down a situation, when feasible, gives officers more time, more options, and better outcomes. Time is on the officer's side more often than it feels like it is." A synthesis of the ICAT and CDM rationale.

7. Officer Training Implications

Moving from a continuum to a reasonableness-and-decision model changes training in concrete ways.

Teach the law as a thinking tool, not a recitation. Officers should be able to apply the Graham factors and the Garner threat standard to a messy, ambiguous scenario, not just recite them. Memorizing three factors is not the same as analyzing a totality.

Train the decision process under stress. The CDM only helps if it has been practiced enough to run when adrenaline is high. That requires repeated scenario practice with realistic stress, not classroom walk-throughs. This is the central argument of Guide 6.

Integrate de-escalation and force training. If de-escalation is part of the force timeline, it should not be a separate course taught by separate instructors. The same scenario that ends in a force decision should begin with the de-escalation window.

Train the reassessment habit. Because reasonableness is assessed moment by moment, the most important trained reflex may be noticing when a threat has ended and stopping. "Continuing force after the threat is over" is one of the most common findings in adverse force reviews, and it is trainable.

Train articulation as a skill. In a reasonableness regime, an officer who acted well but cannot explain why is exposed. Articulation should be trained and practiced, not assumed. This connects directly to Section 8.

Use the team and the tools that buy time. Training should reinforce that calling for resources, using cover, and creating distance are not hesitation. They are tactically sound and they widen the decision space.

8. Documentation: Articulation, Not Conclusions

In the reasonableness model, the use-of-force report is not paperwork after the fact. It is the record of the officer's decision-making, and its quality often determines how the force is judged.

8.1 Facts and articulation, not labels

The recurring documentation failure is the conclusory report: "the subject was actively resisting, so I used a control technique." That sentence states conclusions and legal categories. It does not show the reader what the officer saw and why it mattered. A reviewer cannot assess a conclusion.

The better report describes the specific, observable facts and then articulates the officer's reasoning from them. Not "the subject was aggressive," but what the subject did, said, and how they moved, the officer's distance and positioning, what the officer knew at that moment, and the inference the officer drew. The legal standard is built on the totality of the circumstances as the officer reasonably perceived them, so the report has to actually contain those circumstances and that perception.

8.2 Document the perception, including what turned out wrong

Because reasonableness is judged on what a reasonable officer on the scene perceived, not on hindsight, the report should capture what the officer perceived and believed at the time, including a perception that later proved mistaken. An officer who reasonably perceived a threat that turned out not to exist should document that perception and its basis. Rewriting the report to match what was later learned is both dishonest and, paradoxically, legally weaker, because it abandons the standard that actually protects a reasonable officer.

8.3 Document the whole timeline

A good report does not start at the moment of force. It includes the de-escalation window: what the officer did with time, distance, communication, and positioning before force, and why the situation moved past those options. In a modern review, the absence of that part of the story is itself conspicuous.

8.4 Write it so a stranger understands it

The audience for a use-of-force report is not only the officer's sergeant. It may be a reviewer, an attorney, a jury, or the public, none of whom were there. Acronyms, undefined jargon, and insider shorthand fail that audience. The discipline is to write so that a reasonable person who was not present can follow what happened and why each decision made sense at the time.

Quick reference: documentation. Describe observable facts, not conclusions or legal labels. Articulate the reasoning from facts to decision. Record what you perceived at the time, including mistaken perceptions, not hindsight. Include the de-escalation window. Write so a stranger can follow it. The report is the decision, preserved.

9. Common Evaluator Mistakes

The reasonableness model puts heavy weight on the people who review force. Reviewers, supervisors, training officers, and use-of-force board members, make recurring errors, and a guide a chief would share should name them.

Hindsight bias. The single most common error. Knowing how the encounter ended, especially knowing the subject was unarmed or in crisis, contaminates the assessment of what a reasonable officer should have perceived in the moment. Graham explicitly forbids this, and reviewers still do it. Disciplined review reconstructs what was knowable at each decision point and assesses the decision against that.

Outcome bias. Closely related but distinct: judging the decision by how badly or well it turned out rather than by its quality at the time. A sound decision can have a tragic outcome, and a poor decision can have a lucky one. Reviewing only outcomes teaches officers nothing and rewards luck.

Collapsing the timeline. Evaluating only the final frame, the moment of force, and ignoring the minutes before it. A modern review examines whether earlier officer decisions reasonably created or worsened the final situation, while being careful not to slide into hindsight.

Treating reasonable as a synonym for ideal. The legal question is whether the force was within the range of reasonable responses, not whether it was the single best possible response. Reviewers sometimes fault an officer for not choosing the option that, in calm hindsight, looks optimal. The standard does not require optimal. Conversely, "legally reasonable" is the floor, and policy review can hold officers to more, as long as the reviewer is clear about which question is being asked.

Confusing the questions. A review should be explicit about whether it is asking "was this lawful," "was this within policy," "was this good tactics," or "was this avoidable." These are four questions with four possible answers. A review that blends them produces findings nobody can learn from.

Ignoring articulation gaps as if they were neutral. If an officer cannot explain a decision, that is itself a finding, and usually a training finding. Treating an unexplained decision as automatically fine, or automatically misconduct, both miss the point. The constructive response is usually targeted training.

Inconsistency across reviewers. Without a shared rubric, two reviewers reach two conclusions on the same incident. This is unfair to officers and useless for organizational learning. Structured, criteria-based review, the same logic CodeBlu applies to scenario after-action review, is the corrective. See Guide 6 on assessment.

Pull quote. "The test of reasonableness is not capable of precise definition or mechanical application." Graham v. Connor, 1989. The same caution applies to those who review reasonableness afterward.

10. Quick Reference Summary

The one-page version.

The continuum is fading. It implied a staircase real encounters do not climb, read as permission, underweighted de-escalation, and did not match the legal standard. Decision-making models are replacing it.

Graham v. Connor (1989). Force is judged by objective reasonableness under the Fourth Amendment, from the perspective of a reasonable officer on the scene, not hindsight, under the totality of the circumstances. Factors: severity of the crime, immediate threat, active resistance or flight. The immediate-threat factor usually dominates.

Tennessee v. Garner (1985). Deadly force requires probable cause that the suspect poses a significant threat of death or serious physical injury. The crime alone is not the justification. The threat is.

The floor is not the building. Above the constitutional minimum sit necessity, proportionality, sanctity of life, the duties to intervene and render aid, and a de-escalation expectation. Agency policy and state law often require more restraint.

The Critical Decision-Making Model replaces the ladder with a cycle: collect information, assess threats, consider authority, identify options, act and reassess, around a core value of preserving life.

Documentation is the decision, preserved. Facts not conclusions, articulated reasoning, perception at the time including mistaken perception, the full timeline, written so a stranger understands it.

Evaluators must avoid hindsight bias, outcome bias, collapsing the timeline, confusing "reasonable" with "ideal," blending the lawful/policy/tactics/avoidable questions, and inconsistency across reviewers.

11. Bibliography

Court opinions are public record. Legal statements require review by qualified counsel; this guide states the federal constitutional floor and does not address jurisdiction-specific statutes.

• Graham v. Connor, 490 U.S. 386 (1989). Public record.

• Tennessee v. Garner, 471 U.S. 1 (1985). Public record.

• Scott v. Harris, 550 U.S. 372 (2007). Public record.

• Kingsley v. Hendrickson, 576 U.S. 389 (2015). Public record.

• Police Executive Research Forum. Guiding Principles on Use of Force. PERF, 2016.

• Police Executive Research Forum. ICAT: Integrating Communications, Assessment, and Tactics. A Training Guide for Defusing Critical Incidents. PERF, 2016.

• International Association of Chiefs of Police. National Consensus Policy and Discussion Paper on Use of Force. 2017 and subsequent revisions.

• President's Task Force on 21st Century Policing. Final Report. COPS Office, 2015.

• ABLE Project (Active Bystandership for Law Enforcement), Georgetown University Law Center.

Related CodeBlu guides: The Modern Officer's Guide to De-Escalation | Mental Health Crisis Response | Building Better Training

Related CodeBlu scenarios: Traffic Stop with Escalation Risk | Aggressive Intoxicated Person Outside a Bar | Disorderly Conduct in a Public Park