Policing Needs an Evidence-Based Definition of De-Escalation

Controversies over high profile police shootings have led to a demand for de-escalation in policing. Unfortunately, in their hurry to avoid becoming the “next Ferguson,”many agencies rushed to implement new de-escalation training programs. This is a problem because there is currently no evidence surrounding de-escalation in policing and no consensus on what the term even means. Further, reforms implemented quickly without thought to how they fit within the organizational structure of the agency, or whether they will effectively address the community’s concerns, can actually cause unanticipated consequences.

Many officers, for example, view calls for de-escalation as demands that they hesitate when they really should use force. They believe this expectation will put them and others at serious risk. In public rhetoric, the term “de-escalation” seems to be a buzzword referring to any progressive tactic producing a more positive or less violent outcome. This ambiguity creates a challenge for researchers seeking to generate evidence on the topic, and for agencies balancing the requests of the public with the practical realities of their work.

A few efforts have been made to create an evidence-based definition of de-escalation. The IACP (2017)issued a National Consensus Policy on Use of Force, which viewed de-escalation as both an overarching philosophy for how the police should approach potentially violent situations, and a collection of tactics for handling situations using the least amount of force. De-escalation was conceptualized as follows:

Taking action or communicating verbally or non-verbally during a potential force encounter in an attempt to stabilize the situation and reduce the immediacy of the threat so that more time, options, and resources can be called upon to resolve the situation without the use of force or with a reduction in the force necessary. De-escalation may include the use of such techniques as command presence, advisements, warnings, verbal persuasion, and tactical repositioning (IACP, 2017, p. 2).

Based on a systematic review of de-escalation training evaluations in professions outside the criminal justice field, Engel and colleagues (2019)suggested the following definition: “[the] prevention or management of clients’ violence, aggression, agitation, or similar behaviors, based on a process designed to defuse situations and reduce the likelihood of physical or verbal confrontation between parties.”

Based on interviews with veteran officers, my colleague Dr. Michael White and I defined de-escalation as bringing a situation or citizen in crisis back to a calm state, or preventing a situation from escalating into a conflict, using the least amount of force possible (Todak & White, forthcoming). Most of the tactics described by the officers in our study were verbal – talking to the citizen as a human being and not a suspect, listening to their side of the story, engaging them in decision making, compromising in minor situations, and showing respectfulness and empathy (putting yourself in that person’s shoes).

These studies are just the beginning of the conversation, and a number of questions still need to be answered.

  • What is de-escalation in policing?
  • What are the goals of de-escalation for encounters, agencies, and communities?
  • How is de-escalation performed (i.e. what are its tactics)?
  • In what situations should these tactics be used, and when should they not be used (i.e. in what situations are these tactics unsafe or ineffective)?

More work needs to be done to reach an evidence-based definition of de-escalation. A range of stakeholders should be involved in the conversation, including experienced police officers, administrators and training personnel, members of civilian oversight boards, police experts, and the public, to name a few. And of course, once the evidence begins to accumulate and evidence-based training programs are developed, they should be rigorously evaluated in collaboration with trained researchers. Despite the good intentions of de-escalation reform proponents, and the efforts made by police agencies to make moves to reduce unnecessary use of force by their patrol officers, it is absolutely essential that these changes be grounded in evidence, and that training programs be rigorously evaluated to ensure it does not put officer and citizen lives at risk.


By Natalie Todak

This blog post was adapted from:

Todak, Natalie& Madison March. “De-escalation in policing: Preliminary answers to key questions.” McLean, Alpert and Dunham (Eds.) Critical Issues in Policing, 8thEdition.



Engel, R. S., McManus, H. D., & Herold, T. D. (2019). The deafening demand for de-escalation training: A systematic review and call for evidence in police use of force reform[Unpublished Manuscript]. Cincinnati, OH: IACP / UC Center for Police Research and Policy.

IACP. (2017). National consensus policy on use of force. Alexandria, VA: Contributing Organizations.

Todak, N., & White, M. D. (forthcoming). Expert officer perceptions of de-escalation in policing. Policing: An International Journal.


2 thoughts to “Policing Needs an Evidence-Based Definition of De-Escalation”

  1. There are so many things to discuss with de-escalation and perhaps we should start putting quotations marks around the term like we do others like “best practices” because the term is so ill-defined. Part of the problem with de-escalation is that it appears to be more of a conceptualization of an end state.

    The National Consensus Policy on Use of Force, or at least some sort of consensus among the eleven participating organizations, fails to provide a cogent definition. As you have outlined in your fourth paragraph, the definition mentions actions, communication, and reducing the immediacy of the threat along with some techniques. It is not a definition as you have stated but is instead just a grouping of ideas and many of those are general ideas at best. The suggested definition in the fifth paragraph is obviously clinical. Care should be taken with clinical definitions because police work is rarely clinical.

    As for your proposed definition, it makes sense. However, I would offer that the objective is to bring a situation into a greater level of control or at least move the situation to a pre-crisis state. As I have told firefighters and paramedics coming in to handle shooting and stabbing calls, calm and controlled are not synonyms. The arrival of enough police resources may not lead to calm but can lead to an ability to control the crisis and continue communicating with involved parties so that emotional levels can be reduced.

    I concur that a better definition is needed for “de-escalation” and that such definition should be evidence-based. However, as those in both ASEBP and CAN-SEBP have frequently noted, police training is often poorly evaluated and lightly researched. If we don’t know what training our current “experts” in de-escalation have had or how that training was evaluated, then we must go back to ground zero for the definition. I submit that most evaluation of training stops at Kirkpatrick’s Level Two (Learning) and usually consists of a post-test. As such, we often use field experts to tell us what de-escalation means even though they have a tenuous grasp of the definition. What they do know is what tactics or techniques have been successful for them in reducing a state of crisis. As such, the definition of de-escalation is mired in an anecdotal state.

    I submit that the first place to start with a new definition for de-escalation is communication. Communication appears to be a key element of de-escalation if not the lynchpin. However, we also need to understand that communication is not something that is easily bifurcated into verbal and non-verbal. Verbal, vocal, and visual (non-verbal) communication aspects are inextricably intertwined and failing to align those aspects is often the cause of escalation.

  2. I believe the challenge in this endeavor is the criterion problem. We can measure training programs at Kirkpatrick’s first two levels through end of course critiques and pre and post testing of course content. Kirkpatrick’s third level, changed behavior in an operational setting, is almost never measured. Currently a research endeavor is underway with the Louiseville Metro PD that seems to skip level three and seeks to measure the fourth level, organizational change/impact. The problem I see with not measuring level three change is it increases the likelihood that a confounding variable influences the organizational change/impact.

    So what criterion should we define and possibly measure? The use of force continuum may be out of favor with some but I believe it is an accurate taxonomy of police coercive influence ranging from mere presence to the use of deadly force. De-escalation seems to have emerged as a response to use of force that was lawful but may not have been necessary had other alternative means of influence been utilized earlier in a confrontation. An officer should use the least coercive influence safely possible in order to achieve a legitimate police objective. De-escalation training is focused on enabling officers to move along that continuum if safe.

    We must also consider the context in which coercive influence is utilized. The suspect always gets a vote in the outcome. Officers are essentially adapting to a suspects’ actions. I believe we should explore adaptability as the underlying construct as a criterion we seek to define and measure. At the end of the day we seek to influence cognitive and physical performance under stress – the ability to adapt under stress. There is a very robust body of literature on cognitive and physical performance under stress. The US Army Special Forces has devoted research and training effort to improve adaptability of Special Forces soldiers.

    Without being able to define and measure the behavior we seek to change in an operational setting we cannot establish a causal chain from training to organizational change/impact. The only people who think outside the box are people outside the box. I believe we need to look in many other research boxes outside our own.

    IMHO for what it’s worth.

Leave a Reply