The Invisible Rules That Govern Police Use of Force

As calls for reform intensify in the aftermath of police violence, police departments can be expected to fiercely resist proposed changes that look anything like rules. I recently published a paper that shares the responses of officers who were asked the question: “Can use-of-force practices be reduced to hard and fast rules that limit office discretion?” The response, to summarize, was an emphatic no. Those interviewed thought that hard and fast rules would undermine the discretion officers need to confront unpredictable and often dangerous circumstances. The predominant view was that rules were not just ill advised in the use-of-force context but also incompatible with the demands of policing.

Contrary to these departmental claims, my research reveals that use-of-force rules not only exist but also are common. These rules are created by the very departments that deny their existence. Departments do not publish these rules, let alone write them down. Rather, during trainings, departments communicate hard and fast rules that limit officer discretion and disregard the relevance of factual variability. In reaching this conclusion, my research team reviewed use-of-force training videos used by agencies across the country. In our analysis of these videos, it became apparent that departments, however unknowingly, are imposing rules on their officers. The research also revealed that the nature of these rules, shielded from public scrutiny and input, often distributed the risks of potential (and unnecessary) harm to civilians.

If departments are in fact wrong that rules are nonexistent and incompatible with policing, what is the upshot? Accepting that rules have a place in regulating use of force expands the options for community intervention. Communities seeking use-of-force reform should feel empowered to disregard departmental resistance and insist on legislating rules that curb officer discretion. At a minimum, communities should require departments to transcribe their trainings. By doing so, the invisible rules—communicated during training and already governing police encounters—can be revealed to permit public scrutiny. The point is not that rules are good or bad. Rather,contrary to what departments say, rules exist, and given that rules exist, members of the public should have a say in creating them. At the very least, the public should be apprised of what rules already govern police use of force in their community.

The Rule-Resistant View

My research team interviewed representatives from Peace Officer Standards and Training agencies (POSTs). POSTs regulate officer training requirements in each state. Many representatives previously held leadership positions in police or sheriff departments. Almost all dismissed the possibility that use-of-force practices could be reduced to rules, and emphatically so: “Absolutely not” and “It’s impossible” were representative answers.

The most common rationale for the rule-resistant view was that a rule would undermine officer discretion needed to respond to fluid, unpredictable circumstances. “Every situation will have countless variables.” “You can have two really similar incidents lead to two totally different outcomes.” “Everything varies a million different ways.^” “It’s impossible to have a set of rules that will anticipate a particular situation.” “The only thing black and white about my job is my cruiser.” Representatives tended to view rules as flattening out the factual variability an officer confronts in the field. In their view, requiring officers to act a certain way would result in too little force (losing control) or too much force (causing unnecessary harm). One representative viewed any discussion over rule-making as pure hubris: “A lot of people try to come up with answers. You’re a prophet in your own backyard.”

Some representatives saw rules as unfair: “[S]ubjects don’t follow rules, so how can an officer be expected to do so?” There was concern that any rule would undermine the critical thinking necessary for deescalation: “Making rules puts the gun in [officers’] hands one step sooner.” Some saw any rule overrun by exceptions, obviating its utility: “Can you [really] have a rule that prohibits chokeholds? What if the suspect grabs a gun, and the only option is to put the person in a chokehold?” Some feared that removing officer discretion would be demoralizing: “You can’t legislate [reasonable force]. If we put rules and laws in … it is not gonna work …. No one will want to do this job.”

Very few were open to the imposition of rules. “You can’t say in every situation, if ... this, then do that. But you can have rules at a macro level[,]” said one. “I think we can have some successful rules of engagement. ... [I]f you can anticipate certain things happening, then you can make rules.” He was an outlier.

At the heart of the rule-resistant narrative is Graham v. Connor (1989), in which the Supreme Court interpreted the Fourth Amendment to require that the use of force be “reasonable.” The court held that reasonableness will depend on the “totality of the circumstances,” accounting for the officer’s incomplete information and the stressful, rapidly evolving nature of an incident. Graham is explicitly rule resistant, stating that an officer’s use of force can not be “governed by a single generic standard.” Importantly, the opinion was designed to establish the constitutional floor for officer adherence to the Fourth Amendment, not to deprive communities of the opportunity to impose a more restrictive code of conduct on their officers.

Yet, from hamlet to metropolis, Graham is the law, in large part due to the lobbying efforts of departments. As to the representatives interviewed, a recurring sentiment emerged: In the absence of rules, departments still followed the law, the law being Graham. “Stick to Graham,” were the concise words of one. As Rachel Harmon observed, the Graham standard is so vague it provides little functional limit on the use of force. It was Graham, for example, that permitted use-of-force expert Barry Brodd to sincerely assert that Officer Derek Chauvin, kneeling on a man’s neck for nine minutes, acted reasonably. Graham, contends John Gross, leaves the reasonableness of force “almost entirely up to the individual officer.” Put another way, the Graham standard permits officers to judge the reasonableness of force for the rest of us.

If Rules Exist, Where Are They?

For the vast majority of law enforcement representatives that we interviewed, use-of-force rules do not and cannot exist. They are wrong on both counts. My research reveals that a rule-based approach not only is possible but also is a widespread practice. In coming to this conclusion, my research team analyzed use-of-force videos produced over 10 years by Lexipol, the largest provider of police training in the United States. Lexipol trainers identified conditions that required officers to act a certain way. These are rules. From many examples, here is one with deadly consequences. A Lexipol trainer stated, “Officers cannot wait to react until they are absolutely certain of an individual’s malicious intent. If an officer waits to be certain that the individual is retrieving a weapon, the weapon could easily be used against the officer before he or she has an opportunity to respond.” Another trainer stated, “You see officers hesitate … [trainers] sitting back with perfect perspective [who know the officer will be harmed are thinking], shoot, shoot, shoot, shoot now ….” Pursuant to this training, if an officer perceives a civilian is reaching for a weapon, regardless of the circumstances, the officer is to use deadly force. Many jurisdictions in fact train this practice. I call it the “waistband rule,” as officers so trained will deploy deadly force if they perceive civilians reaching for their waistbands.

Like the waistband rule, rules indeed flatten out factual variability. Take an incident reported by a law student driving her car home from yoga. An officer checked her license plate. The car came up stolen. When this happens, training often dictates that officers call for backup, draw firearms, and order the civilian to walk backward toward officers and submit to cuffing. No exceptions. In the law student’s case, an officer pointed a long gun at her as she walked backward, terrified, toward the group of officers. The reported theft was two years old. But the record’s staleness had no relevance to officer conduct. Training deemed all facts irrelevant but for a single condition: Was the vehicle reported stolen? With this condition present, officers followed a rule that subjected the student to a serious risk of harm, regardless of the unique circumstances known to the officer that reduced any risk surrounding the stop.

From an operational perspective, it should be obvious that departments would impose rules. No employer would give unfettered discretion to employees, and no employees would want to go out in the field without rules of engagement. Despite what they say, departments are no different. And as departments impose rules during training, the stakes are high. Even “mundane police encounters,” contends Alice Ristroph, can quickly lead to violent results.

In fact, training can require an officer to use deadly force in a mundane encounter. Take the case of Paul Heenan, an inebriated man in need of assistance who was shot by Madison Police Officer Steven Heimsness. As locals wondered whether the officer committed murder, another question remained dormant: Did training require Heimsness to shoot Heenan? The Madison Police Department (MPD) provided the answer in a press conference, where trainers described how, according to MPD training, it would have been unreasonable not to have shot Heenan.

The incident occurred on Nov. 9, 2012. Heenan, new to the neighborhood, had joined his household, a young family, for dinner. After seeing some bands, he ended up at the Weary Traveler bar. Owner Bregan Fuller noticed Heenan was intoxicated when he staggered into the lady’s restroom. It took time to coax from Heenan his place of residence—Baldwin Street, a few blocks away—where Fuller dropped Heenan off around 2 a.m. Forty-five minutes later, Megan O’Malley, who lived on Baldwin Street, heard someone at the door. On the porch, her husband Kevin found Heenan, who, shivering, attempted to enter. Ascertaining Heenan was inebriated, Kevin assisted Heenan home, two doors down. Heenan, unaware of where he was, leaned into Kevin. With Heenan in tow, Kevin backed toward his house to get help. Kevin did not know Megan had called 911 to report a possible burglary. Dispatch radioed, “Husband’s wearing plaid pajama bottoms.” Officer Stephen Heimsness arrived at the incident with his firearm drawn. Only after Heimsness yelled, “Get down!” did Kevin realize the figure emerging from the shadows was an officer. As Kevin stepped back, Heenan lost his balance and flailed toward Heimsness. Kevin yelled, “He’s a neighbor!” Heimsness pushed Heenan away and shot him.

The MPD deemed Heimsness’s actions reasonable. Facing public skepticism, the MPD held a public presentation that offered a glimpse into MPD training. First, offering a big-picture view, trainers Jason Freedman and Kimba Tieu emphasized that officers, thrust into situations with “limited options” and “little information,” always abide by three “principles”: (a) retain the advantage, (b) hesitation kills (the officer), and (c) only the suspect chooses to escalate or deescalate a situation. Turning to the incident, the trainers provided answers to their rhetorical questions by acting out possible scenarios. Why did the officer approach with a lethal weapon? In accordance with MPD training, Tieu stated, “in high risk situations such as burglaries and home invasions where the threats are unknown,” “to start at a lower level of force” would potentially allow the suspect to harm the officer or public. “We don’t know the civilian’s intentions, their motivation, their training, the equipment that they carry, or if they’re fleeing from something heinous that they just did,” said Freedman. “Felonious situations will be a firearm day.”

The presentation continued. Freedman commanded, “Get down!” Tieu, playing the suspect, pulled a gun. Freedman reported an untrained person can produce a firearm and shoot the officer in less than a half-second. Tieu clarified that an unarmed civilian still presents a deadly threat. Returning to their places, Freedman said, “Get down!” Tieu lurched forward and grabbed the gun. What if the officer had attempted additional verbal commands to gain compliance? Resetting their positions, Freedman issued a first and then second command. Tieu aggressed. Freedman explained that “by the time my brain physically is able to process” noncompliance, “now he has my weapon, I am unarmed, and [there is a] threat to the community and to myself.” What if the officer had tried to holster his weapon? As Freedman attempted to do so, Tieu hit Freedman with a right hook. What if the officer attempted to step back? Freedman explained it is not possible to move back faster than the suspect moved forward. What if the officer had turned and run? As Freedman turned to retreat, Tieu put him in a headlock. What if the officer had not pushed Heenan away? As the two struggled, Tieu disarmed and shot Freedman. “We think about what it takes to disarm [the officer],” said Tieu, “nothing at all, I’m gonna rip the gun, and I’ll apply deadly force to the officer.”

Can rules be extracted from this impromptu training session? MPD training certainly limits officer discretion once certain conditions are present. Though the officer was responding to a mundane situation (a drunk person needing assistance), MPD training deemed it “high-risk,” which required the officer to draw his firearm. As Seth Stoughton observes, doing so creates a “path dependence, limiting an officer’s ability to use other force options.” This was certainly the case in Madison. When Heenan failed to comply with the command to get down and approached the officer, training dictated that Heimsness use deadly force. Are Madisonians in agreement with this rule that distributes risk to the civilian and increases the potential for unnecessary violence? We don’t know.

In sum, my research indicates that departments often impose rules on officers during training. Again, departments reject this contention. “If the [recruits] keep asking for rules,” remarked one trainer, “they should seek another profession.” “If [the recruit] asks me what to do in this specific case,” said another, “I can’t answer.” But trainers often do what they say they do not.

Looking Forward

What are the consequences of accepting the rule-resistant view? Under this view, public input is limited: Any proposal that interferes with officer discretion is rejected. Instead, a community is left proposing weak-tea standards, for example, that officers “attempt to use time and distance” approaching a scene. Such a standard leaves officer discretion intact; the officer determines what time and distance are appropriate in any situation. In this way, acceptance of the rule-resistant narrative deflects public intervention and waters down any attempted reform. Thus far, departments have had success convincing the public that rules cannot govern the use of force.

Part of the challenge for communities seeking change is to make visible the use-of-force rules that already exist but have long remained invisible. My research suggests communities that do so will discover rules that often distribute risk to civilians. The rule directing officers what to do when a car is reported stolen, for example, is fraught with danger and humiliation for the civilian—as this video shows. The rules followed in the Heenan incident ensured that any harm resulting from the officer’s misperception was to be borne by the civilian.

And the waistband rule trained by Lexipol—that an officer use deadly force when the officer perceives a civilian to be reaching for a weapon—has left a trail of preventable harm and death. In a Los Angeles study, of 100 shootings resulting from the waistband rule, 61 of these individuals had no weapon—the officer was wrong. Of these 100 individuals, then, at least 60 should not have been shot. Under an alternative rule that is not trained—(a) wait to confirm the presence of a weapon and (b) confirm the civilian’s intent to use the weapon—slightly more officers would be injured or killed, but the number of civilians spared from injury or death would be dramatically higher. Perhaps the people of Los Angeles would agree with the sheriff’s department that they prefer officers to follow the waistband rule. We do not know. The people of Los Angeles have no idea they are subject to this rule.

Use-of-force rules often lurk behind tragedies. When officers shot Jacob Blake for resisting arrest, shot Daunte Wright for evading misdemeanor arrest, shot Laquan McDonald for walking with a knife, shot Adam Toledo after he dropped his gun or shot 12-year-old Tamir Rice as he reached for his waistband, the officers were likely acting according to decision trees that, imparted through training, ironed out a need to assess the particulars of an incident and directed officers to act in ways that distributed risk to civilians.

Though departments say otherwise, use of force is often governed by rules that remain invisible to the public. These rules are not the product of compromise. Imposed during training, these rules reflect values that might be unacceptable to the public. The values and assumptions embedded in use-of-force rules deserve public scrutiny. The question, then, is not whether use of force can be governed by rules. The question should be this: Who gets to impose rules governing officer conduct: the police or the communities the police serve?