Rich Pedroncelli / AP

On a late-fall afternoon in 1984, in Charlotte, North Carolina, Dethorne Graham started to feel ill. He understood what was happening; insulin reactions from diabetes were a regular part of his life. Graham asked his friend William to drive him to a local convenience store where he could buy orange juice to offset the effects. Graham walked in, but left quickly after seeing a long line at the counter.

M. S. Connor, a Charlotte police officer, was nearby. Connor, an African American like Graham, thought it was odd that someone would enter and leave the store so quickly. The officer followed Graham and his friend for about a half mile in his squad car before pulling them over. After speaking briefly with the two men, Connor called for backup.

Graham got out of the car. As he wavered in and out of consciousness, he found himself handcuffed and lying on his stomach on the sidewalk. Graham tried to explain his medical condition, but the group of officers didn’t believe him and mistook an insulin reaction as him simply being drunk. The officers pulled him up from behind, slammed his head on the hood of his friend’s car, and pinned him facedown, with an officer leaning heavily on each limb. They then picked him up and threw him inside the patrol car. At one point, a friend familiar with Graham’s condition ran over to the scene with orange juice to help. Graham begged one of the officers for the drink. She responded: “I’m not giving you shit.”

After the clerk at the convenience store confirmed that nothing unusual had happened, the officers drove Graham home. He was left with a broken foot, several lacerations, and what he would later describe as a persistent ringing in his right ear. Little did Graham know as he writhed in pain that this episode would lead, five years later, to one of the most important U.S. Supreme Court decisions in modern history, Graham v. Connor. While the decision stemming from this incident is not well known, its influence has been far-reaching.

Since the beginning of the Black Lives Matter movement, in 2013, stories concerning police use of force have been prominent in the news and on social media. Much of the public conversation has focused on a collective exasperation: How is it that police can beat and kill men and women, many of them unarmed, yet rarely be held accountable?

The answer lies in large part in the 1989 Graham decision. Graham brought a federal claim against the Charlotte police officers, under a civil-rights statute called 42 U.S.C. §1983, in the U.S. District Court for the Western District of North Carolina. He argued that the excessive use of force against him violated substantive due process, or his right to be free from such abuse under the Fourteenth Amendment—one of the Reconstruction amendments ratified after the Civil War to give African Americans full legal equality with whites.

The trial court, as well as the Fourth Circuit Court of Appeals, sided with the officers. But, in deciding to review the case, the Supreme Court made a surprising move. Until that point, the legal standards through which federal courts reviewed claims of excessive force by state and local police were diverse. Many cases used substantive due process under the Fourteenth Amendment, following an earlier Second Circuit Court of Appeals decision in Johnson v. Glick, in which a detained man alleged that a correctional officer had assaulted him. This standard had been criticized, however, for emphasizing officers’ subjective mental state—that is, whether the force was applied in “good faith” or “maliciously and sadistically for the very purpose of causing harm.”

In the Graham decision, the Supreme Court held that substantive due process was not the applicable constitutional standard. Rather, the Court said the proper constitutional test was whether the action was “reasonable” under the Fourth Amendment, which prohibits unreasonable searches and seizures. (Use of force by the police during an arrest or investigative stop is understood to be a type of “seizure.”)

The choice was significant. Turning away from the Fourteenth Amendment as a constitutional standard would come to represent a missed opportunity to situate excessive force in minority communities as a long-standing structural problem. The Fourth Amendment was developed at a time when slavery was condoned by the Constitution, and it is largely preoccupied with the relationship between individuals and the government. The Fourteenth Amendment, on the other hand, has its roots in the post–Civil War effort to extend legal equality to former slaves. Particularly through its clause guaranteeing equal protection of the law, it reflects an awareness of how racial groups, not just individuals, can face state persecution. (Although the Fourteenth Amendment’s due-process clause is what allows the Fourth Amendment to apply to state and local police, as opposed to only the federal government, the Graham decision nonetheless represents a significant retreat from the Fourteenth Amendment’s original purpose of protecting racial minorities from state violence and other inequities.)

The Court’s decision to embrace a Fourth Amendment perspective that frames excessive force as an isolated interaction between police and individuals would impede federal courts’ ability to consider how race and racism can influence an officer’s decision to use force. To be sure, before Graham, substantive-due-process claims concerning police violence focused largely on individual liberty rather than structural conditions. But shifting the constitutional standard for excessive force away from the Fourteenth Amendment would prove to hinder courts’ ability to consider such abuse as a problem tied to issues of equal protection and racial subordination—in turn limiting the types of claims that victims of police violence could successfully bring. The disproportionate policing of racial minorities and the state-sanctioned violence that often ensues (performed by officers of all racial backgrounds) had been a dire problem in communities of color, and would continue to be. Decades after the Graham decision, research would show that black men are three times more likely to be killed by the police than white men. Police violence would come to be understood as a major public-health issue.   

Moreover, in articulating the new standard for judging excessive-force cases, the Graham Court didn’t do much to describe what “reasonable” means. William Rehnquist, the chief justice at the time, provided a few guideposts, such as paying “careful attention to the facts and circumstances of each particular case, including the severity of the crime at issue, whether the suspect poses an immediate threat to the safety of the officers or others, and whether he is actively resisting arrest or attempting to evade arrest by flight.” But he also wrote that reasonableness is difficult to describe, noting that “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.”

The vagueness of the standard for “what counts” as excessive force would have sweeping implications. Initially, some believed that its use would favor plaintiffs, because the “reasonableness” test seemed more objective. But this perspective would prove overly optimistic. It has provided limited tactical guidance for how police officers should treat people and how judges and juries should understand claims of police misconduct. Dethorne Graham saw this firsthand. Following the Supreme Court’s decision, his case went back to the trial court so that the jury could review the evidence under the new rules. Yet the jury concluded that the police officers’ behavior toward him was “reasonable.”

Subsequent decisions by the Supreme Court and lower federal courts have continued the Graham decision’s legacy of ambiguity in considering what constitutes “reasonable” force on the part of officers. Many police departments have also created their own set of administrative rules on when force is appropriate.

But these policies sorely lack specificity. In a study forthcoming in the Cornell Law Review, for which I am the lead author, my co-author and I analyzed use-of-force policies from the 75 largest U.S. cities and showed that they regularly fail to provide meaningful instruction to police on when to use force, or how to do so in ways that increase the likelihood that people will survive. For example, only 31 percent of the policies required officers to exhaust alternatives before using deadly force, and only 17 percent had policies that instructed officers to use force that is proportional to a person’s resistance. Meanwhile, all of the policies we examined restate Graham’s reasonableness standard, many times directly citing the case. The vagueness of this standard creates wide discretion for police and few protections for community members.

Our examination of the case law also showed that when people file lawsuits alleging that police used excessive force, federal courts often reference or defer to police departments’ use-of-force policies as the appropriate legal interpretation of “reasonable.” For example, in a 2004 case before a federal district court in West Virginia, in which an officer fractured the leg of the plaintiff, Kevin Neiswonger, as he tried to restrain him, the court held that the officer “acted reasonably under the circumstances to protect both Mr. Neiswonger and himself, in accordance with the Morgantown City Police Department’s Use of Force Policy,” and thus “did not violate Mr. Neiswonger’s Fourth Amendment right to be free from unreasonable search and seizure.” In this case, as in many others, federal courts were influenced by the idea that as long as an officer’s behavior does not violate the use-of-force rules created by his own department, his actions are not unreasonable and therefore not unconstitutional.

This suggests that the ongoing epidemic of police violence is not simply the result of what former Attorney General Jeff Sessions once described as “individuals within a department that have done wrong.” Instead, by allowing police to largely define what constitutes excessive force, the Court has limited its own judicial oversight of the system, creating the conditions that allow police to use violence with impunity. As a result, the individual bias often found among police officers can quickly translate into violence against minority communities.

Despite the thousands of lives that have been lost to police violence since the Graham decision, the Supreme Court has shown little interest in rethinking its approach. State and local governments, however, have the opportunity to be more proactive. For example, California Governor Gavin Newsom is expected to sign Assembly Bill 392, the California Act to Save Lives, which would change California’s current standard regarding police use of deadly force under state law (which looks at whether it was “reasonable”) by requiring that police use force only when “necessary.” This new term is not clearly defined, and some advocacy groups that fought for the bill have withdrawn their support due to this and other shortcomings. But in shifting from an ambiguous and deferential concept to a more concrete and cautious one, the proposed law seeks to encourage officers to prioritize other ways to resolve situations before using force that often shatters bodies, destroys lives, and fractures communities.   

Laws like A.B. 392 wouldn’t solve the police-violence crisis alone. Many aspects of policing need reform, such as rules on qualified immunity, which can shield officers who use excessive force, and the level of community participation when police departments develop or revise use-of-force policies. However, A.B. 392, despite its limitations, may offer an instructive case study in whether less permissive rules regarding police use of force can change the culture within police departments and perhaps save lives.

This article is part of our project “The Presence of Justice,” which is supported by a grant from the John D. and Catherine T. MacArthur Foundation’s Safety and Justice Challenge.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.