Policing the Police

An officer slams a handcuffed man against a wall. A mother shouts at the police as they arrest her son. A bystander raises a phone to record a stop, and the officer tells him to put it away. We give the police extraordinary power to keep us safe, and the same law that grants that power also decides when its use crosses the line. Often the line sits somewhere very different from where most people assume it does.

In this TalksOnLaw interview, Joel Cohen speaks with Rachel Harmon, the Harrison Robertson Professor of Law, F. Palmer Weber Research Professor of Civil Liberties and Human Rights, and Director of the Center for Criminal Justice at the University of Virginia School of Law, and author of The Law of the Police. They discuss how American law authorizes and limits police force, arrests, and the policing of protest, criticism, and recording, and why Harmon believes the most important checks on the police lie outside the courtroom.

We Create the Police Through Law

Harmon begins with a basic point: the police are not "magical creatures." They exist because the law authorizes them. Without that authority, she says, an arrest would be kidnapping and a use of force would be assault. Policing the police is therefore really a question of how well we govern the people we give that authority to, and how much authority we give them. Concern about police treatment of African Americans and immigrants goes back almost to the founding of American departments in the mid-1800s. What has changed is that more of the country now sees the problems firsthand on video. Lawyers usually point to the Fourth Amendment and to Miranda as the law of the police. Harmon agrees those matter, enforced through the exclusionary rule, prosecutions, and damages suits. But she argues that state law, which authorizes what officers do, and local government, which funds departments and hires and fires chiefs, are often where the rules actually take shape.

Force, Reasonableness, and "De Minimis" Force

A use of force is a seizure under the Fourth Amendment, so it must be reasonable. For force, Harmon explains, courts do not ask about probable cause. They balance the government's interests against the intrusion on the individual, look at whether the person was resisting, fleeing, or threatening anyone, and avoid judging with 20/20 hindsight. That balance has been generous to the government. Many circuits treat shoving a handcuffed person against a wall as de minimis force with no constitutional violation, even when it was unnecessary, because the test is not the minimum force needed. Courts become stricter with weapons, such as Tasers used against passive resistance. Harmon stresses that non-deadly force is far more common than deadly force, injures many more people, and is far less regulated. The fact that the Constitution permits something, she says, does not mean states or communities have to allow it.

Deadly Force and Fleeing Suspects

Deadly force is more tightly restricted, but Harmon says claims against officers remain hard to win. Officers may generally use deadly force when they or others face a threat of death or serious bodily harm, or to stop someone fleeing a dangerous crime. Under Tennessee v. Garner, police cannot shoot a non-dangerous fleeing felon, but they may shoot a dangerous one. The Supreme Court has suggested that being suspected of a violent crime can by itself make someone dangerous. Cohen raises a suspect wanted for a violent rape who might be the wrong person, and Harmon agrees that Garner appears to allow deadly force against him as he runs. Departments and communities cringe at that result, she says, and many states and departments allow less than the Constitution does. The court of public opinion often matters more than the federal courts. In practice, most officers who use deadly force say they believed they or someone else faced a serious threat.

Prosecuting Officers, and Systems Over Bad Apples

Harmon prosecuted civil rights crimes in the Criminal Section of the Justice Department's Civil Rights Division. She explains that the federal criminal civil rights statute requires willfulness, a demanding intent standard that makes charges against officers difficult. It still reaches deliberate excessive force, such as kicking a handcuffed suspect who is already under control. Cohen notes that officers do many important things to keep the public safe, and the conversation is not meant to suggest otherwise. Despite her years as a prosecutor, Harmon does not think prosecution is the best tool for most problems in policing. Prosecution matters, she says, because it treats victims as equal citizens. But policing is about systems: the conduct we authorize, how we structure departments, and the tasks we give them. Problems usually come from institutions, not from bad people.

The Hidden Costs of Lawful Arrests

An arrest requires only probable cause, which Harmon describes as a fairly low bar. Most arrests need no warrant, and an officer who arrests without probable cause can be sued. Her bigger question is why police make so many arrests at all. Even a short stay in custody can cost someone a job, leave children without care, cause trauma, and carry immigration and housing consequences. In over-policed communities, or where people feel targeted, even a justified arrest deepens alienation. Much of Harmon's work asks lawmakers and communities to weigh the costs of policing even when it is legal, and to restrict practices that do more harm and unfairness than good.

"Contempt of Cop" and Retaliatory Arrests

One of Harmon's favorite chapters in her book deals with "policing resistance." The unusual feature, she explains, is that the police decide in the first instance how to treat opposition to themselves. The Supreme Court has held that purely verbal opposition to the police is protected speech, while physically interfering with an arrest is not. The hard cases fall in between: shouting in an officer's ear, standing a few feet away, or stepping between an officer and a suspect. The case law, she says, is messy. "Contempt of cop" is an informal term for arrests or force in response to disrespect. There are dozens of federal cases about people arrested for giving an officer the finger, and she quotes the line "I can't give him a rap, but I can give him a ride." Retaliating for protected speech violates the First Amendment, and whether "fighting words" can ever justify arresting someone who insults an officer remains an open question. Harmon's view is that an arrest for purely verbal opposition is not acceptable. Under Nieves v. Bartlett (2019), however, probable cause generally defeats a retaliatory-arrest damages claim. There is a narrow exception when similarly situated people are not arrested, as with jaywalking. Because officers can almost always find probable cause for something, she calls this a real barrier to accountability. Research also shows officers arrest and use force more against people they see as disrespectful.

Protests and Dispersal Orders

The First Amendment protects political protest, but it allows reasonable time, place, and manner restrictions. Harmon says that gives police broad discretion. An unpermitted march that briefly blocks a street can lead to mass arrests, and a few people throwing rocks can lead a whole crowd to be treated as lawbreakers, or tear-gassed. Much of the right to protest, she says, is decided on the street rather than in court. Some departments see protecting speech as part of the law they enforce, and communities can demand that approach. Every state has riot, unlawful assembly, or dispersal laws. Once an order is given, police must provide notice and an opportunity to leave before arresting, although notice need not reach every person in the crowd. Courts struggle with what counts as enough notice, and after the Ferguson protests some issued preliminary injunctions on this point. But courts rarely second-guess officers after the fact, so Harmon calls for clearer rules up front.

Recording the Police

Harmon explains that every federal circuit court to address the question, and most state high courts, have recognized a First Amendment right to record police in public. The Supreme Court has not ruled on it. The litigation has now moved to what time, place, and manner limits are reasonable, such as how close someone can stand while recording an arrest. She notes harder cases, such as a homicide scene before the family has been notified. For a person recording their own traffic stop, officers may still order hands on the wheel for safety, but she believes that stopping someone from recording an encounter that is already being recorded, without interference or risk, should not be allowed. Recording technology, Harmon says, moved faster than the law, and the law moved faster than police training. She describes how George Holliday first offered his Rodney King video to the police, then to local news, and how it reached national audiences partly because the Gulf War had just ended. Today the media no longer decide what the public sees, and that has driven new legislative attention to policing.

Body Cameras and Who Gets the Footage

Access to body-camera footage, Harmon explains, is almost entirely a matter of state law, and states vary widely in whether footage is a public record and who may see it. A criminal defendant must receive exculpatory footage under Brady, but there is no constitutional right to all of the government's evidence. Many arrests, especially ones involving excessive force, never lead to prosecution, so criminal discovery cannot solve the problem. Police recording also raises privacy concerns, such as a sexual-assault victim who does not want to be filmed, and states have begun passing laws on when officers must record. Harmon closes on her central theme. The law of the police should give clearer guidance up front and provide better remedies afterward, but the main way Americans govern the police is through the political process, in communities deciding what kind of policing they want.

What to Know Now

This conversation was recorded in 2021, and the Supreme Court has since revisited two of the doctrines Harmon describes. In Gonzalez v. Trevino (2024), the Court held that the Nieves exception for retaliatory-arrest claims does not require examples of similar people who were not arrested. Any objective evidence that officers usually do not arrest for that conduct can be enough. In Barnes v. Felix (2025), a unanimous Court rejected the "moment-of-threat" rule some circuits had used in police-shooting cases. It held that excessive-force claims must be judged under the totality of the circumstances, including events before the shooting, but it did not decide how an officer's own conduct in creating the danger should count. Courts are also now testing limits on recording police. In 2025 the Seventh Circuit upheld Indiana's 25-foot police "buffer" law against a First Amendment challenge in Nicodemus v. City of South Bend. A separate panel later affirmed an injunction blocking the law as unconstitutionally vague, and Indiana has amended it to require that an officer reasonably believe the person will interfere. Harmon published a second edition of The Law of the Police in 2024, adding material on retaliatory arrests, state qualified-immunity reform, and community efforts to limit the scope of policing.

Federal policy has shifted sharply. President Biden's 2022 Executive Order 14074 on accountable policing, which created a national database of federal officer misconduct, was revoked on January 20, 2025. In April 2025, Executive Order 14288, "Strengthening and Unleashing America's Law Enforcement," directed the Justice Department to expand legal protections for officers and to review consent decrees. In May 2025 the Justice Department moved to dismiss its proposed consent decrees with Minneapolis and Louisville and retracted findings of constitutional violations against several other departments. Both cities pledged to continue reforms on their own. The George Floyd Justice in Policing Act, which would lower the intent standard for federal prosecution of officers from willfulness to recklessness and limit qualified immunity, was reintroduced in September 2025 but has not been enacted. Reform has continued mainly at the state and local level, as Harmon predicted. According to Mapping Police Violence, police killed 1,365 people in 2024, the most since it began tracking. In 2025 that number fell to 1,314, the first decline in six years, although researchers caution that one year does not establish a trend.