Blog
Consent Decrees and How the Justice Department Polices the Police
What a 1994 law lets the Justice Department do to a police force that violates rights, why the agency dropped the Minneapolis and Louisville cases in 2025, and what cities can still do.
In late May 2025, days before the fifth anniversary of George Floyd's murder, the Justice Department's Civil Rights Division announced that it was walking away. It would ask federal courts to dismiss its cases against the police departments of Minneapolis and Louisville, the two cities whose names had become shorthand for the protests of 2020. It would also withdraw the findings it had issued about several other departments, including those in Phoenix, Memphis and Trenton, and the Louisiana State Police. The cases had taken years to build. The investigators had interviewed officers and residents, read thousands of reports and watched hours of body-camera footage. The division's new leadership described the agreements as federal overreach that tied the hands of local police. Most people reading the headlines had no idea what a pattern-or-practice case was, or that the power to bring one rests on a single paragraph in a crime bill signed more than thirty years ago.
In short: A 1994 law lets the Justice Department sue a police department that routinely violates people's constitutional rights, and most such cases end in a court-supervised agreement called a consent decree. The second Trump administration abandoned the Minneapolis and Louisville cases in 2025 and retracted other findings. The tool's future now depends heavily on states, cities and voters.
The paragraph in the 1994 crime bill
The Violent Crime Control and Law Enforcement Act of 1994 is mostly remembered for things progressives now criticize: money for prison construction, new federal death-penalty crimes, incentives for longer sentences. Tucked into it is a provision that pulls in the other direction. Now codified at 34 U.S.C. 12601, it makes it unlawful for any government authority, or anyone acting for one, to engage in a pattern or practice of conduct by law enforcement officers that deprives people of rights protected by the Constitution or federal law. It then gives the attorney general power to go to court for relief to eliminate that pattern or practice.
Notice what the law does not do. It does not let a person who was beaten sue under it; individuals still have to use other statutes and face doctrines like qualified immunity. It does not award damages. It does not create any new crime for officers. It is aimed at institutions, not individuals: the training, supervision, discipline, policy and culture that produce violations over and over. And only the federal government can use it.
The provision had a specific origin. In March 1991, Los Angeles police officers beat Rodney King on the side of a freeway, and a bystander's video showed the country what many Black residents had been describing for years. Congress spent the next few years debating whether the federal government should be able to address abuse at the level of a whole department. The 1994 bill answered yes.
How a case actually works
A pattern-or-practice investigation usually starts with a public announcement, often after a high-profile incident, sometimes after years of complaints. Lawyers and investigators from the Civil Rights Division's Special Litigation Section then spend many months inside the department. They review use-of-force reports, stops and searches, complaint files, internal-affairs investigations and training materials. They ride along. They talk to officers, supervisors, prosecutors, public defenders and community members.
If they find a pattern, the division publishes a findings report, laying out the evidence. Those reports are some of the most detailed public accounts of American policing that exist. The Minneapolis findings, issued in June 2023, described a department that used excessive force, including unjustified deadly force, and discriminated against Black and Native American people during stops. The Louisville findings, issued in March 2023, described unlawful searches based on invalid warrants, excessive force and discriminatory enforcement, and they discussed the no-knock warrant practices behind the raid that killed Breonna Taylor in March 2020.
After findings, the department and the city usually negotiate. Few cities want a trial against the federal government, and few federal lawyers want to spend years litigating when an agreement can begin changing things sooner. The result is typically a consent decree: a settlement that a federal judge approves and enforces. The decree lists required reforms, such as new use-of-force policies, better supervision, data collection and early-warning systems for problem officers. An independent monitor reports to the court on progress, and the decree ends when the department has stayed in compliance for a set period.
The documents themselves are long and surprisingly concrete. A typical decree spells out when an officer must report force and what the report has to contain, how quickly a supervisor must review it, what counts as a reason to stop someone, how complaints are taken and investigated, and how the department will collect data on stops by race. It often requires crisis-intervention training and new ways of handling calls involving mental illness. Many decrees also set up community advisory bodies, so residents can hear directly from the monitor and the department about what has been done. The monitor's reports are public, which means anyone can read, every few months, exactly where a department is failing.
Pittsburgh entered the first such decree in 1997. Los Angeles signed one in 2001, after the Rampart corruption scandal, and the court ended it in 2013. Seattle, New Orleans, Ferguson, Baltimore, Albuquerque, Cleveland and others followed over the next two decades.
The fair case against decrees
People who oppose consent decrees are not all defenders of police misconduct, and their arguments deserve a real hearing.
The first is cost and duration. Monitors are expensive, and some decrees have lasted far longer than anyone expected, with cities paying millions of dollars a year for compliance work while residents wonder what has changed. The second is democratic accountability. Under a decree, important decisions about a city's police move from the mayor, the council and the chief to a federal judge and a monitor whom no local voter elected. The third is crime. Critics argue that when officers feel watched and second-guessed, they pull back from proactive work, and that some cities under decrees saw violence rise. The fourth is federalism: policing is a local responsibility, and Washington should not be writing training manuals for city departments.
Each of these has something to it. Some decrees have dragged on. Some monitors have been costly. The research on crime effects is genuinely mixed and depends heavily on which city and which years you examine.
Here is the answer, though. The law does not apply to departments with ordinary problems. It applies when federal investigators find, and can prove in court, a pattern of unconstitutional conduct. A city that systematically violates residents' rights has already given up the claim that local control is working. Duration and cost are reasons to write better decrees with clear end points, not to abandon the remedy. And the studies that find reductions in civil-rights lawsuits and use-of-force complaints in some decree cities should count at least as much as the anecdotes about officers pulling back. Constitutional policing and public safety are supposed to be the same project.
Progressives should also own a mistake here. In some cities, reform advocates treated a decree as the finish line, declared victory when it was signed and then stopped showing up at the monitor's public meetings. Compliance work is slow, technical and dull, and it needs the same steady pressure that won the agreement in the first place.
A tool that swings with each administration
Because only the attorney general can bring these cases, their use has risen and fallen with every change of administration. That is the structural weakness of the 1994 design.
Near the end of his time as attorney general in November 2018, Jeff Sessions signed a memo that sharply limited the use of consent decrees with state and local governments, requiring high-level approval and evidence of more than individual misconduct. In April 2021, Attorney General Merrick Garland rescinded it, and the division opened new investigations, including Minneapolis and Louisville. Those two investigations ended in agreements negotiated in the final weeks of the Biden administration, in late 2024 and January 2025, which then needed a judge's approval.
The new administration's lawyers asked the courts to dismiss the cases before those agreements took effect. The division's leadership argued that the findings had been rushed, that the agreements would cost the cities heavily and that the federal government should not be micromanaging local police. Supporters of the agreements, including civil-rights groups and many residents who had testified to investigators, argued that the findings were the product of years of work and that dropping them sent a message to departments everywhere that the federal check was gone.
There is a lesson here for people who want lasting change, and it is not a comfortable one. A reform that one administration can switch off is not a settled reform. If the protection matters, it needs a second anchor that does not depend on who holds the Justice Department.
The states step in
That second anchor is already being built, unevenly. Minneapolis had a second agreement all along: in 2023 the city entered a court-enforceable settlement with the Minnesota Department of Human Rights, which had run its own investigation after Floyd's death. Minneapolis officials said after the federal retreat that they would carry out the reforms in the abandoned federal agreement anyway. Louisville's leaders made a similar promise to implement the reforms on their own.
Chicago's police department has operated since 2019 under a consent decree with the Illinois attorney general rather than the Justice Department. California, Washington and several other states have given their attorneys general authority to investigate local departments. State laws on use of force, decertification of officers who commit serious misconduct, and public access to disciplinary records have spread since 2020.
None of this is as strong as a nationwide federal power that every administration uses. State laws vary, state attorneys general change with elections too, and many states will not act at all. But it means the question of police accountability is now partly on the ballot every time a state elects an attorney general or a legislature, which is where a lot of residents can actually influence it.
What decrees can and cannot do
It helps to be honest about the ceiling. A consent decree can rewrite policies, require better data, build early-warning systems, change training and force supervisors to review force incidents. It can make a department document what it does and explain why. It cannot by itself change who wants to be a police officer, how a union contract handles discipline, how prosecutors treat officers who break the law or how much a city invests in housing, mental-health response and youth programs that reduce the calls police are sent to answer.
The cities that have made the most progress usually paired a decree with local leaders who wanted the reforms and kept pushing after the monitor left. The cities that struggled often treated the decree as an outside imposition to be outlasted. That suggests where the leverage really sits: with mayors, council members, police chiefs and the voters who choose them.
The people on both sides of the badge
It is easy to forget that the most detailed parts of these findings reports come from officers. Investigators routinely describe officers who told them that training was outdated, that supervisors never reviewed their force reports, that the department rewarded arrest numbers over judgment. Many officers want clear rules and real supervision, because they are the ones who carry the consequences when a department fails.
The residents in these reports are often people who never imagined they would talk to a federal lawyer: a woman stopped repeatedly on her way to work, a teenager searched with no reason given, a family whose door was broken in on a bad warrant. Their stories became evidence. When a case is dropped, their testimony does not disappear. It sits in the public record, waiting for the next official willing to act on it.
What you can do
The federal retreat moves the fight to places where you have more say. A few ways in:
- Ask your state lawmakers about their own oversight power. Look up your state legislators on the representatives page, which also shows their recent votes, and ask whether your state attorney general can investigate a local department for a pattern of rights violations. If the answer is no, ask what bill would change that.
- Read the findings for your region. If federal or state investigators have ever examined a department near you, find the report; news outlets and civil-rights groups have kept copies of many of them. Reading one about a department near you is the fastest education in how policing actually works on the ground, and it gives you specifics to raise at council meetings.
- Track the courts. The Supreme Court's decisions on searches, qualified immunity and federal power shape what any of these reforms can achieve; the court-watch page lists the term's opinions with the Court's own summaries.
- Find the local groups already watching. Justice reform and legal aid nonprofits in your state are listed in the organizations directory; many of them run court-watching programs or community police boards that need steady volunteers.
- Keep up with the broader fight. The topic page on democracy and the rule of law follows how the Justice Department's priorities are changing across civil rights, voting and policing.
The 1994 law is still on the books. A future attorney general can pick it up again. Whether there is pressure to do so depends on whether people remember, between administrations, why it was written.