A mural in Baltimore depicting Freddie Gray.Patrick Semansky / AP

When Baltimore police officer Caesar Goodson Jr., was acquitted Thursday of all charges related to the death of Freddie Gray, the one emotion absent from the courtroom, social media, and the crowds of protesters in the city was surprise. The cases of all six officers alleged to be involved in Gray’s April 2015 death have been tossed about in a sea of strange legal wrangling and reshuffling, but without much real suspense. The trial of Officer Edward Nero ended in a judge’s acquittal, and  that of Officer William Porter in a hung jury. All six officers charged in the case remain on administrative, drawing full salaries, pending the outcome of an internal investigation. But it’s likely that these officers will share the fate of most officers accused of killing black people in the line of duty: a return to police work.

The process will grind on. It must. Baltimore State’s Attorney Marilyn Mosby is poised to continue cases against the three remaining charged officers who have not been tried. Judge Barry Williams—who presided over Goodson’s bench trial—will continue to hear them. Porter will be retried after those three, and a process that was originally slated to be completed this summer might stretch into 2017. But  the lesson at the end of that process will likely be the same as it was during trials for officers involved in other high-profile deaths of people of color. The criminal justice system just doesn’t possess the will, tools, or objectivity to consistently deliver justice for victims of police violence. Criminal justice is incapable of policing itself, and that will continue to be the case for as long as its inherent conflicts of purpose and interest remain. But the idea of restorative justice provides a sorely-needed alternative.

Restorative justice is a process that allows for direct mediation between victims of violence and police perpetrators. It also allows direct dialogue between families, communities, and police departments. It relies on a much less staggering burden for action than criminal courts—responsibility—and that action generally involves restitution, an admission of guilt or responsibility, and requiring responsible parties to take action to minimize further harm. It also avoids the near-inevitable letdown of communities hoping for verdicts and provides a remedy to the ineffectiveness of using the criminal-justice system to police the police.

In many ways, Baltimore prosecutors are actually pushing criminal justice beyond where it usually goes in exploring police violence. That verdicts are even being discussed in Baltimore is an anomaly. Indictments of police officers accused of misconduct are much rarer than internal investigations or civil proceedings from victims or their families. Police officers are by definition agents of state violence, protected by the aegis of state authority, and as such agents the standard of proof needed even to charge them with a crime is incredibly high. Even vivid video evidence did result in an indictment for McKinney, Texas, Officer Eric Casebolt; a grand jury declined to indict him for pinning a black teenage girl to the ground after a pool party. It was certainly not enough for Cleveland officers Timothy Loehmann and Frank Garmback, who were not indicted after a video showed Loehmann shooting 12-year-old Tamir Rice as he played in the park with a pellet gun.

Lethality is inevitable when police are armed with lethal weapons, trained to act aggressively, and used as a first response to dysfunction—or to random acts of existence by people of color. Widespread lethality is an obvious and foreseeable consequence of widespread policing, and becomes its implicit teleology. Lethality is enabled by broad protections that police are given under criminal law; without them officers would be unable to function in the aggressive way that they do. There are few ways in which police could go far enough beyond the enormous tolerance for violence against poor people and people of color to merit an indictment by whatever standards of law exist. This tolerance for blood is less an unintended consequence of policing than an integral feature of America’s approach to it, and policing provides the foundation for of all the rest of American criminal justice. Using criminal justice to curtail police violence presents a deep paradox.

Police used to do the business of violence surveilled by no one but those they maimed and killed, lawfully or not. That allowed for the spread of anecdotes from people who lived to tell the tale and a strong body of cultural awareness among black and brown people about the tyranny of police. We all knew about “rough rides” and boys who just disappeared after arrests. We all knew about women who had been “felt up” and assaulted or men who’d had guns pointed at their heads for nothing. Baltimore’s black citizens knew intimately the brutality of its department. But that cultural and anecdotal evidence was almost never enough to indict or punish police officers, nor was the testimony of their victims who had survived, pitted as it was against the credibility of officers on the stand.

Black Lives Matter and the past two years of black protests against police brutality have been built on a historical reversal of this black box of police interaction. Videos of arrests, beatings, and deaths of black people have become near-ubiquitous and have led to increased pressure for indictments when police kill people. And indictments seem to have increased with the protests, as 14 officers were indicted nationally for killings in 2015, a marked increase over the yearly average of six such indictments. But the overwhelming tolerance of police violence has still not been cracked by convictions, much less overturned. Officers still have the latitude to kill whenever they can say they were threatened, and often have the latitude to kill when they cannot. The thought exercise of envisioning scenarios in which officers in a given municipality would be guaranteed prison time for killing someone is morbid and exhausting. What precisely does the legal concept of murder mean for an agent empowered to kill people?

Not only is Mosby’s quest in criminal court destined to fail; its counterpart in civil court is also flawed. Civil suits, claims, and settlements at least provide victims and their families some avenue for restitution, and preponderance of evidence standards for civil lawsuits are lower than the criminal standard of proof beyond a reasonable doubt. Gray’s family received $6.4 million in the largest police settlement in Baltimore history. But settlements are often granted by city leadership and rarely serve to hold individual officers, police departments, or policing structures accountable. Settlements and judgments are also paid for out of the city’s coffers, meaning they are paid for in part by taxes from the very communities and families demanding justice. Too often, they are used as magic wands to wish problems away, as opposed to real methods of accountability. This seems to be how Chicago is handling its rampant police violence problem. It gave Laquan McDonald’s family $5 million after video of his killing set off mass protests and investigations into the police department, and has since doled out millions in settlements to victims.

Restorative justice stands as a remedy for much of what criminal and civil proceedings lack in holding police accountable, and it is being taken more seriously as a potential way to bring justice to the victims of police violence and to communities such as Baltimore’s that have been affected by it. In Baltimore, restorative justice might involve restitution from the six officers, the police union, or the department; officers losing their badges; or officers being required to perform community service. The lack of the prospect of a long prison sentence would lower the stakes, and reduce the adversarial nature of the dialogue between community and police. That might make police more willing to root out problems in their ranks, instead of resorting to obstruction and grandstanding.

Granted, restorative justice may not be palatable to communities hurt by police. Communities like Baltimore in 2015 and Los Angeles in 1992 that were on the edge after routine and brazen policing abuses, might reject alternatives to prison time for officers. Restorative justice is not entirely congruous with modern-day policing, and the broad authority that police have to kill necessarily makes restorative justice a conversation between actors with starkly different incentives to embrace change.

But, for all but the cases with the most damning evidence, restorative justice offers the most likely way to provide some justice for individual victims. It requires that police accept guilt and blame for misconduct or poor decisions, and also opens up possibilities for financial and non-financial restitution. It is more likely and less bloodthirsty than criminal justice and more comprehensive than civil claims.

On a communal level, restorative justice could serve another important function. Rage in the streets of Baltimore and other cities that have experienced police violence comes both from the actions of police—and from government denial that policing is or could be racist. People who know the reality of unjust policing and want to change it are often forced to validate their direct experiences by securing criminal convictions or assembling hard data in order to move any sort of policy forward. Since police are rarely indicted for the excessive use of force, let alone convicted—and even more seldom convicted of displaying racial bias—the available data points are mostly the same anecdotes of everyday brutality that are shared daily in black neighborhoods. Even with video evidence brought forward by black protesters and vigilant citizens, the burden of proving injustice is as intimidating as proving the guilt of a violent police officer. If restorative justice even slightly eases that burden, it will provide a sliver of hope that change is possible.

This article is part of our Next America: Criminal Justice project, which is supported by a grant from the John D. and Catherine T. MacArthur Foundation.

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