Then-Attorney General Jeff Sessions speaks in Washington on May 3. (Susan Walsh/AP)
Opinion writer

Over at National Review, Walter Olson has written a partial defense of Attorney General Jeff Sessions’s last official act, a memo that put new restrictions on the use of consent decrees. Olson also adds in some criticism of the media, which he writes was “primed . . . to fit Sessions’s every move into a pre-set frame of criticism” — which is to say, defending cops from accusations of abuse. Olson is a senior fellow at the Cato Institute and co-founder of the Overlawyered blog. He’s also a very smart guy and a friend. But I think his article gets some important things wrong. It’s also the most concise and well-argued piece in opposition to consent decrees, so it’s worth addressing at length.

A consent decree is a binding agreement between the Justice Department and some large or official entity, usually coming after the agency has found evidence of ongoing wrongdoing. The other party to a consent decree can sometimes be a corporation, but more often it’s a local or state government. The most well-known variety are consent decrees aimed at reforming police departments, but there are lots of others. Most, including all of those associated with policing, aim at correcting mass violations of constitutional rights, although they can also be used to force compliance with environmental, labor or other regulations.

Olson makes some good points on the problems with how consent decrees are implemented. He’s also right that public discussion of consent often focuses on their application to policing, and tends to overlook their use — and often misuse — in other areas. But his piece also overreaches, particularly in his criticism of media reaction to Sessions’s order and his dismissal of the fear that the order will make it difficult to target systemic police abuse. His article is also a useful vehicle to examine Sessions’s orders in more detail, as well as to take a closer look at the costs and benefits of the consent-decree model for police reform. As you might guess, contrary to Olson, I think they do a lot of good.

Let’s start with Olson’s media criticism. He writes that “Critics promptly assailed [Sessions’s order] as motivated by a wish to let brutal police off the hook,” then notes:

"If you look at a copy of the order itself, though, you may be struck by something: Not once in its seven pages does the word “police” even appear. That’s a clue that the press missed much of the story.

The debate in policy and legal circles over consent decrees goes back decades, and has only recently begun to overlap with the debate over police misconduct. The Justice Department’s website lists numerous decrees its negotiations have extracted from state and local governments, few of which have anything to do with cops.

Olson also correctly points out that Sessions has long objected to consent decrees, and in the past has decried their effects on all sorts of agencies that have nothing to do with law enforcement.

All of these things are true. And yet it wasn’t unfair for the media to speculate that halting federal oversight of police agencies specifically was Sessions’s primary motivation for his order. Why? Because of Sessions’s record. While he has issued broader condemnations of consent decrees from the campaign trail to the time he took office as attorney general, Sessions made it clear that one of his primary goals was to dramatically scale back federal oversight of law enforcement. He explicitly said he thinks consent decrees both increase crime and hurt officer morale. He also objected to the very notion that there could be systemic problems in police departments (such as in, say, Little Rock). He said in his confirmation hearing that mere criticism of a police department damages all police departments and, therefore, hinders effective policing.

One of Sessions’s first acts upon taking office was a memo ordering a review of all current agreements between the Justice Department and police departments across the country, from joint task forces to information-sharing agreements to consent decrees. The memo laid down a set of priorities that department personnel should emphasize when conducting such reviews. Among them: promoting “officer morale” and “public respect for their work.” The memo also prioritized “local control” and “local accountability,” emphasized that “it is not the responsibility of the federal government to manage non-federal law enforcement agencies” and cautioned that the “misdeeds of individual bad actors” shouldn’t malign the honor and hard work of law enforcement agencies.

This is a central tenet of what you might call the “bad apple theory.” Sessions doesn’t deny that some police officers can be abusive. In fact, during his time at the Justice Department, Sessions was fairly proactive at bringing federal civil rights charges against individual officers, particularly with respect to abuses at county jails. But Sessions has rejected the idea that there might systemic abuse in some departments.

Again, the memo calling for the review of police agreements with the federal government was one of Sessions’s first actions as attorney general. So it doesn’t seem unfair to speculate that the priorities underlying his first memo may have motivated his last. (I also looked for examples of Sessions criticizing other types of consent decrees after he became attorney general. Perhaps they exist, but I couldn’t find them. They certainly weren’t a high enough priority for him to spend much time on them in public appearances.)

Olson is right that Sessions’s orders will affect all consent decrees, not just those associated with policing. But he also suggests near the end of his article that the alarm over how the orders will affect police oversight was overblown, and he makes several other more general criticisms of consent decrees, as they apply to policing. He writes:

They let outside critics manage (and micro-manage) local agencies. Decrees, which may be hundreds of pages long, install DOJ (or some other lawsuit-filer) to oversee and second-guess the operations of the sued city or state, in an enviable position of power without accountability. The deal often includes the appointment of a monitor who might even move in to the subject agency’s offices on a full- or part-time basis.

It isn’t clear to me why this is inherently a bad thing. The police culture described in Justice Department reports from Ferguson,Mo., Chicago, Baltimore and elsewhere (which Sessions criticized but later conceded he hadn’t bothered to read) has existed for years or even decades. Any professional culture ingrained for that long will resist change. It’s hard to see the harm in putting new people on the ground to oversee that change.

One of the independent monitors' key responsibility is to hold regular meetings with the public to assess where new policies are succeeding and failing. Where the police relationship with some marginalized communities is particularly bad, members of those communities may feel more comfortable talking to an independent monitor than with a representative of the same department they feel has been disrespectful or abusive. Monitors also meet with rank-and-file officers to talk about how the policies are affecting them — again, officers whose opinions differ from the status quo would presumably be more comfortable talking honestly to someone from outside the agency than inside of it. In an interview for this piece, Christy Lopez, who worked in the Justice Department’s Office of Civil Rights under President Barack Obama and led the investigation into the Ferguson police department, said that of the jurisdictions in which consent decrees have failed to significantly change how policing is conducted, most were decrees that were not enforced by independent monitors.

Here’s another Olson criticism:

They last and last. Having acquired this valuable power, the feds or other plaintiffs can be leisurely about relinquishing it. Definitions of what constitutes compliance can be vague, complex, and doubtfully practical, and even if the defendants manage to show that they have crossed every “t” and dotted every “i,” they may still need to prove that they are not likely to backslide when taken off the hook. So the process drags on — sometimes for decades, sometimes indefinitely.

There are definitely some compelling examples of this, particularly in the reports and congressional testimony that Olson cites (though most do not involve law enforcement agencies). And I share his concerns that the incentives here can be problematic — a monitor who declares an agency to be “fixed” also ends his or her job as a monitor.

But it may also be that in some of these examples, the decree dragged on because the local agency failed to improve. The Post reported in 2015 that this was true of several early consent decrees with police agencies. A number of jurisdictions, including Detroit, Los Angeles and Prince George’s County, Md., had been under decrees for up to or more than 10 years, at considerable expense to those cities and the county. That may be because the policies aren’t working, but it could also be because police culture is difficult to change overnight. But the early examples also came as the program was just getting started (Congress didn’t authorize the use of consent decrees to reform police departments until 1994), and organizers were still trying to figure out which policies work, which don’t and even how to measure effectiveness.

In a more positive 2017 report, the Justice Department noted that the duration of policing consent decrees grew significantly shorter during the Obama administration. The report also referenced independent research that found notable improvements in policing after decrees in, among other places, Pittsburgh, Cincinnati, Washington, Seattle and New Jersey.

More from Olson:

Things get done behind closed doors . . . ordinary taxpayers, parents, and other affected interests are sure to wind up on the outside.

This may be true when cities form agreements in response to litigation from private groups or nonprofit advocacy groups, but city and state agreements with the Justice Department generally come with a period of public comment. The Chicago Police Department agreement, for example, came with a guaranteed two to four weeks for public input. On the first day, more than 200 people came to offer their comments before the federal judge who considered it. The public comment period almost didn’t happen, not because the city, the DOJ’s staff or the ACLU wanted to keep it all secret, but because the police unions and Sessions himself tried to get the agreement thrown out, then attempted to drag out litigation well into the period of public debate.

Back to Olson:

They frustrate democracy. Who answers to local voters? Not the control group, as it has been termed, of federal civil servants or other plaintiffs managing the decree. As for the local agency, once its hands are legally tied, mayors and city councils can come and go and it doesn’t matter: It’s unlawful to change direction even if local voters want to.

At least with policing, a consent decree comes only after a Justice Department investigation has revealed practices of policing that cause large-scale violations of constitutional rights — revelations that, if disputed, must also then be proven in court. Under the 14th Amendment, the federal government has an obligation to protect the constitutional rights of U.S. citizens when state or local officials either violate those rights or fail to adequately protect them. This is true even if a majority of voters or local elected officials support the violations, don’t believe violations are occurring, or simply don’t care. We don’t and shouldn’t subject constitutional rights and their enforcement to a popular vote.

“You have to understand that the courts role in a consent decree is to protect people who don’t have access to political power,” Lopez says. “They’re used in places where we’ve had to prove that local officials have repeatedly and systematically failed to protect people, where marginalized groups have been failed by the political process.” Taking decision-making power away from those officials — at least those decisions that pertain to constitutional policing — is precisely the point.

Furthermore, the “local power” argument is particularly difficult to apply to Sessions, who is hardly a principled advocate for local control. He violated the wishes of local voters and local officials when it suited other priorities. The best example is Sessions’s threat to withhold federal funding from sanctuary cities, despite the fact that police leaders in those cities say enforcing federal immigration law would make their cities less safe. (And there’s data to back them up.)

Interestingly, Sessions could have taken a similar approach to police abuse. He could have threatened to withhold funding to police agencies that have shown a pattern of abuse, and he’d likely have been on firmer legal ground. That he only went this route for immigration enforcement tells us something about his priorities. Sessions also fought to throw out the consent decrees in Chicago and Baltimore even though local officials — including the heads of those cities' police departments — wanted them.

It may seem counterintuitive for local officials to welcome Justice Department scrutiny, but it isn’t uncommon. The city of Elkhart, Ind., also recently asked for a DOJ investigation after ProPublica reported shockingly high rates of misconduct and police shootings there. There are also good reasons for some local leaders to welcome federal oversight. As former Seattle police chief Norm Stamper wrote in his book “Breaking Rank,” when mayors, city managers and other local officials negotiate police contracts, they’re often faced with a limited budget. So they’ll often compensate with other benefits, such as increased job protections for cops accused of misconduct, a “police officer bill of rights,” or sending disciplinary cases to officer-friendly arbitrators. They and their successors are then bound to these contracts, which can help enforce the “blue wall of silence” and contribute to destructive police culture. To get a federal court to toss them out as part of a consent decree may often be the only way out. Yet when city leaders wanted that option, Sessions fought them.

Sessions also altered a program in which cities could voluntarily ask the Justice Department to review their police agencies to ensure they’re using the best practices and policies to protect the constitutional rights of the people they serve. Sessions didn’t change the program out of some devotion to federalism or limited government. Instead, he refocused the program toward providing grants for traditional policing such as anti-drug and anti-gang enforcement. The program still pushes federal policy on local police, but instead of offering an incentive for reform, it offers an incentive to continue with more aggressive, reactionary policing.

Olson closes his article by arguing that Sessions’s orders are “modest,” and paints them as common-sense proposals aimed at reining in consent decrees that have outlived their purpose. Again, I can only speak to the orders as they apply to policing, but it seems clear that the orders will make it more difficult for the Justice Department to investigate and oversee problematic police agencies.

For example, one order requires that “provisions of the consent decree must be narrowly tailored to remedy the injury caused by the alleged legal violation.” Strictly interpreted, this means that Justice Department investigators would be unable to pursue any tangential issues or underlying causes that may be causing the constitutional violations. For example, in Ferguson the police targeting of black residents for petty infractions was driven by a dizzying, predatory system in which municipalities in St. Louis counties are heavily incentivized to supplement their budgets with revenue from city courts — and that revenue largely comes from fines for traffic violations and other petty crimes. If Justice Department investigators had been limited to investigating only violations of the Fourth and 14th amendments, they could well have been prevented from ever delving into the role of the municipal courts. “I think we succeeded in showing there was a direct connection between the municipal courts and how the police were violating the rights of Ferguson residents,” Lopez says. “But under Sessions’s memo and the spirit in which it was written, we would not have been able to address the courts' role. I’m confident that the current administration never would have let that through.”

Chiraag Bains, who co-authored the Ferguson report, agrees. “Under this memo and this administration, there would never have been either an investigation or a consent degree in Ferguson,” he says. “The problem with limiting an investigation or an agreement to a specific issue like shootings or the use of police dogs is that you can’t get to the underlying culture and systemic problems that give rise to those issues. In some of these departments, you need a multi-pronged approach. Without that, the same problems are going to crop up as soon as you leave."

Another part of Sessions’s order requires that consent decrees and settlements be approved by “senior department leadership.” Olson characterizes this as an effort stop “underlings” from cutting “consent-decree deals without high-level supervision.” But it isn’t about supervision so much as it’s about putting the ultimate decision about consent decrees with high-level political appointees who, in this administration, tend to be pretty adamantly opposed to consent decrees. On the one hand, there’s the argument that elections have consequences, and a president is entitled to put important policy decisions in the hands of the policymakers he appoints. But once again, the counterargument here is that constitutional rights are not beholden to the whims of voters.

Olson also complains about the inflexibility of agreements as they’re currently enforced, and it’s a fair criticism. But Sessions’s order actually makes it more difficult to alter an agreement once it’s in place. “The order significantly raises the bar if you want to modify an existing agreement,” Lopez says. “It’s already really difficult. This would make it nearly impossible. As a result, there would be a strong incentive to sue instead of to settle, and that can be very expensive, for both the federal government and the city.”

One final and particularly problematic part of the order is, somewhat paradoxically, the one way it would make it easier for the Justice Department to pursue consent decrees — the restrictions are relaxed if a consent decree would protect a city from a third-party lawsuit. Lopez explains. “Let’s say the DOJ knew that the ACLU or NAACP was about to file a lawsuit against a city’s police department,” she says. "And they know that if successful, the lawsuit would bring reforms that someone like Sessions thinks are excessive or burdensome to police officers. This allows the federal government to hastily offer a consent decree that would essentially shield the city from the third-party lawsuit.” That doesn’t seem very democratic, either.

Perhaps the most important question concerning consent decrees is — do they work? The evidence is mixed. The 2015 Post investigation made a compelling case that some of the longer-lasting agreements weren’t achieving much change. But Lopez points out that in addition to the fact that most of the examples from that investigation began in the 1990s, critics also cited metrics that may not be the best way to measure success. “I’ll give you one example,” she says. "We often see people cite use of force incidents as a way to measure whether a consent decree is working. But one of the most common problems we see in police departments is a failure to report uses of force, and one of the most common reforms is a requirement that force be well-documented, with punishment for officers who fail to do so. So we should see use of force incidents go up after a consent decree, particularly in the short term.”

Bains agrees: “I don’t think it’s fair to look at the longest-lasting decrees and cite them to say the entire program isn’t working. There’s a reason those agreements lasted so long. They tended to have the least amount of cooperation from local officials and police leaders. One thing we tried to do in the Obama administration was get input and support at the local level, not just from politicians, but also from civic groups and activists. It makes a huge difference.”

The bulk of the timeline for those early, long-lasting consent decrees also came during the George W. Bush administration, which, like the Trump administration, wasn’t particularly friendly to the sort of systemic reviews undertaken during the Obama years.

But more recently, three years after Seattle’s consent decree went into effect in 2013, public trust in the police rose from 60 percent to 72 percent. More importantly, trust among blacks rose from 49 percent to 62 percent, and among Latinos from 54 percent to 74 percent. After the New Orleans consent decree in 2013, public trust in the police soared. A 2016 poll found that 69 percent of residents thought the city police culture had improved since the decree. A 2018 poll that 87 percent of respondents said their interaction with a New Orleans police officer had been safe and courteous, and 83 percent felt safe in their own homes. The city still has a high crime rate, and there’s still dissatisfaction with response times, but both could be attributed at least in part to the department’s shortage of personnel.

Lopez says this metric — public trust, particularly among minority communities — is the most effective measure of success. “There are just way too many variables that can affect use of force incidents, or crime rates, or officer shootings,” she says. “Even surveys can be misleading, but I think they’re much more valuable. We do surveys of the community in general, of marginalized communities, and of police officers. If we can show that community-officer relations have improved, then we’re on the right track. And since New Orleans, nearly all of our agreements have shown improvement in that area.”

More trust means people are more willing to cooperate with police to report and help solve crimes. It also means more appreciation for police, which can only be good for morale, even if, as Sessions insists, morale takes a temporary hit at the thought of more oversight, paperwork or criticism. That may be why city leaders in places such as Detroit and East Haven, Conn., were supporters of consent decrees by the time those agreements ended.

If Trump were to lose in 2020, a Democratic administration could easily revoke Sessions’s order, but in the meantime, not only will the administration cease investigating new police departments, but existing decrees could also be left to atrophy. If consent decrees were any other policy area, it would be easy to cite their mixed results, note their expense and ridicule the fact that even the people implementing them have acknowledged that they’ve been figuring it out as they go. But this isn’t just any policy area. The Justice Department reports from Chicago, Ferguson, Baltimore, Cleveland, New Orleans and other cities documented systematic abuse, some of it jaw-dropping. Those allegations have been supported by journalists, activists and lawsuits. Legally, the federal government isn’t just authorized to investigate and try to correct these problems, there’s a good argument that it’s obligated to do so. But morally, the case is even clearer. Given what’s at stake, the fact that these agreements haven’t always been successful doesn’t relieve of us of the imperative to keep trying.

Read more:

Video: Exposing police misconduct in Little Rock isn’t hard. Finding justice is.

Radley Balko: Jeff Sessions, the doughty bigot

Kimbriell Kelly, Sarah Childress and Steven Rich: Forced reforms, mixed results

Radley Balko: Consent decrees have a mixed record of success, but Sessions’s plan to end them is still worrisome