This story was produced in partnership with the Garrison Project, an independent, nonpartisan organization addressing the crisis of mass incarceration and policing.
It shouldn’t come as a surprise that the rise in shootings and homicides that began amid the turmoil of 2020 has undermined support for progressive-leaning criminal legal reform. But the apparent schism within Democrats—with many moderate Democratic leaders, including the president, calling for fairly conventional policing responses to the rise in violence—may be more unexpected.
Notably, the break within the Democrats runs alongside a fault line across institutions. The divide isn’t just between “moderate” and “progressive” Democrats. It’s between moderate mayors and progressive prosecutors. It’s Eric Adams vs. Alvin Bragg in New York City, Lori Lightfoot vs. Kim Foxx in Chicago, London Breed vs. Chesa Boudin in San Francisco, LaToya Cantrell vs. Jason Williams in New Orleans.
Yet this fracture exposes a long-running but underappreciated tension within the movement to address both the scope of and the disparities within the justice system, one that makes political sense but also raises serious policy problems.
At its heart is this: There is no such thing as “criminal legal reform,” because there is no such thing as the “criminal legal system.” What we call the “criminal legal system” is actually a chaotic web of overlapping, often inimical systems—plural.
Arrests are made by city police who respond to a police chief appointed by a locally elected mayor, or by sheriff’s deputies who report to a sheriff elected by county voters. Prosecutions are determined by prosecutors elected by county voters, while the judges who oversee the system are either elected by city or county voters or appointed by various executive officials (and that’s without getting to the fact that in some places judges who set bail are chosen differently than those who impose sentences). And the complexities keep growing. Prison sentences are overseen by state-run departments of correction, but shorter jail terms are served in county-funded facilities; parole is determined by a state parole board, but probation is overseen by agencies that are state-run in some places and county-run or locally run in others.
Conceivably, this mélange of institutions could be run in a coherent way, but we fracture political and budgetary responsibility so that any real cohesion across all these groups is essentially impossible. In other words, much of the moderate-progressive schism isn’t over “criminal legal reform” writ large, but rather reflects that the term homogenizes a complex set of systems, one where people hold different views about the different pieces.
For example, over the 2010s, a tenuous bipartisan consensus formed for scaling back prison populations, at least for nonviolent crimes (one that would quickly break down as soon as the conversation so much as shifted to violent offenses); it’s likely that this consensus still persists, even in the wake of all the upheavals of 2020. And this is an issue that progressive prosecutors are well situated to address.
But there has always been much less of a clear desire for reducing policing, especially for white voters. White confidence in policing has barely budged over the last three decades, holding fairly steady at around 60 percent. For Black people, confidence in the police rose to 27 percent in 2021 from a contemporary low of 19 percent in 2020 (which is still lower than any other year since 1993, when Gallup started asking the question). A separate Gallup poll in June–July 2020 found that overall only 19 percent of Black people and 12 percent of white people wanted to see less policing in their neighborhoods. Note, though, that when the Gallup poll zoomed in on Black people who saw the police in their neighborhoods often or very often, 34 percent, not 19 percent, said they wanted to see less policing.
More broadly, polls suggest that Black support for policing is often deeply ambivalent. A Data for Progress poll conducted in 2021 reported that among white respondents, 81 percent said that police patrols made them feel safe, and 72 percent said that “most police officers can be trusted.” Among Black respondents, however, a somewhat similar 65 percent said that police patrols made them feel safer, but 63 percent also rejected the choice of “most police officers can be trusted” in favor of “you can’t be too careful in dealing with police officers” (admittedly, the only two choices the question permitted).
These results echo the findings of James Forman’s Pulitzer Prize–winning 2017 book Locking Up Our Own, which chronicles D.C. from the 1970s to 1990s and the demand for tough-on-crime policing from the majority-Black City Council. While the Council’s Black members were aware of the costs of policing, they also understood they needed the police to get the guy with the gun off the street corner today. But their demand for policing was seen as a short-run solution to permit less police-centric social investments to bear fruit (investments that D.C.’s congressional overseers ultimately declined to fund), not so much as an optimal way to address the risk of crime over the long haul—an attitude that surely persists to this day.
Mayoral support for policing, then, makes significantly more sense in this context. And, on a policy front, it is even more understandable because the evidence is clear that more policing does reduce crime, even though it often comes with significant opportunity costs: Given the levels of police spending we already have, it’s likely that an extra dollar of spending has bigger net social gains invested elsewhere, like in drug treatment. In addition, our usual measures of the cost of policing focus just on the budgetary costs (what we spend on police), not the social costs (like the murder of George Floyd, or the costs of routine stops, which are likely much larger). Still, unlike tougher prison sentences, policing does have empirical support.
But policing is just one component of our sprawling criminal legal system. And there is a real and underappreciated cost of prosecutors, instead of mayors, being the leading proponents of progressive reforms. Progressive prosecutors, by and large, can minimize some criminal legal system harms, but they have no ability to push for solutions that fall outside of the justice system.
Here’s a concrete example of this problem and its risks. In 2018, one of the policies Rachael Rollins announced after winning the Democratic primary to be the next Suffolk County (Boston) DA was that she wouldn’t prosecute cases of breaking and entering into unoccupied places where no damage was done. Her logic was that she didn’t want to effectively criminalize homelessness. These cases often involved people who broke into places like construction sites to seek shelter at night. So Rollins’ policies meant that she was effectively announcing that the Suffolk County jail was no longer Boston’s backup homeless shelter.
Which, policy-wise, is surely a great idea. As the never-ending crisis at Rikers tells us, jails are terrible places for people to be, and stable housing likely reduces offending and victimization but provides other benefits as well. But the Rollins proposal would have been far better coming from the mayor. Rollins had no ability to expand access to homeless shelters or build affordable housing or shape eviction policies. She could close the door to the county jail—but she couldn’t open up any other doors.
That’s politically risky. It’s possible that her position would force the mayor, the City Council, and state officials to adapt by building public housing or shelters or creating vouchers. But it’s also possible that they would hope Rollins would bear the brunt of voters’ anger while they dragged their feet, lose her next election, and be replaced by a prosecutor who would embrace the old status quo.
And it’s not just housing. Studies show that better lighting leads to less crime. That falls under the control of the mayor. So do school skills programs and summer jobs. Caseworkers for those who are homeless. Civic, noncriminal drug treatment. Lead abatement. Mayor, mayor, mayor (and governor, and president, but not the DA).
Which means that the all-too-common framing—that the recent electoral success of moderate Democrats is somehow a general referendum against progressives—fundamentally misstates the issue. This is not an either-or situation, but more a both-and: Voters likely want moderate mayors and progressive prosecutors. After all, two prominent progressive prosecutors—Larry Krasner in Philadelphia and Kim Foxx in Chicago—won contested elections amid the rising violence of 2020. And in Manhattan in 2021, Alvin Bragg defeated several more punitive candidates in the primary and rejected his Republican opponent in the general.
The split between progressive prosecutors and more punitive mayors is rooted in real schisms permeating the criminal legal morass, but it comes with real costs for advancing reforms. Those with the greatest ability to put forth empirically supported alternatives to policing appear to be least willing to do so.