Myths and Realities: Prosecutors and Criminal Justice Reform

Prosecutors play a vital role in the criminal justice system, determining not just which cases to pursue but also what charges to file and penalties to seek. Over the past decade, some prosecutors have developed approaches that aim to reduce racial and economic disparities and unjust outcomes in the legal system — such as excessive sentences or the criminalization of poverty through cash bail — while preserving public safety.

Far from a unified group, these prosecutors bring many different approaches to their work. They span the political spectrum and are found in urban and rural jurisdictions alike. They are often grouped under the label “progressive prosecutors” or “reform-minded prosecutors.” For the purposes of this analysis, we use the term “pro-reform prosecutors” to indicate chief district attorneys, county attorneys, commonwealth attorneys, and state attorneys who campaigned on or promised to reimagine the role of their office to broadly reduce unjust disparities in the justice system and decrease unnecessary incarceration.

Some critics have alleged that pro-reform strategies have driven increases in crime. We evaluate those claims, drawing on recent crime data and an understanding of how the criminal justice system works in practice, and find that these claims lack support. In fact, there is no evidence that pro-reform prosecutors are responsible for crime rising or falling.

First, we use data from the last decade, which includes the period of time during the Covid-19 pandemic, to evaluate how crime trends have changed in cities with and without pro-reform prosecutors. Previous research on this subject has, with some exceptions, found little to no relationship between the inauguration of a pro-reform prosecutor and a measurable increase in crime, even after using sophisticated statistical strategies.

Our analysis, described below, also finds no clear relationship between the pro-reform prosecutorial approach and the incidence of crime. Using data collected by the Council on Criminal Justice, we compared aggravated assault, larceny, and homicide trends in cities with pro-reform prosecutors to trends in cities without pro-reform prosecutors. Assault and larceny were selected because of their frequency, allowing clearer analysis, and because they are more likely to be affected by prosecutorial decision-making. Murder was chosen because of its seriousness and because those crimes spiked sharply during the first two years of the Covid-19 pandemic.

Researchers have previously relied on several criteria to define pro-reform prosecutors, including a focus on the use of diversion policies, increased resources devoted to police accountability units, use of data to mitigate racial bias, reduced support for excessively long punishments, reviews for claims of wrongful convictions, and limitations on the use of money bail. In this article, we build on these methods and select only those prosecutors who have been consistently categorized as pro-reform by other researchers in the field.

As illustrated below, trends are not meaningfully different between cities with pro-reform prosecutors and cities without pro-reform prosecutors. For each group, we plot the average percentage change in crime rates from a 2018 baseline. The effects of seasonal variation and the Covid-19 pandemic are broadly similar, and the net changes in aggravated assault, larceny, and homicide rates between 2018 and 2024 are remarkably alike. Neither group sees drastically better or worse outcomes.

In sum, this research undermines the idea that pro-reform prosecutors cause crime to increase. To be sure, our analysis does not account for differences in local conditions or contexts. For example, the cities in our sample with pro-reform prosecutors are generally larger than those with more traditional prosecutors. Yet the lack of differentiation between the two groups ultimately suggests that other variables, including economic, political, and social factors, may provide more plausible explanations for crime trends.

Jurisdictions that elected pro-reform prosecutors midway through our dataset were excluded from this first analysis. Looking at them individually, however, we see similar results. Below, we compare trends from each excluded city to aggregated trends from a comparison group of cities. Creating a comparison group that closely matches a city’s trends before a prosecutor is inaugurated allows us to evaluate how trends might diverge (or fail to diverge) after a prosecutor assumes office. In each figure, the shaded area represents the administration of the pro-reform prosecutor. Note that homicide trends are excluded because the relative rarity of the offense results in highly erratic patterns, making reliable comparisons between individual cities and small groups of cities difficult.

Notably, the graphs show that crime trends in pro-reform prosecutor jurisdictions largely match those in their comparison groups. Where they do not match, they indicate lower crime rates in cities with pro-reform prosecutors. That is not what we would expect to see if, as some critics claim, jurisdictions with pro-reform prosecutors experience rising or higher crime. In Los Angeles, a pandemic-era rise and plateau in aggravated assault rates is mirrored by trends in the comparison group of cities — both before and after the inauguration of pro-reform prosecutor George Gascón. In Austin, the decline in larceny rates overseen by José Garza, another district attorney elected on a reform platform, outpaces that of the comparison group. And assault trends in Boston are particularly notable. Aggravated assault rates began to decline sharply in the first year of then–District Attorney Rachael Rollins’s administration and have remained significantly below 2018 rates since then, even after Rollins’s departure in 2022. No other city or combination of cities in our sample could match Boston’s steep drop in aggravated assaults.

These analyses offer a new way of thinking about crime trends in cities with and without pro-reform prosecutors. And like most other research on this subject, they suggest that the relationship between pro-reform prosecutors and crime trends is limited, especially when weighed against the impacts of social and economic factors such as poverty, economic opportunities, and the general social instability caused by the Covid-19 pandemic.

These results should not be surprising. All prosecutors — of every political affiliation — must decide how to prioritize cases. That is what they are elected to do. No office has enough resources to prosecute every case, and offices must often make complicated decisions by balancing the resources they have and the needs of the particular community they serve. In doing so, prosecutors can (and should) weigh whether it is in the interest of justice to prosecute certain cases and whether public safety allows for, and the community would benefit from, alternative resolutions. Most prosecutorial agencies prioritize violent crime, particularly crimes committed by people with a history of justice system contact.

Some pro-reform prosecutors have attempted to balance these competing demands by determining, as a matter of policy and safety, that some violations of the penal code should not result in prosecution. Some such “declination” policies have drawn criticism, especially where they target broad categories of offenses. But these policies are exceptionally rare. When they exist they often focus on low-level offenses such as marijuana possession — which the public favors decriminalizing — while continuing to charge and prioritize violent crime. Another common declination policy focuses on not prosecuting doctors or patients for abortions, consistent with local community wishes not to criminalize health care. In New York City, former Manhattan District Attorney Cyrus Vance announced in 2018 that his office would decline to prosecute most cases of subway turnstile jumping unless there was a related public safety risk.

Indeed, declining to prosecute certain categories of crime often stems from prosecutors acknowledging that in these cases prosecution does not increase public safety. Declining to prosecute also does not preclude other forms of accountability. Often cases that are diverted from the criminal justice system still require defendants to complete a form of treatment or make amends for harm they have caused.

There is evidence that these policies work. Broadly speaking, a recent paper found no correlation between crime and non-prosecution rates. Turning to specific cases, a 2022 study of Suffolk County (Boston) found that the decision not to prosecute a certain type of case had wide-ranging positive effects by reducing future prosecutions and convictions — vastly shrinking both the short- and long-term footprint of the criminal justice system by averting the cascading impacts of misdemeanor prosecution on the person charged. At the same time, there was no evidence that the policy change had detrimental effects on public safety. In that paper, the researchers studied a policy implemented by then–Suffolk County District Attorney Rachael Rollins under which her office declined to prosecute many nonviolent misdemeanors, such as disorderly conduct and minor drug possession.

For another example, Baltimore State’s Attorney Marilyn Mosby adopted a policy of declining prostitution and lower-level drug offenses during the Covid-19 pandemic. Subsequent research showed very low recidivism among those who had their cases declined, and no increase in these offenses citywide. In 2019, Cook County State Attorney Kim Foxx instructed attorneys in her office not to prosecute people accused of driving with licenses that were suspended or revoked for financial reasons, like failure to pay child support, tolls, or parking tickets. This allowed prosecutors in her office to spend more of their time prosecuting serious and violent crimes.

Additionally, in Minnesota, Ramsey County (St. Paul) Attorney John Choi limited the use of non–public safety traffic stops and searches in 2021. This policy has since measurably reduced racial disparities in which drivers were pulled over and has not negatively impacted public safety.

Prosecutors’ offices do not singlehandedly reduce crime. Prosecutors occupy an inherently reactive role; they are tasked largely with prosecuting crime after it happens. Successful crime prevention strategies require input from community members, law enforcement, social service providers, elected officials, and faith-based groups, among others, all working together to improve public safety. For example, research indicates that addressing socioeconomic hardships in a community — through drug treatment, access to housing, or even simply creating green spaces and improving lighting in urban areas — helps prevent crime. Research also shows that increasing the likelihood that someone is apprehended, which prosecutors have little hand in, does more to deter crime than the severity of the prospective punishment.

To be sure, some pro-reform prosecutors have started to address the causes and conflicts behind crime. In New York City, investments into early-intervention diversion programs have been made with promising results. The Brooklyn District Attorney’s Office’s Project CLEAR (Collaborative Legal Engagement Assistance Response) offers clinical and community services for low-level drug possession cases before the accused ever see a courtroom. Similarly, in 2022, the Manhattan District Attorney’s Office established a $9 million initiative to invest in community-based peer navigators for people in need of housing, mental health care, and substance use support. The initiative aims to build public safety outside the criminal justice system by addressing basic needs through voluntary, personal engagement. New York is hardly alone in these initiatives.

However, it is rare for prosecutors’ offices to have the resources to invest in crime prevention in this way. To increase public safety, policymakers must invest in strategies that address root causes of crime and poverty to build up disinvested communities — prioritizing affordable housing, improving education, and increasing access to jobs.

While elected prosecutors can impact office policies, they are also part of a complex system that does not turn on a dime. Public safety impacts from changes in prosecutorial policies can take years to propagate through the system. In the meantime, many pro-reform prosecutors have faced substantial pushback once elected, underscoring how challenging the process of policy change can be. This pushback has taken many forms: some rank-and-file employees have declined to follow reform directives, for example; elsewhere, lawmakers and well financed opposition groups have called for impeachment inquiries or recall elections. State legislators have also worked to limit the discretion of prosecutors, and in some states, governors have gone as far as removing prosecutors whose policies they object to. These efforts not only undermine the office’s independence but also thwart the will of voters. They are also based on false assumptions about the impact of reform policies on crime rates.

Taken together, even when it occurs, an immediate uptick or downturn in crime rates correlated with the election of a prosecutor is unlikely to be the result of office policies. As shown, these rates trend nationally and are minimally affected by the tenure of progressive prosecutors — and when they are affected, they tend to be slightly lower in progressive jurisdictions.

Historically, prosecutors have won elections on rhetoric and promises centering increased punishment — the same policies that have disproportionately harmed low-income communities of color. Similarly, success for a prosecutor has typically been measured by the number of convictions and severity of the punishments imposed. These traditional priorities and policies have contributed to mass incarceration in the United States and have not meaningfully enhanced public safety.

Before criticizing pro-reform prosecutors who are attempting to create a more equitable justice system that also protects public safety, critics should familiarize themselves with the data to better understand that crime trends in prosecutors’ jurisdictions across the political spectrum tend to be about the same. Prosecutors have limited effect on crime rates, but their policies can mitigate or increase some of the harms caused by decades of over-prosecution.

Brennan Center researchers analyzed crime data collected by the Council on Criminal Justice for its year-end 2023 report on crime trends. The sample of 38 cities is not necessarily representative of national trends. Data for aggravated assault, larceny, and homicide were available for 25, 32, and 30 of the 38 cities, respectively. Within each charge type, city baselines were determined by their average rates across 2018 to account for differences in seasonal swings.

We categorized prosecutors according to an analysis of published literature. There are three recent papers that code whether prosecutors are “progressive”: Amanda Agan, Jennifer L. Doleac, and Anna Harvey, “Prosecutorial Reform and Local Crime Rates,” Law & Economics Center at George Mason University Scalia Law School Research Paper Series, No. 22–011 (October 2021); Nick Petersen, Ojmarrh Mitchell, and Shi Yan, “Do Progressive Prosecutors Increase Crime? A Quasi-Experimental Analysis of Crime Rates in the 100 Largest Counties, 2000–2020,” Criminology & Public Policy 23, no. 2 (May 2024): 459–490; and Thomas P. Hogan, “De-prosecution and Death: A Synthetic Control Analysis of the Impact of De-prosecution on Homicides,” Criminology & Public Policy 21, no. 3 (August 2022): 489–534. To determine our sample of pro-reform prosecutors, we selected only those prosecutors who were unanimously recognized in these papers as pro-reform (see below for list).

The comparison group for figures 1–3 comprises all prosecutors in our dataset who did not receive unanimous recognition as pro-reform in the recent studies. Comparison groups for figures 4–9 were composed by selecting comparison cities whose aggregated trends matched pre-treatment trends for each pro-reform prosecutor city. As a result, prosecutors may have enacted some pro-reform policies and still be placed in the comparison group. To check for robustness, researchers tested alternative coding rules. Neither recategorizing individual cities (“leave-one-out”) nor testing non-unanimous-recognition rules fundamentally changes the findings of this analysis. Details and results of these checks are provided in the published Appendix.

In some cases, a jurisdiction had a pro-reform prosecutor for only part of the sample period. If a city was covered by a pro-reform prosecutor for 54 of 72 months (75 percent) or more, it was coded as a pro-reform jurisdiction; for 18 of 72 months (25 percent) or less, it was coded as a control jurisdiction; and for 19–53 of 72 months, it was dropped from the main analysis sample. Austin, Boston, and Los Angeles were dropped for this reason. Ninety percent of city-months in the main pro-reform sample had a pro-reform prosecutor in office.

Additionally, most prosecutors are elected at the county level. Since data was available by city, researchers matched each city to its nearest county to assign a prosecutor. In cases where cities overlapped multiple counties, the county containing the plurality of the city’s residents was chosen.

For figures 1–3, the pro-reform prosecutor jurisdictions are Baltimore, Chicago, Denver, Philadelphia, San Francisco, Seattle, and St. Louis. The full list of comparison prosecutor jurisdictions includes Arlington (TX), Atlanta, Buffalo (NY), Chandler (AZ), Charlotte (NC), Chattanooga (TN), Cincinnati, Colorado Springs (CO), Dallas, Detroit, Lincoln (NE), Little Rock (AR), Memphis (TN), Minneapolis, Nashville (TN), New York City (all five boroughs), Norfolk (VA), Omaha (NE), Phoenix, Raleigh (NC), Richmond (VA), Rochester (NY), St. Paul (MN), St. Petersburg (FL), Syracuse (NY), and Washington, DC. For each charge, a small number of cities were dropped due to inadequate data availability.

For figures 4–9, researchers identified smaller comparison groups of three or four comparison cities by matching aggregated trends to pre-inauguration trends for the jurisdictions that elected a pro-reform prosecutor midway through the sample period. This matching serves only as a benchmark to illustrate how post-inauguration trends could have looked, given pre-inauguration trends.

For their review and guidance, we thank Mona Sahaf, director of Vera’s Reshaping Prosecution Initiative; Andrew Taylor, senior research associate with Vera’s Reshaping Prosecution Initiative; and Rachel Marshall, director of the Institute for Innovation in Prosecution at John Jay College. We also thank Ben Nyblade, Josy Hahn, Ruby Nidiry, and Josh Bell for their research and writing support and Maris Mapolski for her work fact-checking this analysis.