Prosecutorial Reform and Local Crime Rates

Prosecutors often have substantial discretion in enforcing criminal laws in their jurisdictions, creating various pathways through which changes in prosecutorial leadership may influence case outcomes and public safety. Although other factors affect case outcomes—including police discretion in making arrests, defendants’ choices to participate in diversion programs or accept a plea bargain, and judges’ decisions about pretrial detention and sentencing—prosecutors nonetheless influence outcomes at several stages throughout a case. After an arrest, prosecutors must decide which charges, if any, to file and whether to request pretrial detention, offer diversion and/​or a plea bargain, and drop charges at any point before trial. If a conviction occurs, prosecutors usually make sentencing recommendations to judges.

Our research uses variation in the timing of when reform prosecutors took office across the country to estimate the resulting changes in criminal case outcomes, local reported crime rates, and drug overdose mortality rates between 2005 and 2023. Our findings reveal that reform prosecutors typically reduced per capita charge and conviction rates shortly after taking office, especially for lower-level offenses. Specifically, the induction of a reform prosecutor reduced convictions per capita by about 23–25 percent on average in their jurisdiction. Our findings are consistent with some of the reforms on which these prosecutors typically campaign, such as presumptive declination policies for certain minor offenses and increased use of pretrial diversion programming.

Our study also analyzes the effects of reform prosecutors separately for misdemeanors and felonies. Our findings suggest that the induction of a reform prosecutor reduced the number of misdemeanor charges per capita by 19 percent. Charge rates appear to have decreased across all misdemeanor types, though the evidence is strongest for a 42 percent decrease in drug charges. Similarly, reform prosecutors reduced the share of charges resulting in misdemeanor convictions in their jurisdictions by 17–24 percent. The reduced number of misdemeanor charges per capita and the reduced share of charges resulting in conviction led to a 23 percent reduction in misdemeanor convictions per capita. Our estimates suggest that these prosecutors reduced convictions for all misdemeanor types, with the largest reduction being for drug offenses at 39 percent. Convictions for violent misdemeanors also decreased by 26 percent, which appears mainly driven by a decrease in the share of charges resulting in conviction rather than the charge rate.

Furthermore, our research finds that the induction of a reform prosecutor reduced the number of felony charges per capita by 22 percent. Consistent with our misdemeanor findings, the evidence suggests that charge rates decreased for all offense types, with the largest decrease being for drug offenses. The share of felony charges resulting in conviction also fell by 5 percent, mostly due to fewer drug convictions. A 57 percent reduction in convictions per capita for drug offenses, along with the overall reduction in felony charges per capita and felony charges resulting in conviction, reduced the number of felony convictions per capita by 25 percent. However, the conviction rate fell for all offense types, including violent felonies, which decreased by 25 percent.

These substantial reductions in legal sanctions did not appear to cause similarly large increases in reported crime rates. Our findings indicate that reform prosecutors may not have affected reported crime. Specifically, we can rule out increases of more than 10 percent for property crimes and violent crimes and increases of more than 19 percent for drug offenses. One limitation of our analysis, however, is that we were largely unable to disentangle the effects on underlying crime rates from the effects on police officers’ recording behavior and civilians’ reporting behavior. To address this concern, we estimated effects on local drug mortality rates because this measure of drug use does not rely on police recording. Our research finds no convincing evidence of an effect on drug mortality rates. There is some evidence of a modest increase, but these estimates are not precise enough to rule out no effect.

Additionally, our research finds no convincing evidence that reform prosecutors decreased prison incarceration rates. Nevertheless, prior research suggests that forgoing charges and convictions may benefit defendants without harming public safety. For example, misdemeanor charges and convictions tend to increase the likelihood of rearrest, especially for defendants without a prior criminal record. Likewise, for felony defendants, noncarceral convictions increase rearrest rates relative to nonconvictions. Furthermore, prior charges and convictions may result in harsher sanctions for future offenses, including increases in incarceration.

Moreover, a growing body of research reveals that criminal charges and convictions have substantial collateral consequences. For instance, employers are less likely to return calls from applicants who report a past charge or conviction, and convictions can substantially reduce long-term earnings for reasons other than incarceration itself and can persist even after a conviction has been expunged. Legal sanctions can also cause other economic hardships for defendants and their families, such as exclusion from social safety net programs, increased debt from fines, difficulty securing rental housing, and reduced access to higher education. Forgoing certain charges and convictions is likely to shield some defendants from these consequences.

Bringing fewer cases may have additional benefits for the justice system, such as reducing court-appointed attorneys’ caseloads; this, in turn, reduces the pressure on these attorneys and may lead to better outcomes for defendants. Therefore, reform prosecutors’ less punitive approach may benefit both defendants and the justice system as a whole without imposing significant adverse effects on public safety. Future research should examine how far leniency can extend—and for which crimes—before reduced deterrence begins to harm public safety.

Note
This research brief is based on Amanda Agan et al., “Prosecutorial Reform and Local Crime Rates,” GitHub, May 5, 2026.