How Does Electing a Reform Prosecutor Affect Charging and Conviction Rates?

Agan et al. (2026) examined whether electing a reform-minded prosecutor changes misdemeanor and felony charging and conviction rates. They analyzed county-level administrative criminal justice data from the Criminal Justice Administrative Records System (CJARS) covering 25 jurisdictions, using variation in the timing of reform prosecutors taking office and a staggered difference-in-differences design. They found that reform prosecutors reduced misdemeanor convictions by about 258 per 100,000 residents annually (17%) and misdemeanor charges by about 443 per 100,000 residents (12%). Drug misdemeanor charges declined by about 25%, while felony convictions fell by about 8%, with no significant change in felony charging rates.

Why This Article Was Selected for The Policy Scientist

The policy choices made by elected prosecutors shape how criminal justice systems operate, influencing the number of individuals who enter and remain in the system. As more jurisdictions have elected reform-minded prosecutors, rigorous evidence about their effects has become increasingly important for both scholars and policymakers. Agan and her colleagues have been leading contributors to this emerging literature, making this study a timely extension of their previous work. Published in AEA Papers and Proceedings, a highly respected American Economic Association journal, the article combines high-quality administrative CJARS data with a strong staggered difference-in-differences causal inference design. Although the jurisdictions studied are not fully representative of the United States, the findings likely generalize to many large counties with similar prosecutorial structures. The study also builds incrementally on influential prior research examining prosecutorial discretion while providing broader evidence across multiple jurisdictions rather than focusing on individual prosecutors.

Full Citation and Link to Article

Agan, A., Doleac, J. L., Harvey, A., Kyriazis, A., & Schechter, L. (2026). Reform prosecutors and the footprint of the criminal justice system. AEA Papers and Proceedings, 116, 388–392. https://doi.org/10.1257/pandp.20261051

Central Research Question

This study examines whether the election of reform-minded prosecutors changes the size of the criminal justice system by altering misdemeanor and felony charging and conviction rates. Reform prosecutors generally seek to reduce reliance on criminal prosecution through greater use of prosecutorial discretion, diversion programs, and alternatives to traditional criminal sanctions. While these policies have attracted substantial political and scholarly attention, most previous evidence has focused on individual prosecutors or specific jurisdictions. The authors therefore ask whether reform prosecutors produce consistent changes across multiple jurisdictions. More specifically, they investigate whether the inauguration of a reform prosecutor reduces criminal charges filed, lowers conviction rates conditional on charges, and ultimately decreases convictions per capita. The study also examines whether these effects differ between misdemeanor and felony offenses and whether certain offense categories, such as drug crimes, experience larger changes than others.

Previous Literature

The literature on prosecutorial discretion has expanded rapidly over the past decade as reform prosecutors have become increasingly prominent in many American jurisdictions. Earlier research has demonstrated that prosecutors exercise considerable influence over criminal justice outcomes because they determine whether charges are filed, what charges are pursued, and how cases are resolved. Much of the existing empirical evidence has concentrated on high-profile prosecutors serving in individual jurisdictions. Studies of prosecutors such as Rachael Rollins in Massachusetts and Larry Krasner in Philadelphia found substantial changes in charging practices for lower-level offenses after they assumed office. Additional work examining progressive prosecutors in Florida reported reductions in incarceration and racial disparities associated with prosecutorial policy changes.

This study builds upon that literature by shifting attention from individual case studies to broader patterns across multiple counties. Rather than asking whether one particular prosecutor affected criminal justice outcomes, the authors investigate whether reform-minded prosecutors consistently produce measurable changes after taking office across many jurisdictions. This broader approach improves understanding of whether previously observed results reflect isolated local circumstances or represent a more general pattern. The study therefore extends an emerging literature while complementing earlier work on prosecutorial discretion and criminal justice reform.

Data

The analysis combines two primary sources of information. First, the authors obtained a list of reform-minded prosecutors from a nonprofit organization that coordinates a network of prosecutors committed to reducing the footprint of the criminal justice system. These prosecutors vary in the specific reforms they implement, but they generally share similar policy objectives emphasizing alternatives to prosecution and reduced criminal system involvement.

Second, the study relies on administrative data from the Criminal Justice Administrative Records System (CJARS) Justice Outcomes Explorer. This database aggregates criminal justice records from numerous jurisdictions throughout the United States and provides annual county-level information on criminal charges and convictions. The authors construct measures of misdemeanor and felony charges per 100,000 residents, conviction rates conditional on charges, and convictions per 100,000 residents. They also examine offense-specific categories, including drug, property, and violent crimes.

Because complete administrative records are not available for every county and year, the analysis includes separate samples for misdemeanor and felony outcomes. The misdemeanor analysis contains 24 counties that eventually elected reform prosecutors, while the felony analysis includes 21 treated counties. Hundreds of additional counties serve as comparison jurisdictions. The CJARS database represents one of the strongest currently available administrative data sources for evaluating criminal justice policy because it provides standardized measures across numerous jurisdictions while minimizing reporting bias commonly associated with survey data.

Methods

The authors employ a staggered difference-in-differences research design, one of the strongest quasi-experimental approaches available for evaluating policy interventions when randomized experiments are not feasible. Reform prosecutors assumed office at different times across jurisdictions, creating variation that allows the researchers to compare changes occurring after each inauguration with contemporaneous changes in jurisdictions that had not yet elected a reform prosecutor or never did.

The analysis follows the modern difference-in-differences framework developed by Callaway and Sant’Anna, which addresses several methodological limitations associated with traditional two-way fixed effects estimators when treatment timing varies across units. County fixed effects account for stable differences among jurisdictions, while year fixed effects capture national changes affecting all counties simultaneously. Population weighting ensures that estimates appropriately reflect differences in county size. Standard errors are clustered at the county level and estimated using a wild cluster bootstrap procedure to improve statistical reliability given the relatively small number of treated counties.

The authors also examine pretreatment trends through event-study analyses and report that treated and comparison counties exhibited similar outcome trajectories before reform prosecutors assumed office. This evidence strengthens the credibility of the causal interpretation by supporting the parallel trends assumption underlying the research design. Although the study cannot distinguish whether observed changes arise from prosecutorial decisions, policing behavior, or changes in criminal activity itself, the identification strategy provides credible evidence regarding the overall effects of reform prosecutor elections on criminal justice outcomes.

Findings/Size Effects

The most substantial effects appear in misdemeanor case outcomes. After reform prosecutors assumed office, misdemeanor charging rates declined by approximately 443 charges per 100,000 residents annually, representing roughly a 12 percent reduction. Misdemeanor convictions declined even more substantially, falling by approximately 258 convictions per 100,000 residents each year, or about 17 percent. These estimates indicate that reform prosecutors were associated with meaningful reductions in misdemeanor involvement within the criminal justice system.

The evidence suggests that much of the reduction in misdemeanor convictions resulted from fewer charges being filed rather than major changes in conviction rates after charges were initiated. Although the estimated reduction in the proportion of misdemeanor charges resulting in conviction was negative, it did not reach conventional levels of statistical significance.

The largest offense-specific changes occurred for misdemeanor drug offenses. Drug misdemeanor charges declined by approximately 164 cases per 100,000 residents annually, representing nearly a 25 percent reduction. By comparison, estimated decreases in misdemeanor property and violent offense charges were relatively small and statistically insignificant.

Felony outcomes exhibited more modest changes. Felony convictions declined by approximately 52 convictions per 100,000 residents annually, corresponding to roughly an 8 percent reduction. However, overall felony charging rates showed no statistically significant decline, and offense-specific felony categories likewise produced relatively small and statistically insignificant estimates. These findings suggest that reform prosecutors concentrated their largest measurable effects on lower-level offenses rather than serious felony prosecutions.

Across nearly every outcome examined, estimated effects moved in the direction of reduced criminal justice system involvement. The consistency of these estimates across multiple measures strengthens confidence that reform prosecutors generally reduced the footprint of the criminal justice system, particularly for misdemeanor offenses.

Conclusion

This study provides one of the first multi-jurisdiction evaluations of reform prosecutors using modern causal inference methods and comprehensive administrative criminal justice data. Rather than focusing on a single prosecutor or jurisdiction, it demonstrates that the election of reform-minded prosecutors is associated with systematic reductions in misdemeanor charging and conviction rates across numerous counties. The largest reductions occurred in misdemeanor drug cases, while felony outcomes changed more modestly.

Several features make the study particularly valuable. The CJARS administrative database provides high-quality, objective measures of criminal justice outcomes across many jurisdictions. The staggered difference-in-differences design represents a rigorous quasi-experimental approach that substantially strengthens causal inference compared with conventional observational regression analyses. Although randomized controlled trials are generally considered the strongest research design, they are not practical for elections or prosecutorial appointments. Consequently, the research design employed here represents an appropriate and methodologically sophisticated alternative.

The findings are most directly applicable to counties with prosecutorial structures and criminal justice systems similar to those included in the sample. Although the participating jurisdictions are not fully representative of every county in the United States, the consistency of the estimated effects across multiple jurisdictions suggests that similar patterns may emerge elsewhere under comparable institutional conditions. Overall, the study significantly expands empirical knowledge regarding prosecutorial discretion and provides stronger evidence about the system-wide consequences of electing reform-minded prosecutors than previously available.