NEW YORK—Prosecuting attorneys have the greatest latitude in the U.S. justice system in decision making that influences criminal case outcomes. They play their pivotal role with little oversight from the public, the press, or members of the judiciary—prompting some observers to refer to the “black box” of prosecutorial decision making.

New research from the Vera Institute of Justice, with support from the National Institute of Justice, provides an unprecedented look inside the workings of prosecutors’ offices to understand how prosecuting attorneys make choices. The Anatomy of Discretion: an Analysis of Prosecutorial Decision Making goes beyond previous research that mainly explored relationships between specific case-level factors and case outcomes. It examines the processes that prosecutors use to reach their decisions and the contextual factors that influence those decisions—prosecutors’ characteristics, such as gender, experience, supervisory role, unit, caseload, and preferred plea bargaining strategies; office structures and policies; resource constraints; and social context (relationships among participants in the courtroom, for example). Furthermore, the new research explores how prosecutors’ conceptions of justice and fairness influence their decision making.  

The Institute’s research also provides a more integrated look at the impact of case-level factors across case processing stages than most prior studies, which have tended to focus on single decision points or a limited number of factors. Principal investigators Bruce Frederick, a senior research associate at Vera, and Don Stemen, an assistant professor in the Department of Criminal Justice and Criminology at Loyola University Chicago examined initial case screening and charging decisions, plea offers, sentence recommendations, and post-filing dismissals for multiple offense types. At each decision point, they analyzed the impact on case outcomes of legal factors such as crime seriousness, strength of evidence, and criminal history and other factors such as defendant’s age, gender, employment, parental status, substance abuse, mental health status, treatment history, victim-offender relationships, and defendant’s demeanor. 

“Prosecutors expressed strong commitments to the ideals of justice and fairness—to doing ‘the right thing’—and judged what would be just and fair primarily on the basis of strength of evidence, crime seriousness, criminal history, and the potential consequences of their decisions for victims and defendants,” said Bruce Frederick. “However, contextual factors such as office rules, resource constraints, and relationships with other actors sometimes led prosecutors to make compromises they would not consider to be fully consistent with those ideals.”

Four accompanying podcasts feature conversations with scholars and justice practitioners about the ramifications of the study’s findings, which include:

• Prosecutors’ decisions were guided by two basic questions: “Can I prove the case?” and “Should I prove the case?”
• Prosecutors were guided by an overarching philosophy of doing justice—or “the right thing.”
• Strength of evidence was the primary consideration at screening and continued to influence decisions throughout the processing of a case.
• Seriousness of the offense influenced decisions throughout the processing of a case.
• Victims’ characteristics, circumstances, wishes, and willingness to testify affected prosecutors’ evaluations of both the strength of the evidence and the merits of the case.
• Defendants’ personal characteristics and circumstances affected how prosecutors evaluated the fairness of potential case outcomes.
• Shortages of courtrooms, judges, clerks, court reporters, and scheduled court hours—and especially unscheduled reductions in court hours—often forced prosecutors to undercharge, reevaluate and change plea offers, or dismiss cases.
• A decrease in emphasis on investigations resulted in a decline in the quality of information coming from police departments and an increase in cases declined for prosecution, pended for additional information, or dismissed because of poor follow-up investigation.