Topic(s) Racial & Gender Justice, Case Processing

Download PDF
See Related Press and Media
Back to Case Processing

Alignment with New York City’s Pretrial Release Assessment: Results for the Five Boroughs

This report examines the alignment of New York City judges’ pretrial release decisions with the recommendations of the Pretrial Release Assessment, a validated tool that calculates the likelihood people will return to court if they are released before trial. Drawing on 251,917 New York City arraignments subject to pretrial release decisions in 2021, 2022, and 2023, we looked at what the assessment recommended, how often judges followed these recommendations, and cross-borough differences over time.

Key Findings:

  • Most People are Recommended for Release on Recognizance (ROR): From 2021 to 2023, the Release Assessment recommended releasing people on their own recognizance (ROR) for 88% of cases, including 79% of violent felonies, 77% of nonviolent felonies, and 92% of misdemeanors. Yet judges granted ROR in only 25% of violent felonies, 42% of nonviolent felonies, and 78% of misdemeanors.
  • The Assessment Was Consistent Across Each Race/Ethnicity: The assessment recommended 87% of Black, 88% of Hispanic, and 85% of white people for ROR. For people charged with a violent felony, the assessment recommended a statistically identical 78% of Black and Hispanic and 77% of white people for ROR.
  • Alignment with the Release Assessment’s Recommendations: From 2021 to 2023, judges infrequently followed ROR recommendations for violent felony cases (30%), only followed such recommendations about half the time (51%) in nonviolent felony cases, while adhering at a high rate for misdemeanors (83%). Conversely, in violent felony cases that are virtually all legally eligible for bail under the State’s bail reform law, judges set bail or remand in 41% of the cases where the Release Assessment had recommended ROR.
  • Judges Aligned with the Assessment at Racially Disparate Rates: When the assessment recommended ROR in a violent felony case, judges set it least often for Black people (26%), somewhat more for Hispanic people (32%), and most often for white people (43%). In these same cases, judges were more likely to impose bail or remand on Black (44%) than Hispanic (39%) or white people (29%). Further analysis linked these racial differences to a tendency of judges to overrate certain risk factors correlated with race/ethnicity.
  • Overrating Certain Risk Factors: Although criminal history and living situation are already factored into the assessment’s recommendations, judges were more likely to adhere to a ROR recommendation when people had no prior warrants, no prior misdemeanor or felony convictions, and a stable abode than when one of these risk factors was present.
  • Alignment Varied Across the City: Bronx and Brooklyn judges followed a ROR recommendation at the highest and Staten Island judges at the lowest rate. Nonviolent felonies saw especially wide borough variability. (For example, people facing a nonviolent felony charge and recommended for ROR were 1.8 times more likely to receive it in the Bronx than in Manhattan or Staten Island.)