CJA’s updated Release Assessment is consistent with the goals of bail reform
it recommends release on recognizance for the vast majority of individuals who are likely to attend all their court appearances.
- Does the updated Release Assessment accurately classify the likelihood of court appearance?
- How often were individuals recommended for release on recognizance?
- Are there racial/ethnic variations in the release recommendations?
- How often do judges’ decisions align with the Release Assessment recommendation?
- Are there racial/ethnic variations in judges’ release decisions?
See How The Release Assessment Is Calculated
This interactive tool shows the eight factors used in the Release Assessment, factor by factor. Read the reasoning behind each factor, how each factor is scored, and how the final score guides the release recommendation.
This online version of the Release Assessment is for illustrative purposes only. Scores on the official Release Assessment administered for the courts may differ.
This Release Assessment and its scoring rubric apply in most cases for which an interview is conducted. In some situations, particular criminal charges, histories, and other factors may impact this process. In those cases, CJA may not make a recommendation or may not conduct an interview.
In figure 4A, rates of Remand were low enough as to barely appear on the graph. The amounts are: 0.2% of Total/All Recommendations, 0.1% of Recommended for ROR, 0.6% of Not Recommended for ROR.
In figure 4B, rates of Remand were low enough as to barely appear on the graph. The amounts are: 0.6% of Total/All Recommendations, 0.5% of Recommended for ROR, 0.8% of Consider all Options, 0.8% of Not Recommended for ROR.
In figure 4C, rates of Remand were low enough as to barely appear on the graph. The amounts are: 1.5% of Total/All Recommendations, 1.3% of Recommended for ROR, 2.3% of Consider all Options, 2.4% of Not Recommended for ROR.
In figure 5, rates of Remand were low enough as to barely appear on the graph. The amounts are: 0.5% of Black and Black Hispanic, 0.5% of Non-Hispanic White, 0.4% of Hispanic White and Hispanic Other.
The following principles guided the research: the assessment must be evidence-based and informed by data science; it should be developed in conjunction with judges, court actors, advocates, and affected communities and individuals; and it must be transparent and validated.
For more info on the release assessment
To learn more about the process of building the updated Release Assessment, see the full research report here.