Assessing the Impact of Pre-Adjudication Assessment Approaches on Racial/Ethnic Disparities in Oregon

Problem

The problem addressed: 

The study examines the impact of pre-adjudication risk assessments (PAAs) on racial/ethnic disparities in criminal sentencing outcomes in Oregon. Despite widespread use of risk assessments post-adjudication, there was limited research on their role before court decisions.

General impact on the system and/or public: 

PAAs have potential as tools to guide sentencing decisions and reduce prison populations. However, concerns arise about exacerbating racial/ethnic disparities due to biases embedded in risk assessment tools, often related to proxies like criminal history.
 

Research Questions Answered:

1. How have PAAs impacted racial/ethnic disparities in sentencing outcomes?
2. Do different PAA implementation methods affect disparities?
3. How do PAAs influence case negotiations and decision-making, and are they applied differently across racial/ethnic groups?
4. Are there predictive biases in PAA tools, and do these impact disparities?
   
    

Method and Analysis

Program Evaluated: 

The study analyzed PAAs used in two Oregon counties:

1. Multnomah County: Employed a formal judicial settlement conference.
2. Yamhill County: Relied on probation recommendations without mandated conferences.
    

Data and Sample Size: 

• Multnomah County: Data from 3,930 PAA cases and 1,153 historical controls.
• Yamhill County: Data from 290 PAA cases and 2,986 historical controls. 

Data included demographic details, risk assessments, case outcomes, and sentencing decisions.
 

Analysis Used:

• Quantitative Analysis: Quasi-experimental design using Propensity Score Matching (PSM) to compare sentencing outcomes and racial disparities.
• Qualitative Analysis: Interviews with 75 stakeholders (judges, prosecutors, defense attorneys, and probation officers) and observations of judicial conferences.

Outcome

Key Findings:

• Impact on Disparities: PAAs did not exacerbate racial/ethnic disparities in sentencing outcomes.
  • In Multnomah County, PAAs significantly reduced prison sentences across racial groups without differential effects.
  • In Yamhill County, PAAs had minimal impact on prison likelihood but slightly increased sentence length.
• PAA Implementation Differences:
  • Multnomah County’s approach, incorporating judicial settlement conferences, was more effective at reducing incarceration rates.
  • Yamhill County’s reliance on probation recommendations led to inconsistencies, with many recommendations for community supervision not followed.
• Predictive Bias: There was no evidence that PAAs increased racial bias. The observed racial disparities predate PAA use, reflecting broader systemic issues.
• Stakeholder Perceptions: The risk score itself was not a primary factor in decision-making. Instead, judges and attorneys emphasized qualitative factors such as life narratives and criminogenic needs.
  
   

Implications or Recommendations: 

• Policy Implications: PAAs can reduce overall incarceration rates without worsening racial disparities, particularly with structured implementation like judicial conferences.
• Further Research: Examine the role of qualitative factors in sentencing and how disparities persist despite equitable PAA applications.
• Practice Recommendations: Enhance training for stakeholders to focus on dynamic needs and ensure consistent application of probation recommendations.

This structured approach to pre-adjudication assessments offers insights into reducing incarceration equitably while addressing persistent systemic disparities.