Gauging Detention Dosage: Assessing the Impact of Pretrial Detention on Sentencing Outcomes Using Propensity Score Modeling

Problem

The problem the study aimed to address: 

The study investigates whether pretrial detention impacts sentencing outcomes, specifically focusing on the probability and length of incarceration sentences. This research addresses concerns that pretrial detention may inadvertently contribute to harsher sentencing, such as increased odds of incarceration and longer sentence durations.

General impact on the system and/or public: 

The findings have broad implications for justice reform, particularly in reducing reliance on incarceration and mitigating biases in sentencing. The study supports efforts to refine pretrial practices, which can significantly impact defendants' lives and systemic equity.

Research Questions:

1. Does pretrial detention increase the likelihood of incarceration (prison or jail)?
2. Does the "dosage" or duration of pretrial detention affect sentencing outcomes, including incarceration likelihood and sentence length?
    

Method and Analysis

Program Evaluated/Gaps Addressed: 

The study fills methodological gaps in prior research by using quasi-experimental design and propensity score modeling (PSM) to assess the impact of pretrial detention. It also explores the nuanced effects of detention duration (dosage) beyond binary detention status.
 

Data and Sample Size: 

The research analyzed a sample of 3,390 criminal defendants from nine counties in Oregon, using administrative data linked across court, jail, and correctional systems for 2016–2017 felony cases.
 

Analysis Used:

The study applied PSM techniques—Inverse Probability of Treatment Weighting (IPTW) and Marginal Means Weighting through Stratification (MMW-S)—to simulate random assignment and mitigate selection bias. Double-robust regression models further controlled for legal and extralegal factors.
 

Outcome

Key Findings:

1. Likelihood of Incarceration: Pretrial detention more than doubles the likelihood of receiving an incarceration sentence. Fully detained defendants had odds 2.2 times higher than those released to receive incarceration and 2.6 times higher for prison sentences specifically.
2. Detention Dosage Effects: Longer detention durations increased the likelihood of incarceration but did not significantly influence sentence length after accounting for time served.
3. Extralegal Factors: Male defendants were significantly more likely to receive incarceration sentences. High-risk individuals, as per the Public Safety Checklist, were also disproportionately affected.

Implications or Recommendations: 

The findings advocate for reforms to reduce the use and duration of pretrial detention. Key recommendations include:

• Exploring alternatives to detention, such as supervised release.
• Addressing potential systemic biases that lead to inequitable outcomes.
• Further research into judicial decision-making processes and implicit biases in pretrial and sentencing practices.

This summary highlights critical elements relevant to practitioners and policymakers, emphasizing the systemic and individual-level impacts of pretrial detention practices.