Providing Feedback to Judges in Texas and New York

Last registered on May 30, 2023


Trial Information

General Information

Providing Feedback to Judges in Texas and New York
Initial registration date
December 09, 2022

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
December 16, 2022, 4:39 PM EST

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Last updated
May 30, 2023, 1:40 PM EDT

Last updated is the most recent time when changes to the trial's registration were published.


Primary Investigator


Other Primary Investigator(s)

PI Affiliation
Harvard Law School

Additional Trial Information

In development
Start date
End date
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Racial disparities exist at every stage of the U.S. criminal justice system and are particularly prominent in the setting of bail. In settings like Harris County, TX, black defendants are more than 34 percent more likely to be detained compared to whites. In this project, we are testing the effectiveness of detailed private feedback and personalized tips to judges in New York State; Dallas County, TX; and Bexar County, TX. The private feedback will consist of the judges’ own outcomes over time, giving them the motivation, information, and tools necessary to reduce racial disparities in their pretrial decisions. We will estimate the causal effect of this intervention on pretrial release and misconduct rates using a randomized control trial.
External Link(s)

Registration Citation

Dobbie, Will and Crystal Yang. 2023. "Providing Feedback to Judges in Texas and New York." AEA RCT Registry. May 30.
Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information
Experimental Details


We have designed an intervention to improve decision-making and reduce racial disparities at the pretrial justice stage. We will provide judges with individualized feedback on their past performance, including their own release and good conduct rates, where the good conduct rate is defined as the percent of released defendants who appear for their scheduled court appearances and who are not rearrested for another crime pretrial in the three months following their initial bail hearing. We would start by providing historical information, before providing updated information on a regular basis (e.g., every 6 months) as the experiment progresses.
Intervention Start Date
Intervention End Date

Primary Outcomes

Primary Outcomes (end points)
Pretrial Release Rates, Pretrial Misconduct Rates, Racial Disparities in Pretrial Release Rates, Racial Disparities in Pretrial Misconduct Rates
Primary Outcomes (explanation)
We will evaluate the effect of our combined intervention on pretrial release rates, misconduct rates, and racial disparities in these outcomes.

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
We will conduct a randomized controlled trial (RCT) to identify the causal impact of our intervention on pretrial release rates and pretrial good conduct rates (on average and on racial disparities).

Randomization and Data Collection: We will use a two-step randomization plan. The first step is to randomize judges into two mutually exclusive groups following a short planning stage. The first group of judges will continue making decisions under the status quo and will serve as the control group for the study. The second group of judges will receive the intervention and will serve as the main treatment group for the study. Judges will remain in these two groups for twenty-four to thirty months in total, depending on final sample size, which will maximize our statistical power for estimating the impact of the full set of interventions on medium-run outcomes. Additional groups of judges who join the experiment at a later time period will also be randomized into the treatment or control group based on the two-step randomization plan. Half of the additional judges will be randomly assigned to the treatment group and the other half to the control group. In the case of an odd number, N, of additional judges and an odd number of original judges, ⌊N/2⌋+1 of the additional judges will be randomly assigned to the group with a smaller group of original judges and ⌊N/2⌋ of the additional judges will be randomly assigned to the larger group of original judges. Both during and after the intervention, we will obtain publicly-available administrative data on pretrial release decisions and good conduct outcomes at the case level.
Experimental Design Details
Not available
Randomization Method
Randomization will be conducted using a computer random number generator.
Randomization Unit
We will randomize at the level of the individual judge within each location.
Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters
Sample size: planned number of observations
Approximately 180 judges.
Sample size (or number of clusters) by treatment arms
Approximately 90 judges in the control arm and 90 judges in treatment arm.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

Institutional Review Boards (IRBs)

IRB Name
Harvard University Area Committee on the Use of Human Subjects
IRB Approval Date
IRB Approval Number
Analysis Plan

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information