Nationwide, people of color experience unequal treatment at all levels of the criminal justice system:
There’s a lot of work to do to make the system fairer. Over the years, policies have been introduced to support incarcerated people and set them up for success. But evidence shows that these improvements are not made available in an equitable way. Fortunately, Recidiviz is working with state departments of corrections to ensure that these solutions truly help everyone. By combining a simple analysis with a technology tool designed to put the relevant information into the hands of decision-makers, it is possible to chip away at racial disparities across the criminal justice system.
When creating new policies and programs, the first step that corrections leaders take is to develop criteria to determine who will be eligible for which opportunity. In evaluating the potential criteria, agencies can use eligibility analyses to examine how these policies and programs may affect different groups of people, with a particular eye toward racial disparities.
Historically, it has been difficult for corrections leaders to analyze the effect of each step in the eligibility funnel. Why? Because the data that determines eligibility for every individual resides in multiple data systems. For example, determining eligibility based on a risk assessment may require accessing and pulling data from one database, while looking at outstanding fines and fees or length of time on supervision may require information from another.
This analysis is critical when determining whether a newly-proposed program may have unintended racial disparity outcomes. Suppose there are questions about how many people of color on parole are eligible for a new program created by a state legislature. Right now, agencies would not be able to answer that question due to the limitations of their existing technology. Recidiviz solves this challenge by making it possible to analyze eligibility funnels by matching and stitching together multiple datasets and putting all of that information into the hands of agency research teams. That view of every step in the funnel provides insight into program access, enrollment, and outcomes, including where racial disparities exist.
Let’s look at an example of how this works when evaluating a policy meant to support individuals on probation or parole. Probation and parole officers (POs) often have caseloads of more than 70 people – and as a result, end up spending a lot of time with people who are already succeeding on supervision. This creates two issues. First, POs spend less time with the people on their caseload who really need support. And second, they over-supervise low-risk people, which can actually increase re-arrests and re-incarceration – doing more harm than good. To better support staff and people on parole, the Tennessee Department of Correction (TDOC) introduced a more flexible form of community supervision called compliant reporting. Moving people to this less intensive, virtual form of supervision saves staff time, rewards the people already succeeding, and directs more resources toward the clients most in need of attention.
But qualifying for compliant reporting isn’t easy. People in Tennessee who are on supervision must meet 13 distinct eligibility criteria, ranging from being on medium supervision or less for over 18 months to passing a drug screen within the past year to having fines of less than $2,000. Each of these steps in the eligibility criteria funnel introduces the potential for unintended racial disparities.
To help TDOC look for these unintended consequences, Recidiviz built a tool that made it possible to simulate the effect of every step of the eligibility funnel, looking specifically for variation by race. In the image below, you’ll see that for one step of that funnel (the fines and fees requirement) 62% of the supervision population in Tennessee is white, while 68% of those who meet the requirement are white. This shows that this particular step disproportionally disfavors people of color. Looking at the entire funnel, criteria by criteria, it becomes possible for corrections agencies to pinpoint which specific conditions are leading to racial disparities. After uncovering inequality, corrections leaders can then make a plan to address the imbalance. Due to the fines and fees disparity discovery, for example, TDOC is currently evaluating potential ways to reform the fines and fee requirements.
In addition to fines and fees, Recidiviz has also helped the state look at eligibility based on offense type. TDOC’s requirement excluded anyone with a history of driving under the influence, domestic assault, or crime against a person resulting in bodily harm – including if these offenses were historical, having been committed and served years or decades prior. By analyzing the eligibility funnel to see who would be eligible for compliant reporting based on their current rather than their historical offense, TDOC realized that this requirement led to fewer people of color getting access to the compliant reporting opportunity. By looking closely at how to evaluate eligible offense types, TDOC realized it was possible to reduce racial disparities by focusing only on the current offense – and ultimately adjusted the policy accordingly.
Before evaluating every step in the eligibility criteria funnel, TDOC and Recidiviz could see racial disparities in the overall outcome of who was placed on compliant reporting, but couldn’t see why and which steps caused those disparities to occur. Now, TDOC can find and address what’s driving disparities, and change policy and practice accordingly.
“Now, we’re applying what we learned with Recidiviz’s eligibility analysis to improve administrative policy department-wide,” says TDOC district director for community supervision Joshua Graham, who leads 70 employees in charge of more than 5,000 people on probation and parole.
Analyzing the eligibility criteria for policies and programs designed to support incarcerated people is one effective way to ensure the system is more equitable. Better data and tools help practitioners visualize which parts of the system are leading to these disparities, and how to make them fairer over time. Now, Tennessee’s data-driven approach is inspiring other agencies to work toward more equitable outcomes for all people.