The First Step Act (First Step) offers people behind bars the opportunity to earn credits toward early release from Federal Bureau of Prisons (BOP) institutions.

People who haven’t been convicted of one of a long list of offenses are eligible to earn credits for early release by successfully completing evidence-based educational, employment, or substance use treatment programs or by engaging in other productive activities. Many of the details about these programs and activities are still to be defined by BOP, and how BOP chooses to implement them will ultimately affect the impact of the earned credit system.   

As required by First Step, the Department of Justice (DOJ) developed a new risk assessment tool called PATTERN, described in a recent report (PDF), to support the earned credit system. PATTERN considers several factors to determine whether a currently incarcerated person is at high, medium, low, or minimum risk of reoffending.

These designations affect how much time people can earn and when accrued credits can be redeemed for early release. First Step states that only low- or minimum-risk people are eligible for early release from prison, either into home confinement, a halfway house, or community supervision. The law also allows people to become eligible for early release if approved by the prison warden, regardless of risk level *

About half of the population described in the DOJ report scored at high or medium risk. The report says that nearly everyone at these risk levels who is eligible to accrue credits should be able to move to low or minimum risk. Based on our understanding of the tool and BOP’s programming capacity, it could be challenging for everyone at high and medium risk to sufficiently reduce their risk levels.

We developed an interactive version of DOJ’s risk assessment tool that shows users how the tool would score people based on several inputs and what factors could increase or decrease a person’s score.

Applying the tool to a hypothetical person with a high-risk score

We also used the risk assessment tool to calculate the risk level for a hypothetical high-risk man. The table below shows his characteristics and how many points each of those characteristics adds or subtracts to his score.

“Static” factors are fixed, while “dynamic” factors can change over the course of a person’s incarceration. With a total score of 50, this person would need to reduce his score by at least 17 points to become eligible for early release.

Based on this tool, two factors that can reduce someone’s risk score are age (studies have shown that the risk of reoffending goes down as people get older) and participation in evidence-based programming, including drug treatment and education.

Meanwhile, any infractions for misconduct could increase a person’s score. Although this hypothetical person could reduce his score enough to become eligible for release, it would require completion of residential drug treatment, at least one technical or vocational course, and at least 10 other programs, as well as an absence of any infractions for misconduct. For example, under some circumstances, one infraction could wipe out the benefit of one program.

Whether he can feasibly meet these requirements will depend partly on how long it will take to complete enough programs and partly on the length of his remaining sentence. Alternatively, if his remaining sentence is long enough, his score would decrease automatically as he got older.

Our analysis raises questions about whether people presently assessed as high risk can actually—as opposed to theoretically—move from high- to low-risk designations.

How easily can people with high-risk scores complete evidence-based programs?

Based on the report, currently about half of people exiting BOP institutions have not completed any programs, and only 5 percent have completed more than 10. We don’t know what or how many programs will qualify toward earning time credits (BOP will release a list of programs by early next year), whether BOP will have the capacity for everyone interested in programming to participate, how it will prioritize programming for high- and medium-risk people (as required by First Step), or how long it will take to complete each program.

All these decisions will affect the feasibility of people assessed at high- and medium-risk reducing their scores. For example, would our hypothetical person be able to enroll in and complete the 10 programs he would need to reduce his score in addition to completing a technical or vocational course and residential drug treatment?

How would this system impact people of color?

The effect of this system on racial groups will largely depend on its implementation. According to the report, African American men are more likely to score as high-risk with this tool than white or Hispanic men—likely because of more extensive contact with the criminal justice system, and thus, higher criminal history scores.

If the program capacity is large enough and the high-risk group is prioritized for program participation, the system could have a significant favorable impact on African American men’s risk scores. But if the high-risk group lacks access to sufficient programming to significantly lower their scores, the system could disproportionately disadvantage them.

After developing this new risk assessment tool, BOP will need to administer it to the entire federal prison population. They will also need to make a number of decisions about how they will implement the earned time credit system, along with several other provisions in the law. DOJ has solicited public comments about its new risk tool and may modify the tool based on that feedback. Continued transparency about the implementation process will allow the public to determine who in BOP will benefit from First Step and how.