A Report on Behalf of the American Bar Association Criminal Justice Section Task Force on the Implementation of the First Step Act

The First Step Act (FSA), signed into law on December 21, 2018, sought to improve criminal justice outcomes, reduce recidivism, and reduce the federal prison population.1 Since its inception, the FSA and the programs/tools it created have undergone both formal changes and evolutions due to the current pandemic. The ABA is following these developments in an effort to ensure the FSA’s intentions are actualized. In particular, the Criminal Justice Section of the ABA created a Task Force on the Implementation of the FSA shortly after its passage to monitor the progress of realizing the goals of the Act.2 Our Task Force includes a diverse cross section of defense attorneys, prosecutors, academics, as well as a judge. The Task Force offers this Report to summarize its efforts to date and to alert stakeholders to our concerns and unanswered questions regarding the implementation of the FSA. The Report offers a brief overview of the FSA, then focuses on a particular aspect of the Act—the Risk and Needs Assessment System—that has caused us the greatest concerns regarding whether the FSA has been implemented with fidelity to its Congressional purposes.The FSA includes several key parts: (1) amendments to certain sentencing provisions;3 (2) elimination of the Bureau of Prisons’ gatekeeping role for compassionate release petitions;4 (3) a requirement that the BOP create a risk and needs assessment system for tailoring programs to address incarcerated individuals’ needs and the provision of time credits for participating individuals that may result in earlier release from secure custody;5 and (4) creation of an Independent Review Committee to evaluate the implementation of the risk and needs assessment system.6The FSA amended several sentencing provisions, most notably reducing mandatory minimum penalties and changing when they apply.7 For defendants with prior serious drug or serious violent felonies under 21 U.S.C. § 841, the FSA reduced certain enhanced penalties for repeat offenders.8 Additionally, Section 402 of the FSA broadens the safety valve criteria of 18 U.S.C. § 3553(f), which allows more drug offenders to be relieved of mandatory minimums or receive a reduction in their sentences.9 The FSA also eliminated mandatory consecutive sentencing, or “sentence stacking” under 18 U.S.C. § 924(c) for offenses involving using, carrying, possessing, brandishing or discharging a firearm in furtherance of a violent or drug trafficking offense.10 Further, the FSA provided that application of the reduced penalties for crack cocaine offenses that were enacted under the 2010 Fair Sentencing Act11 could be applied retroactively to sentences imposed before passage of that Act.Under the FSA, incarcerated individuals may themselves file motions for compassionate release.12 The Sentencing Reform Act of 1984 established the opportunity for incarcerated individuals to seek compassionate release and allowed courts to grant reduced sentences for “extraordinary and compelling” reasons.13 However, under that Act, prisoners had to petition the BOP to advocate on their behalf for the reduction of their sentence through compassionate release.14 The FSA removed the gatekeeping role of the BOP.15As directed by the FSA, the Department of Justice (DOJ) developed a risk and needs assessment tool called the Prisoner Assessment Tool Targeting Estimated Risk and Need (PATTERN) in July 2019.16 The tool, which has since been revised, is intended to measure an incarcerated individual’s risk of recidivism and assess the needs of the individual to determine relevant programming.17 It is designed to predict the chance of both general and violent recidivism three years post-release.18 While incarcerated individuals have long been assessed at intake to determine needs, PATTERN formalizes and standardizes this process.19The tool assigns a risk level classification of minimum, low, medium, or high.20 So long as incarcerated individuals have not been convicted of a disqualifying offense,21 they are eligible for certain Evidence-Based Recidivism Reduction (EBRR) programs and Productive Activities (PA) to address their needs.22Incarcerated individuals who successfully complete EBRR programs and productive activities earn time credits, depending on their risk classification level.23 Medium and high risk individuals can earn 10 days of time credit for every 30 days of successful participation in applicable programming and activities, while those in minimal or low risk categories can earn an additional 5 days of time credits for every 30 days of participation.24 Individuals in the minimal and low risk categories can apply these credits to earn early pre-release custody or up to a 12-month sentence reduction.25 Incarcerated individuals in the medium and high risk categories can apply their credits for other benefits such as increased phone and visitation privileges or transfers to institutions closer to the individual’s release residence.26 In addition, all incarcerated individuals may lower their risk category through participation in EBRR programs and productive activities.27The FSA created an Independent Review Committee (IRC), comprised of outside experts, to aid the DOJ with the creation of the risk and needs assessment system as well as the programming to be offered pursuant to these assessments.28 Under the FSA, the IRC is required to submit a report regarding the effect of the BOP’s risk and needs assessment system under § 107 of the FSA.29 In this year’s report, the statistical information was current as of September 26, 2020; the changes are in their infancy and to date there are no data on long-term post-release recidivism.30The DOJ set to work on the development of PATTERN in early 2019, and in April and May of 2019 it hosted “listening sessions” with interested stakeholders. Our Task Force participated in the May listening session, although our ability to provide input was limited because no information had yet been released regarding the DOJ’s work. In general, we pressed for transparency in the data and process to be used to build the tool and expressed a concern that the tool should avoid racially disparate impacts.On July 19, 2019, as mandated by the FSA, the DOJ released its initial report unveiling PATTERN.31 The report described the development of PATTERN and some of the considerations and decisions that underly its creation.32 The report did not release any of the data used to create the tool, but did provide the factors to be scored and the scores for each factor.33 The factors for both men and women included the following: (1) age at first conviction; (2) age at the time of assessment; (3) infraction convictions within the last ten years; (4) the number of technical and vocational programs completed; (5) drug treatment and education while incarcerated; (7) whether the instant offense was violent; (8) whether the inmate is a sex offender; (9) criminal history; (10) history of violence; (11) history of escape; (12) whether the defendant was permitted to voluntarily surrender after sentencing; and (13) the inmate’s education.34 Additionally, for women only, the factors included: (14) federal prison industry employment; and (15) non-compliance with financial responsibility.35The DOJ Report also described its methodology for establishing “cut points”—the score thresholds for each of the four categories of risk.36 The cut points were determined relative to the “base rate,” which is roughly the “average” rate of recidivism for the entire data set.37 However, the DOJ did not set the middle cut point—the threshold between low and medium risk—at this base or “average” rate of recidivism. Rather, it elected to set the middle cut point at one-half of the base rate.38 The minimum risk category was set at one-quarter of the base rate, while the cut point between medium and high risk categories was set at roughly two-thirds above the base rate.39 The report explains that this set of cut points was one of nearly a dozen that were examined or tested, but the other sets of cut points were not identified. The only explanation given for the cut points selected was the need to balance between maximizing the number of incarcerated individuals eligible to earn early release “while also considering public safety and the risk of recidivism upon release.”40PATTERN calculates two risk scores, a score for general recidivism risk and a score for violent recidivism risk.41 PATTERN also models men and women separately because of their gender-specific risk factors and pathways to crime.42 Other research, as well as the PATTERN’s own dataset, indicate that incarcerated women are outnumbered by men almost 6 to 1, so that if one model were used women’s scores would be centered on men’s characteristics.43 The developers concluded that gender-tailored risk and needs assessments will better address needs.44Following release of its initial report and PATTERN scoresheets, the DOJ solicited additional public comment regarding PATTERN.45 The DOJ received almost 200 comments and statements in response.46 As a part of this comment process, the President of the ABA sent a letter to the DOJ on August 26, 2019, setting forth a series of questions and concerns regarding, among other matters, the selection of the cut points, the definition of violence, the lack of transparency in the data used to develop the tool, the potential for racially disparate impacts, and other questions regarding the implementation of the overall risk and needs assessment system.47 The DOJ did not respond to our letter.The DOJ also held additional listening sessions regarding PATTERN in September of 2019.48 Members of our Task Force participated in a listening session on September 11, 2019. We referenced the matters addressed in the letter from the ABA President and expressed our continuing concerns about the lack of transparency regarding the data underlying the tool. We also reiterated our concerns about the potential for racially disparate impacts due to differing police practices in communities of color, particularly in light of the significance placed by the tool on factors such as criminal history and age at first arrest.In response to the public comment and continued study, the DOJ released a revised version of PATTERN in January of 2020.49 The DOJ’s report accompanying the updated version of PATTERN described some of the changes made in the new version.50 For example, two additional dynamic factors were added so that incarcerated individuals could have a greater opportunity to lower their risk assessment scores through programming and occasional reassessments.51 First, the new version of the tool considers “the offender’s ‘infraction free’ period during his or her current term of incarceration,” which indicates the period the individual was free of infractions.52 The original infraction conviction factor reflecting infractions within the prior ten years was changed to reflect the duration of time in which the individual was infraction free.53 Second, the eligible programs were modified to include certain psychology treatment programs.54 The January 2020 report also explained that some factors had been removed in the new version of PATTERN in an effort to mitigate the potential for racially disparate impacts.55 The factor for age of first arrest/conviction was removed on the ground it would to while the risk factor of surrender was removed on the it would to the of this version of the the IRC two that into the the of the First, it that of federal are eligible for the FSA’s time credit Second, it no between the of BOP’s and inmate which that the BOP’s criteria are the new version of PATTERN included changes that were explained by the DOJ in its January 2020 In particular, the to general criminal history increased by roughly while the violent criminal history was The of these changes be For for some individuals the points added for violent criminal history in the new version of PATTERN to any benefits from of EBRR programs and productive activities they Other the age of the at the time of no factor is given more in PATTERN criminal The changes to this factor in the of PATTERN is the DOJ has elected to these changes is of The individuals who will be from were designed to incarcerated to in recidivism reducing the individuals who are to the greatest risk to public It no to that would to in activities that are designed to reduce their risk of Task Force to have unanswered questions and concerns regarding number of these are of the PATTERN’s and potential is only if the factors it the its risk and needs and its underlying statistical are made The DOJ’s PATTERN in January of 2020 a to data but to the relevant information The BOP has with an outside to the data practices and it has also to an report on and recidivism of us on the outside in have several (1) a number of incarcerated individuals from early (2) there be racially disparate impacts on those who are to from early (3) the data and underlying the development of the tool be made to the public and other particular concern is that the potential for and be addressed data transparency and of individuals’ data and data a risk to a Risk Assessment tool, the FSA also directed the DOJ to develop a Needs Assessment system programs at reducing risk through an individual’s treatment It not that the DOJ has yet with this aspect of the the DOJ has to an of practices by the BOP to inmate these practices are as by the Independent Review “the of system to have had in with FSA We the DOJ to its work and create a and needs assessment system to assessments of inmate risk with programs and activities to their individual to our concerns regarding the implementation of the FSA, we also additional should also be to its benefits more The FSA included a of offenses that incarcerated individuals from time to the more of all individuals incarcerated in federal are by the as the IRC that there to be no in the risk of those who are eligible for time credits and those who are a lack of underlying the of disqualifying As the IRC has and as the the BOP to release those of disqualifying offenses under in the FSA. Our Task Force with the that this a the of recidivism among those released not between those convicted of disqualifying offenses and may need to the of disqualifying offenses with an the PATTERN risk assessment tool into the incarcerated individual’s participation in programming in credit for early However, there is information about the the programs that are offered or the to which participation in can reduce The DOJ’s its needs assessment has been modified to include treatment and and work While additional programming will improve the of of incarcerated more information is on these and other productive activities to which programs and activities are made to prisoners is also for The FSA required the BOP to complete a risk and needs assessment of every individual incarcerated in a federal prison by January and to provide EBRR programs and productive activities to the needs of every incarcerated individual by January on the BOP’s of it that only a of the programming that will be is yet in particular concern is that the BOP of the programs that are most to improve a chance for earlier release or other of the programs on the only of are or and only of these programs are at all BOP The BOP to have a long to in the to its programs and activities to be in with the of concern is whether the between participation and sentence reduction for The BOP has a that would a of participation in an EBRR or productive as an period of a or for its or The if applied in with the information in BOP’s of EBRR programs and productive activities, would to the credits for of its The in with the BOP also to no FSA credit at all for inmate outside such could be a productive within the of the FSA, and the BOP has offered no explanation for it has not Further, the would provide no time credits for incarcerated individuals participation in a is successful but who are to complete the through no of their such as when an inmate is to a that not have the the also time credits for incarcerated individuals in programs and activities in a or on to the requirement in the FSA that all individuals in BOP custody be activities, and of our Task Force the is at with the of the FSA, and to ensure FSA credits for inmate better would provide that a a and that every an inmate successfully in a or required inmate work is a that should for FSA time PATTERN risk assessment tool to predict recidivism so that incarcerated individuals are released when they have in the programming and are to key is recidivism is The assessment as a to BOP custody or a within three years of release from BOP as well as under the and while recidivism is a for a of within three years of release from BOP definition but is not limited sex and can be by police and that not result in a or conviction are not a measure of recidivism. Further, the definition of general recidivism that offenders who are only at risk for be early the to it has that a and risk is individual a and risk score that is then on a comprised of all recidivism risk scores of individuals released by BOP the recidivism base individual’s risk category is determined by the within this cut points are used to categories of As risk is made up of scores that roughly the base rate, risk is set at under one-quarter of the base Our Task Force is to the to set the middle cut point between low and medium risk at one-half of the base rate is or in with the and of the FSA. given the increased scores for criminal some incarcerated individuals will be to from the FSA of or years they may themselves through programming and productive PATTERN risk assessment tool both and factors to a risk and the developers the of dynamic factors as an in and among the of dynamic factors are history of violence, history of and infraction factors are not which the DOJ as that with programming and or could be by the inmate’s while these of factors may be relevant to an assessment of they are not relevant to whether a has successfully reduced his or her PATTERN risk assessment is an tool. The and the tool has may the long it is that the tool with opportunity for and on the data used to develop the tool. The has for years the in transparency that can the risk and needs assessment system established by the FSA could the to who work and in have and because of the prison is not to the in which incarcerated individuals and and can among the prison to an by The as of January there have been a of of in the federal prison The first incarcerated individuals in federal custody from on 2020 and since then there has been a of of incarcerated to respond to the of the from several courts have the to reduce an incarcerated individual’s sentence and a term of on release if the that “extraordinary and such a or that the individual is at years of the individual has at 30 years of his or her and the BOP has determined that the individual is not a to the safety of any other or the The BOP is also to certain and individuals on The also to the President of the to grant for federal which can include an individual’s sentence to time there is the and Act an response to the in the The Act in intended to those who are in the custody of the BOP through to BOP and and particular the of for under the Act is on the PATTERN risk assessment to the passage of the Act, incarcerated individuals were allowed to be released on for of their term of or 6 is 26, before the Act was signed into a to BOP which directed to the of the BOP’s to grant for individuals in with the that are some who are and a minimal of recidivism and who be their sentences in in BOP which individuals should be the BOP was to the of the for each incarcerated which included but was not limited to their PATTERN risk assessment The individuals who receive a medium or high risk score from treatment under factors that to be included: The age and of the inmate to in with the for and level of the the with given to in low and medium inmate’s in with who have in violent or in prison or who have BOP within the last not treatment under this the inmate has a and that will recidivism and public that the under which the inmate would be upon release would a lower risk of the inmate would in his or her BOP inmate’s of and assessment of the by the inmate to the such as sex will an inmate for Other serious offenses should more for April to BOP which that the has limited to monitor on and that the is to monitor of in the and the BOP to individuals to if was not long as BOP in every such that so is and with to public due in part to on the PATTERN tool, the has only a of the federal inmate In December BOP to the Committee that individuals had been to with an additional individuals to be about a this number may the BOP has individuals in its only about of the inmate has from placed on should also be that the DOJ a on January that individuals who have been placed on during the be to prison after the those who are eligible for under Act at 18 U.S.C. will be allowed to on any of this will under the is of the PATTERN to who should be released during the has been on a number of that development and of the PATTERN did not into the and of the and assess the public safety risk that an individual to PATTERN the is also because it is to in a letter these and other the on and and other the a letter to that his 26, 2020 The letter that “the of a tool PATTERN to or decisions is and to of of of and at the of the The in the letter were First, those as the PATTERN only given for under the BOP’s for the of the federal prison when to other that are as low, medium or high Second, only of men in the are as under the PATTERN tool, to of the PATTERN tool not into the that of have an increased risk of and from because of the in the and most of recidivism and data are relevant to any assessment of public safety risk during this the letter a key factor in a PATTERN score is the individual’s participation in but incarcerated individuals have not had time since the implementation of PATTERN to in programming that will have any effect on their initial PATTERN Additionally, with the limited in federal from efforts to the of programming at may not be as as it was and may have been FSA was intended to prisoners to in its the law to the early release of eligible earn 10 days of time credits for every 30 days of successful participation in recidivism reduction programming or productive and minimum or low risk prisoners earn an additional days of time credits for every 30 of such Our Task Force has the implementation of these and has unanswered questions and continuing this Report with the that our questions and concerns will be and that the of the First Step Act will be through more in this Report are on behalf of the Criminal Justice Section First Step Act Implementation Task The Report has not been by the of or the of of the and may not be as the of the