A pattern of bias in PATTERN
The First Step Act, implemented on Dec. 21, 2018, required the Department of Justice (DOJ) to “develop a risk needs assessment system to be used by the Federal Bureau of Prisons (BOP) to assess the recidivism risk and criminogenic needs of all federal prisoners and to place prisoners in recidivism reducing programs and productive activities to address their needs and reduce the risk.” Critics instantly predicted this plan would create issues of racial and gender bias.
In July 2019, the DOJ announced its new risk and needs assessment system, titled the Prisoner Assessment Tool Targeting Estimated Risk and Need (PATTERN). The assessment categorizes inmates into four risk levels: minimum, low, medium and high. These categories can be used to determine an inmate’s eligibility for rehabilitative programming and the incentives they may be offered, including early release, home confinement or alternative housing. Rehabilitative programs are intended to help convicted persons re-establish their lives after release from prison. They incentivize inmates to participate in programs aimed at reducing recidivism with the knowledge that they can lower their scores and be eligible for better programs.
PATTERN was derived from a long history of Risk-Needs-Responsivity (RNR) models. In the past, the RNR models focused on treatment to address the criminogenic needs of prisoners that contribute to criminal behavior, which was delivered in a manner consistent with the convicted persons’ ability and learning style. Previous risk assessments only looked at static factors in determining risk—such as age at first arrest, classification of offense as violent or nonviolent, sex offender status, criminal history score and history of violence—which ultimately created racially disparate impacts on people of color in the system. More recent models, including Correctional Offender Management Profiling for Alternative Sanctions (COMPAS), the Level of Service Inventory-Revised (LSI-R) and Static Risk and Offender Needs Guide for Recidivism (STRONG-R), which all influenced PATTERN, now consider dynamic factors. Considering factors such as antisocial attitudes, peers, personality, substance abuse, marital/family issues, employment, education and use of leisure time help address the inequalities these tools create. The sample size of the tests and the underlying issues behind the factors continue to put persons of color at a disadvantage in the criminal legal system.
Risk assessments are often implemented because they appear to be an unbiased and successful means of helping prisoners get appropriate help. However, PATTERN has consistently overpredicted the risk that Black, Hispanic and Asian individuals will recidivate after leaving prison. The DOJ found that in the initial use of PATTERN, only 7% of Black people in the sample were classified as a minimum-level risk as compared to 21% of white people. Some of the leading causes for this disproportion include the types of crimes in the inmate’s criminal history, education level and whether the inmate paid restitution to the victims. Newer data continues to show a strange mix of racial disparities. In 2021, results showed that, as compared to white inmates, Black females were overpredicted by 7% for general recidivism. Further, there was a 2 -6% overprediction for Hispanic individuals and a 5-8% overprediction for Asian individuals. Surprisingly, there was a 12 -15% underprediction of recidivism for Indigenous Americans compared to white inmates, since tribal reservations were reportedly not required to provide arrest information; this prevents individuals from receiving the treatment that best fits their needs. These results show the persistent racial disparities that put persons of color at a disadvantage.
There has been a lack of awareness of PATTERN and its impact on the criminal legal system since it was implemented in 2020, even though it was immediately tested and used for purposes beyond its original intention at the beginning of the COVID-19 pandemic. When PATTERN was just two months old, the COVID-19 pandemic began, and the tool was not only used for programming but also to determine who was eligible for compassionate release. Only inmates who pose a low or minimal risk of recidivating can qualify for these programs, which is problematic, considering 14,170 prisoners have been assigned to erroneous risk categories as of late 2020. The tool was not intended to be used for compassionate release purposes and does not consider important factors like an inmate’s vulnerability to the virus. Since inmates were released during the pandemic based on their PATTERN scores, persons of color were often exposed to the prison system for longer than white individuals.
PATTERN has been used to release thousands of federal inmates in the beginning of 2022 alone, and it will likely continue to produce results that disproportionately impact persons of color. In a Congressional Hearing on Jan. 21, 2022, Representative Nadler, Chairman of the House Judiciary Committee, commented that the Biden Administration has already begun revising PATTERN calculations for Earned Timed Credits, hopefully indicating a willingness to address and fix the problems associated with PATTERN.
Many legal practitioners have called for PATTERN to be halted until the racial concerns can be fixed. Since suspension is unlikely, there are several steps the DOJ can take to continue to address PATTERN’s problems. First, it can release more information about the tool to the public. There is very little known about the tool besides reports from the National Institute for Justice, and there is no anonymized version of the datasets available for independent researchers to assess. More transparency about the tool might allow the DOJ to receive more feedback and advice. Second, the DOJ should take further steps to avoid inaccurate ratings. Wrongly classifying 14,170 people as “high risk” inmates signifies this tool is flawed and must be resolved so that inmates can receive the treatment that best suits their needs. Third, and most importantly, the DOJ must continue brainstorming ways to reduce racial disparities, or else this PATTERN will exacerbate the current epidemic of mass incarceration.