In courtrooms across the United States, many judges decide whether to incarcerate youth on the basis of a “risk and needs assessment” tool that has proven to be biased.
These assessment tools allow justice officials to place youth along a spectrum from low- to high-risk offender. The ranking is also used to identify the services and supports aimed at preventing these youths from reoffending.
In a recent Child Trends report, we found that risk and needs assessments may misclassify youth of color as being high risk, which may perpetuate racial biases in the justice system.
That is a cause for concern. Black, Native American, and Latinx youth in the U.S. are already more likely to be incarcerated than their white counterparts. As the use of risk and needs assessments increases, now is the time to develop strategies to address this systematic bias.
Our review expanded on prior research showing that the assessments do not always classify youth of color at correct overall or subcategory risk levels, and that racial differences in risk and needs assessment classifications may be used to disparately impact youth of color.
Misclassifying youth-risk level has long-term implications. Low-risk individuals exposed to heightened levels of justice interventions are more likely to break the law in the future and reenter the justice system.
Risk and needs assessments can also have a disparate impact on youth of color (compared to their white peers) because of the way these assessments are constructed. First, they commonly define “risk” as an expression of the chance of reoffending, often measured by rearrests or reconvictions.
Research shows that this definition of risk is systematically biased against youth of color due to historical and disproportionate racial differences in stops and arrests, sentencing, and community supervision outcomes.
Second, the assessments use domains to determine a youth’s level of risk. Some domains and their associated measurements are influenced by racially biased systematic differences, such as criminal / juvenile history.
Similarly, the way such categories are weighted and scored to determine a youth’s overall risk level may lead justice officials to disproportionately classify youth of color at higher risk levels.
We offer four recommendations to policymakers and practitioners to reduce the likelihood that risk and needs assessments will perpetuate systemic racial bias:
- Policymakers or practitioners should enact protocols or policies that require assessments to be proven to work on the targeted jurisdiction’s populationregardless of whether the assessment has been validated elsewhere;
- These systems should also require a review of assessments to make sure they reliably predict the risks they purport to measure within the targeted jurisdiction;
- When evaluating risk assessments for racial bias and disparate impact on youth of color, evaluators should carefully review every domain and item, cut-score, and measure for reoffending;
- Finally, policymakers and practitioners can engage in quality improvement by continuously refining the assessment itself and ensuring that the decision makers using the tool (e.g., probation officers and judges) “buy in,” or believe in its value, and possess the training and resources to implement it as intended.
To be clear, research shows that risk and needs assessments are more accurate at identifying a youth’s level of risk than the professional judgement of justice officials, such as a judge or probation officer.
While this research supports claims that these assessments create a more equitable system, it does not eliminate concerns that risk and needs assessments have a disparate impact on youth of color. In addition, some researchers contend that there is statistical difficulty in creating a singular risk and needs assessment that produces fair outcomes for people of all races, genders, etc., and that multiple aspects of fairness often cannot all be simultaneously satisfied.
And some researchers stand by the opposite argument that there is no evidence to indicate that risk and needs assessments are biased.
The evidence indicating that risk and needs assessments may misclassify the risk level of youth and have a disparate negative impact on youth of color calls for decision makers to develop and implement strategies that can ensure these assessments do not cause further harm to youth of color who are involved with the justice system.
Victor St. John is a research scientist at Child Trends. Reachable by Twitter at @vjstjohn, Kelly Murphy is a deputy program director at Child Trends. Twitter handles: @childtrends @evalkell. Akiva Liberman is a senior research scholar at Child Trends.