Pretrial Risk Assessments
by Jeffrey Clayton
(excerpt from Mediate.com)
Nearly everyone both within and outside of the legal profession has heard of bail reform. Couched typically as “America’s cash bail system,” it has been branded as evil and in desperate need of repair despite being one of the ten core constitutional rights protected by the Bill of Rights. Reformers loudly sold their premise that the system is biased against certain defendants based on either their race or poverty level. This begged the question: what can we do about it?
The answer arrived about a decade ago in the form of pretrial risk assessment algorithms. A one-size-fits-all solution, it neatly evaluated criminal defendants on their probability of failing to appear in court or committing a new crime, then presented the risk score to a judge. Not all scores were created equal, however. Some carried heavier weights than others, which drove significantly higher negative outcomes. For example, a person charged with a felony who was determined to be “low risk” might have been released on a promise to reappear in court. A different person charged with exactly the same crime could be labeled “high risk," and thus have bail set at an unattainable level or simply be locked-up outright without bail, as is presently the case in New Jersey.
Civil rights groups quickly lined-up to get on the bandwagon to advocate for the wholesale adoption of this system based on the use of risk assessment tools. They claimed it could replace money bail, while substantially reducing any inherent bias. The goal of a color-blind criminal meritocracy was the idea, and many national groups of elected and public officials began to buy in. The most significant public stance was taken by the Conference of Chief Justices. Comprised of the top justices and judges in the United States and its territories, this august body, as we shall see, wound up on the dead-wrong side of things.
In 2015, the Conference of Chief Justices endorsed the universal use of risk assessment tools across the United States, but within a year, cracks in their implementation began to show. At first, there was merely concern as to potential racial and gender bias, as well as discrimination against those with disabilities. But there was also evidence of problems with due process, in which actuarial group data was applied to an individual. Concerns also emerged over transparency, including the determination of who exactly determined the risk percentages of a new crimes that counted as “high, medium or low.”