Researchers warn that risk-based bail systems will lead to an increase in pretrial incarceration in new study.
June 19 2018
As legislators and policy-makers across the country continue to deal with the promises of pretrial “pilot” counties funded by the deep pockets of social “reformers” and proposed legislation intended to eliminate the money-based system and replace it with a risk-based system using computer algorithms, research continues to mount against the 3rd generation of bail reform.
In the most recent article to date, John Logan Koepke, Policy Analyst at Upturn, and David G. Robinson, Managing Director and co-founder, Upturn and Adjunct Professor of Law, Georgetown University Law Center, authored a forthcoming paper in the Washington Law Review entitled Danger Ahead: Risk Assessment and the Future of Bail Reform.
The paper is an examination of the pitfalls of bail reform and makes some important points that policy-makers should consider. The conclusion of the article, which takes on aspects of bail reform directly, is probably worth reading first:
Pretrial risk assessment instruments, as they are currently used, cannot safely be assumed to advance reformist goals of reducing incarceration and enhancing the bail system’s fairness. Early evidence remains sparse, and risk assessment instruments may yet prove themselves effective tools in the arsenal of bail reform. But they have not done so to date. Shifting to risk-based bail will not necessarily reduce incarceration. Without stronger data practices and open governance, we believe it is likelier than not that these tools will perpetuate or worsen the very problems reform advocates hope to solve.
The article makes several other key points of critical importance that we have been making in the conversation about bail reform. Critically, since risk-assessments lock us in the past, we will ultimately create a backward-looking system that will stop true reforms.
On the issue of over-supervision as an alternative to bail, the research looked at the New Jersey system and backed up our conclusion. We have openly been concerned and feared that judges would trammel the liberties offenders by blanketing them with non-monetary conditions of release including supervision by pretrial agencies and a litany of correctional technologies. Our fears were corroborated in New Jersey: “Here, nearly one-third more defendants were subject to the most restrictive conditions of release — or were denied release — than was expected. And while nearly one quarter of defendants were expected to be released on recognizance, only 7.5 percent of defendants were.”
Another key point worth discussing is that these researchers call the decision-making frameworks, which have been implemented in various jurisdictions through pilot programs, “frameworks of moral judgment,” and then call for greater transparency and due process. These frameworks are implemented without due process and with no required transparency in terms of how local officials will make these moral judgments. Said the authors…
“Determining how much risk a society should tolerate — and then formalizing those answers inside decision-making frameworks — is a difficult political and moral question, not a primarily technical one.”
While the authors stop short of calling for the elimination of algorithms in pretrial risk assessments, they did note that there are many proprietary algorithms being used in criminal justice, and there was a need to address that issue. They noted in order for reforms to succeed “the public needs a chance to find — and to press authorities to address — the kinds of risks we have described elsewhere in this paper. Claims of trade secrecy or confidentiality provisions in contracts immunize pretrial risk assessment tools from meaningful public inspection — including from judges.”
On preventative detention, the authors argue that using algorithms reinforces the need for preventative detention and continues the justification for the Salerno decision that first opened the door to mass pretrial incarceration. This is so because risk assessments “tacitly presume that Salerno’s core holding was correct.” The researchers noted several jurisdictions where risk assessments serve to increase pretrial incarceration, and then immediately noted that the current reality of risk based bail reform in practice has turned Salerno on its head:
In upholding the 1984 Bail Reform Act, the Chief Justice Rehnquist wrote that, “[i]n our society, liberty is the norm, and detention prior to trial or without trial is the carefully limited exception.” Of course, this stated truth rings hollow when compared to today’s stark reality.
The scholars’ conclusion about the long-term issues should certainly cause any true-believers in the third generation of bail reform serious pause:
[T]he embrace of pretrial risk assessment instruments creates longer-term doctrinal and policy risks for advocates of bail reform. Specifically, these new tools risk giving an imprimatur of scientific objectivity to ill-defined concepts of “dangerousness”; pave the way for a possible increase in preventive detention; and may entrench the Supreme Court’s historically recent blessing of preventive detention for dangerousness.” We find that pretrial risk assessment instruments, as they are currently used, cannot safely be assumed to advance reformist goals of reducing incarceration and enhancing the bail system’s fairness. Early evidence remains sparse, and risk assessment instruments may yet prove themselves effective tools in the arsenal of bail reform. But they have not done so to date. Shifting to risk-based bail will not necessarily reduce incarceration. Without careful design and open governance, we believe it is likelier than not that these tools will perpetuate or worsen the very problems reform advocates hope to solve.
It is time for the legislature to stop the movement of legislation that will lock states in the past and prevent meaningful reforms. It’s time to end the conversation of allowing “black box” algorithms to replace judicial discretion.
Read the research…Danger Ahead: Risk Assessment and the Future of Bail Reform.Danger Ahead