In the Press
Thursday, December 1, 2022EU Proposes Special Court for Russian Crimes BBC World Service Newshour
Tuesday, November 29, 2022Supreme Court Should Separate Sleazy Lobbying from the Criminal Kind — A Commentary by Stephen L. Carter ’79 The Washington Post
Monday, November 28, 2022Racial Discrimination by Veterans Affairs Spans Decades, Lawsuit Says The Washington Post
Monday, November 28, 2022A Black Vietnam Veteran is Suing the VA for Discrimination NPR
Thursday, November 17, 2022
MFIA Clinic Fights for Release of Parole Algorithm Information
This week, the Media Freedom and Information Access Clinic (MFIA) at Yale Law School filed a lawsuit in the New York State Supreme Court challenging a state agency’s refusal to release basic information about an algorithm integral to the criminal justice system.
Fordham Law School Professor Martha Rayner filed a Freedom of Information (FOI) request with the New York State Department of Community Corrections and Supervision earlier this year, seeking information about an algorithmic tool used in every parole decision in New York. Following the denial of her request, MFIA took up Rayner’s case to ensure that information about the controversial algorithm is released to the public. The suit was filed Nov. 15.
“This algorithm is a crucial component of the parole system in New York,” said Clinical Lecturer in Law and MFIA Local Journalism Fellow Stephen Stich ’17. “Shedding light on it is essential for the public to oversee the Parole Board’s decisions and for prospective parolees to understand and challenge them.”
Every year, the Correctional Offender Management Profiling for Alternative Sanctions (COMPAS), a privately developed reentry risk assessment software, is used in tens of thousands of decisions on whether to release people incarcerated in New York who are serving indeterminate sentences. After such people serve their minimum sentence, their release depends entirely on the New York State Parole Board, which is legally required to consider the algorithm’s recommendations in every one of its parole decisions.
COMPAS, like similar risk assessment tools, is used to predict an offender’s likelihood of recidivism on release. It does so by comparing offenders’ data, collected through a questionnaire, with the data of a class of individuals called the norm core data group. Based on this calculation, it generates a risk score for each inmate.
Such tools have faced intense public scrutiny and debate and received criticism for being inaccurate, racially discriminatory, and rigged against the poor, according to the clinic’s filing. Further, the way in which COMPAS makes its assessments is poorly understood, depriving offenders and the public of the ability to understand the basis of a parole decision.
Rayner filed the FOI request in January fore more details about how the COMPAS instrument calculates risk scores. She sought information including the formulas it uses; the cutoff scores for categorizing individuals as low, medium, and high risk; and the norm core data. She also sought studies validating COMPAS, which is used by public agencies across the country. The department repeatedly delayed its responses to the request before denying it in full in June, according to the lawsuit.
The department claimed that Rayner’s request was overbroad and that releasing the information would violate the copyright and business interests of Northpointe, Inc, COMPAS’s for-profit developer. The company has sold various customized COMPAS products in New York since 1998 without undergoing a competitive procurement process.
“The Department’s denial exemplifies a troubling trend of agencies outsourcing their core functions to private corporations and thereby preventing the public from understanding or overseeing their decisions,” Rayner's petition stated. “The public is entitled to a meaningful explanation of the basis for the state’s decisions to grant or withhold parole. The Court should order the Department to produce the records that will help inform the public on this important issue.”
In response to the denial, the clinic showed that the department denied Rayner’s request as overbroad only by misconstruing what the request seeks. The clinic also argued that there are no applicable copyright protections and that any harm to Northpointe’s business interests would be entirely speculative.
“Criminal justice officials in New York and across the country are increasingly using algorithmic tools to make life-changing decisions about who stays incarcerated and who does not,” said MFIA student Arjun Malik ’24. “It is hard to imagine an issue where public oversight is more sorely needed than this one. We look forward to pressing Professor Rayner’s claims in court and ensuring that New Yorkers learn how their government is making parole decisions.”
Yale Law School students Dorrin Akbari ’24, Arjun Malik ’24 and Samantha Stroman ’23 also worked on the petition.
The Media Freedom and Information Access Clinic (MFIA) is a law student clinic dedicated to increasing government transparency, defending the essential work of news gatherers, and protecting freedom of expression by providing pro bono legal services, pursuing impact litigation and developing policy initiatives.