City government’s apparent success identifying and releasing individuals at low-risk of reoffense begs question of why policy is not universal

January 19, 2022


MEDIA CONTACT: Chad Topaz, Ph.D., QSIDE Co-Founder and Executive Director of Research

MASSACHUSETTS—The Institute for the Quantitative Study of Inclusion, Diversity, and Equity (QSIDE) today announced publication of its research paper “New York City Jails: COVID Discharge Policy, Data Transparency, and Reform,” in the journal PLOS One. This study analyzes Mayor Bill DeBlasio’s spring 2020 policy releasing inmates who were deemed to have a high risk of COVID infection and/or low-risk of criminal reoffense. The research seeks to determine the degree of reincarceration of released individuals, as well as to identify any potential bias based on race/ethnicity, gender, or other intersectional identity characteristics of those released under the policy.

“We found low rates of reincarceration among those who were granted their freedom from Rikers Island under de Blasio’s March 2020 policy,” said Chad Topaz, Ph.D., QSIDE Co-Founder and Executive Director of Research. “The success of the release program suggests that the city government has the capability to identify incarcerated individuals who are unlikely to reoffend.”

The fact that the City has these capabilities raises questions as to why so many individuals in the system remain incarcerated, according to Topaz. “If in March 2020 the city was able to tell who was unlikely to reoffend, and if the city was willing to release them, then why do so many low-risk individuals remain in jail? Why is the March 2020 policy not an ongoing policy?”

Releasing low-risk individuals from jails both makes sense from an ethical perspective and likely has an economic upside. “Over the past six years, if incarcerations had been capped at one year in duration, we estimate that the city jail system would have experienced a 15% decrease in total incarceration. With a cap of 100 days, the reduction would have been over 50%. Implementing such caps could save the city millions of dollars, and would be a major move towards the city’s stated goal of closing Rikers Island,” said Eli Miller, lead author and recent graduate of the Williams College Department of Mathematics and Statistics.

The study also seeks to determine if there was demographic bias in who was released under de Blasio’s policy, for instance, if white individuals were released at significantly different rates than their BIPOC counterparts. The study’s authors were shocked by the state of the data collected by the New York City jail system and shared on the city’s open data portal. “In the public data, there are only three possible categories used to describe an incarcerated individual’s race: ‘Asian,’ ‘Black,’ and ‘Unknown’,” Miller said. “With this data, there is no way to assess bias.”

“This is low-hanging fruit,” Miller concluded. “The city should use a more complete and more humane set of categories to provide visibility into the role of race. The city’s current public data renders this type of analysis impossible, and hinders public accountability.”

Researchers are available to speak with the media; please contact QSIDE for scheduling.


About The Institute for the Quantitative Study of Inclusion, Diversity, and Equity (QSIDE)

QSIDE is the pre-eminent research-into-action institute applying mathematics, data science, and computation to social justice. We use cutting-edge tools to find, secure, analyze, and model data that is difficult to aggregate or is intentionally obscured.