Unique mix: Researchers and stakeholders consider big data use in criminal justice

An event this fall at Moritz College of Law brought together an unprecedented mix of participants to examine the use of big data in the criminal justice system. Academics who study big data, criminology, and policing were joined by people who work in criminal justice, including prosecuting and defense attorneys; judges; city council members; Ohio state legislators; representatives from the ACLU of Ohio, Ohio Criminal Sentencing Commission, and Supreme Court of Ohio; and more.

The translational approach expanded the examination of a socially significant topic beyond the realm of research to include end-users.

“It’s the first one I know of that involved stakeholders,” said co-organizer Ric Simmons, professor of law at Moritz College. Even the variety of academic areas represented was unusual; people from fields as diverse as criminal law, computer science, statistics, and anthropology were there. “It’s now becoming much more recognized as a significant topic,” he said.

Dennis Hirsch

Dennis Hirsch

“It raises fascinating issues like the proper balance between the human decision-maker and the machine decision-maker,” said co-organizer Dennis Hirsch, a TDAI affiliate, Moritz professor of law, and director of the college’s Program on Data and Governance.

Simmons and Hirsch developed the idea for the invitation-only event, which was held September 29 and 30 at the Barrister Club. “The criminal justice system is increasingly using big data to determine bail, parole, sentences, and the allocation of police resources,” said Hirsch. “There are a variety of ways it is being used or could be used in the criminal justice system.”

As with many innovations, big data presents as many opportunities as challenges. “You could reduce bias in decision-making,” points out Hirsch; if data is used to determine a reasonable sentence or a parolee’s flight risk, for example, it could remove human bias from the judge’s decision-making process. However, human discretion is deliberately built into our justice system for cases with extenuating circumstances.

“In our judicial system we have judges and others who can exercise discretion,” said Hirsch. “They may feel the rules don’t fully capture the human situation here.” In addition, humans aren’t the only source of bias. One of the factors used to consider the likelihood of recidivism is prior arrests, with more arrests assumed to indicate a person who is more likely to reoffend. “That data might be biased itself,” Hirsch said. “We have seen that poor and minority communities are more heavily policed” and thus people are more likely to be arrested for crimes than they would be in another area.

“We are turning over decisions to machines and algorithms,” Simmons said, “but we need to consider: Is there some sort of minimum at which humans must be involved?”

The September event took place in two parts. On day one, the academic participants met to present their research and participate in a roundtable. In advance, attendees reviewed idea papers by 12 participants that formed the basis of the afternoon  discussion.

On September 30, 26 people from the community representing various aspects of the criminal justice system joined the mix: judges, law enforcement officials, public defenders, people who work in corrections, employees of legal watchdog organizations, and politicians.

“Involving the stakeholders helped expand the reach of the discussion,” Simmons said. Breakout discussions moderated by an academic participant each centered on a topic: predictive policing, big data in surveillance, the use of big data for criminal risk assessment, the balance of human versus machine decision-making, and potential innovative uses of big data.

“We compiled a list of research questions from the sessions,” said Hirsch. “We also asked the stakeholders what they would like to know more about.”

One recurring observation by stakeholders was the disparities in how data are organized.

“There were recurring pleas for standards in using and collecting this data,” Simmons said. “Every group catalogs or organizes it differently. There’s no consistency in what this data means, there’s no consistency across jurisdictions.”

Simmons noted that the public tends to hear stories about algorithms failinga person who was paroled and then reoffendedand distrust the use of big data. However, there are many times big data has been used successfully and accurately. In those cases, police resources are saved and individual liberty is maximized, he said. It’s partly a matter of gathering more data and organizing it correctly. “I want people to look at the big picture to see what these numbers will be,” he said.

Conversations initiated at the event will continue with the publishing of the event’s idea papers and questions to guide future research in the January 2018 The Ohio State Journal of Criminal Law, spurring further work in this area. “A lot of people at the end of the second day were talking about continuing the conversation,” Simmons said.

Share this page
Suggested Articles
Visualization challenge teaches the art of telling data’s stories

The winning team’s presentation At first blush, Translational Data Analytics and creativity might seem worlds apart. When it comes to data collection, after all, “creative” is often a euphemism for...

Call for papers: Political Networks Conference and Workshops

TDA is co-sponsoring the 10th Annual Political Networks Conference and Workshops, which Ohio State will host June 14-17. The program chairs and host committee are pleased to invite proposal submissions for...

User-generated data is a social science goldmine

As increasing amounts of digital data are produced and stored online it is important to remember that humans produce much of that data. In an era in which people express...

TDA to host "Big Data for Good" Fall Forum

Does big data really matter? It does if you put it to work in solving today’s complex challenges. TDA’s 2016 Fall Forum will focus on “Big Data for Good,” a special...

New data visualization book: Know thy audience

When it comes to data visualization, meaning is in the eye of the beholder. For this reason, knowing your audience is critical for creating data visualizations that fulfill their purpose....