Criminal Justice in the Age of Big Data

By Nick Sinai, Adjunct Lecturer in Public Policy, Harvard Kennedy School

Left to right: Nick Sinai, Cynthia Rudin, Clarence Wardell, Nicole Wong.

“Our criminal justice system isn’t as smart as it should be.” — President Obama, July 15, 2015

This summer, President Obama addressed the 106th NAACP convention to lay out the reasons we need to reform the criminal justice system.

The numbers speak for themselves. The U.S. has 2.2 million prisoners, up 400% since 1977. The U.S. has 25% of the world’s prisoners, but the U.S. is only 5% of the global population. And 60% of U.S. prisoners are black or Latino.

As a society, we’re using criminal justice in place of social services for some of our most sick and vulnerable citizens — 68% have substance abuse issues, and 50% have a serious mental illness.

But it’s not just the President who is pushing for reform. Republican presidential candidates, like Rand Paul, Ted Cruz, and Chris Christie, are also advocating for a fairer system. The Koch brothers have collaborated with the Center for American Progress and the Arnold Foundation on efforts to reduce mass incarceration.

We are reaching a pivot point in our criminal justice system.

To explore how criminal justice is changing, especially in an age of technology and big data, Harvard Kennedy School hosted a panel discussion on November 13 entitled Criminal Justice in the Age of Big Data. Co-sponsored by the Shorenstein Center on Media, Politics and Public Policy, the Center for Public Leadership, and the Ash Center for Democratic Governance and Innovation, the panel featured:

  • Nicole Wong. Former U.S. Deputy Chief Technology Officer; former Legal Director of Products, Twitter; and former VP & Deputy General Counsel, Google.
  • Clarence Wardell, Former Presidential Innovation Fellow, working on the Police Data Initiative; Berkman Center affiliate, and currently in the White House U.S. Digital Service.
  • Cynthia Rudin, Associate Professor of Statistics, MIT CSAIL and Sloan School of Management; Berkman Center faculty affiliate.

Nicole Wong opened the panel with historical context: one of the pioneering data scientists in criminal justice was African-American journalist Ida B. Wells, who used data from newspaper articles in the late nineteenth century to disprove that lynching was primarily in response to rapes and murders, and to argue that “our country’s national crime is lynching.”

As one of the principal authors of the White House’s 2014 Big Data study, Nicole highlighted the Obama administration’s findings about how big data is transforming how we live and work, and how the private and public sectors can seize new opportunities while managing the risks. The report highlighted ways to make government more data-driven, but also drew attention to the challenge of data discrimination, which can arise when algorithms or machine learning are built from data that is incomplete or collected with bias.

Clarence Wardell provided background about the president’s Police Data Initiative. In the aftermath of high profile officer-related deaths, the White House created a 90-day Task Force on 21st Century Policing last December. The task force traveled the country and listened to law enforcement, citizens, and activists to gather feedback on how to improve policing. It then delivered a report with 59 recommendations, 14 of which used data and technology to increase transparency and build greater trust in criminal justice institutions.

As a result, the White House launched the Police Data Initiative. Clarence and colleagues initially collaborated with 21 local police departments that each committed to release three new open, machine-readable data sets about policing, to help communities better understand policing in their neighborhoods. Data sets about police-citizen encounters, such as officer-involved shootings, citizen complaints, traffic and pedestrian stops, and use of force incidents will bring needed transparency to help build community trust. The initiative, which has grown to include 27 police departments, is working to establish data and transparency as a central and ongoing part of community policing.

But data is only valuable if put to constructive use. Cynthia Rudin gave an example from her research about how data can be used to help police focus their resources and fight crime more effectively. Cynthia and her collaborators (Tong Wang, Daniel Wagner, Rich Sevieri) are working on automatically detecting crime series, or groups of crimes thought to be committed by the same individual or group. By automating the processes of data extraction and pattern analysis, crime analysts save hours of time and can better focus their expertise on helping create proactive enforcement efforts.

Additionally, the panelists emphasized the difference between data and truth. Data is not always neutral, and new types of data, like video, raise a host of questions. For example, with the growing use of police body cameras, how does a police department establish protocols for when the camera gets turned on or turned off? How does the act of capturing video change the relationship between police officer and citizen? And what are the rights and privacy implications for victims of domestic violence who get captured on a police body camera in a private home?

It was clear from the ensuing conversation that data and technology aren’t a panacea. But better data and new tools are informing the national conversation on race and policing, and can be an important piece of criminal justice reform.

Listen to audio from the panel discussion: