The enormity of the country’s criminal justice system — 15,000 state and local courts, 18,000 local law enforcement agencies, more than two million prisoners — looks even more daunting when you consider how little we know about what is actually going on in there.
Want to know who we prosecute and why? Good luck. Curious about how many people are charged with misdemeanors each year? Can’t tell you. How about how many people reoffend after prison? We don’t really know that, either. In an age when everything is measured — when data determines the television we watch, the clothes we buy and the posts we see on Facebook — the justice system is a disturbing exception. Agencies exist in silos, and their data stays with them. Instead, we make policy based on anecdote, heavily filtered through a political lens.
This week the nonprofit Measures for Justice is launching an online tool meant to shine a high beam into these dark corners.
It is gathering numbers from key criminal justice players — prosecutors offices, public defenders, courts, probation departments — in each of America’s more than 3,000 counties. Staffers clean the data, assemble it in an apples-to-apples format, use it to answer a standard set of basic questions, and make the results free and easy to access and understand.
It’s the kind of task you’d expect a federal agency like the Bureau of Justice Statistics, which has an average annual budget of $97 million, to take on. Instead, the 22 people at Measures for Justice’s Rochester, N.Y., offices are doing the work themselves on an annual budget of $4.6 million, donated mostly from foundations.
So far they’ve tackled six states: Washington, Utah, Wisconsin, Pennsylvania, North Carolina and Florida, gathering most of the numbers one county at a time. Together, these make up 10 percent of the nation’s counties. The team chose those six states for their geographical diversity and — to ease the data gathering in the project’s early phases — because they had unified statewide court databases. The hope is to complete 15 more states by 2020, while updating the statistics from the first six states every two years.
“We’re giving people data they’ve never had access to before,” says Amy Bach, the founder and executive director of Measures for Justice. “We’re telling them stories about their communities and their counties that they’ve never heard before.”
The project, which has as its motto “you can’t change what you can’t see,” centers on 32 “core measures”: yardsticks to determine how well local criminal justice systems are working. How often do people plead guilty without a lawyer? How often do prosecutors dismiss charges? How long do people have to wait for a court hearing? Users can also slice the answers to these questions in different ways, using “companion measures” such as race and political affiliation.
The portal itself is like a video game for criminal justice nerds. Users can compare counties, click on interactive maps and bar charts, and layer one data point upon another. The interface is clean and easy to use, if a little wonky. (The organization wants to present data in context, so each infographic is followed by a screen full of fine print and footnotes.) It’s meant for everyone — not just professors and policy wonks.
If you lived in Montgomery County, Pa., for example, you might click around to find that people in your community are jailed for lack of $500 bail or less at twice the statewide rate. If you lived in Winnebago County, Wisc., you might learn that your local pretrial diversion program is disproportionately enrolling white defendants over black by almost 2 to 1.
“When you start looking at objective data, all of a sudden you’ve got some explaining to do,” says Christian Gossett, the district attorney in Winnebago County. Gossett says that since he learned the results of Measures for Justice’s analysis, he went digging. He learned the disparity was arising not from who was being offered the diversion program but who was accepting it — a relief, he admitted, but still a problem. He has reached out to researchers at the University of Wisconsin to study why and how to address it.
“This is our goal,” Bach says. “We want to generate questions,” especially among decisionmakers in the system such as Gossett. “We want to provide comprehensive, reliable, accurate data, that they can use to ask questions that weren’t being asked before.”
Still, the task is huge enough that it’s inevitably incomplete. Only six of the 32 core measures are available for North Carolina. For Florida, 20. Fiona Maazel, the organization’s spokeswoman, says they expect things to improve. “Data breeds data,” she says. Once county officials see how better numbers can lead to better outcomes, she says, “they’ll say, ‘we want to be a part of it. Let’s improve our data collection practices.’”