How to Make Federal Justice Data More User Friendly

Chris Cutler was born in Foxborough, MA, and lived just 10 minutes from Gillette Stadium, where the New England Patriots play. But football wasn’t his focus in middle school and high school. The crime show CSI was. That sparked an interest initially in forensic science and then data science, where he has made his mark in his work at Abt for the Bureau of Justice Statistics (BJS) on the Federal Justice Statistics Program (FJSP). A graduate of the University of Miami with a master’s in criminal justice from State University of New York, Albany, he likes managing large data sets and investigating data issues to make a more harmonious final product for end users. Data quality has always been his primary focus and where he gets the most joy from his work. Chris sat down with us to discuss what he’s been doing and his plans for the future.

Q: Tell us about your background, how you got into what you're doing.

A:  I began college as a biology major with a goal of becoming a forensic scientist. My career trajectory changed during my junior year after an in-lab internship on a cancer study at the University of Miami Medical School. One of my primary tasks was to count the number of cells that had migrated from one side of the membrane to another every two hours. I ended up counting hundreds of cells each week, often with little tangible data to show for my efforts. At the end of that semester, I thought, “this is awful” and realized that a career working with a microscope wasn’t for me. At the same time, I began taking more criminal justice classes to broaden my skillset and eventually completed a second major in criminal justice. 

Then I was accepted into a criminal justice Ph.D. program at SUNY Albany. While working on a research project with Dr. Shawn Bushway, I had my first opportunity to write a statistical program. It's funny to look back at my trajectory, because growing up, I loved math but was not a big computer person. My typing was, and still is, hunt-and-peck. The fact that all my work is on a computer and it's coding-based is funny to me. 

I've always been a detail-oriented person, which sometimes leads me to miss the forest while examining the trees. Tracking down that 1 percent of cases that no one cares about because they don’t impact the whole story--that was where I found my niche. I liked trying to understand what was happening and why. What was different about these cases? What could we learn to improve the product?

I quickly realized that a Ph.D. was not a great fit for me. I left after I got my master's and returned home to Boston. A couple of months later, I got a job at Abt, in April of 2014. I've been working on our criminal justice project for more than 10 years. It's just crazy to think about.

Q: Can you explain the Federal Justice Statistics Program FJSP?

A: The FJSP collects and standardizes administrative data on individuals involved in the federal justice system and has been operated by the Bureau of Justice Statistics (BJS) since the 1980s. On this project, we collect data from federal agencies spanning arrest and booking (U.S. Marshals Service), investigation (Executive Office for U.S. Attorneys), court proceedings (Administrative Office of the U.S. Courts), sentencing (United States Sentencing Commission), and incarceration (Federal Bureau of Prisons). New data are processed each year and standardized to align information across years and agencies.

There are two primary strengths of the FJSP data collection: the longevity of the series and the dyad link files, which connect individuals in adjacent agencies. Presently, the FJSP spans nearly 30 years (fiscal years 1994 – 2022). Deidentified versions of these data are available for researcher use with an approved Institutional Review Board package through the National Archive of Criminal Justice Data (NACJD). A limited set of variables for fiscal years 1998 – 2022 are available for querying through the project’s public web tool, the Federal Criminal Case Processing Statistics (FCCPS). Individuals interested in more information on the project or the data available should go to the BJS website or NACJD.

The dyad link files connect individuals in adjacent stages using common person and case level identifiers. The capability allows users to follow cohorts of individuals from arrest and booking through incarceration, which expands the types of research questions that can be investigated. 

Abt has been the contractor on this award since 2013, but we haven’t rested on our laurels. We are perpetually looking for ways to improve the quality, timeliness, and breadth of the data collection. When we received data for the first time from the Executive Office for Immigration Review (EOIR) for individuals in immigration courts, I led the development of processing programs to create annual data files from the raw agency data. After months of painstaking work, we presented our standardization approach to the agency. During this meeting, representatives from EOIR indicated that our efforts were the closest that any external group had come to replicating their published statistics.

FJSP has been an underutilized resource for the criminal justice research community. We're always looking for ways to excite the next user who's going to find a dissertation topic or the next paper idea from the FJSP.

Q: Can you explain the difference between the old tool for using data and the upgrade? 

A: The FJSP web query tool is called the Federal Criminal Case Processing Statistics (FCCPS) Data Tool (ojp.gov). It's been around for at least a decade now, and the previous iteration was coming to the end of its life span. The BJS was no longer supporting the technology it was originally built on. The previous version of the FCCPS was good at producing tables but wasn’t super modern in its utility and offered users limited flexibility. 

In November 2023, the FCCPS was transitioned to a Tableau-based platform. We've added a lot of flexibility to enhance the user experience, including increased filtering capabilities and multiple data visualization options (such as bar graphs, line graphs, and heat maps to show the interaction of two different variables). In the original tool, users could select one or two variables to display in a table and filter one additional variable. Each selection was a separate step. That limited how deeply you could dig into the data. With the updated tool, researchers can filter multiple variables at the same time. It gives users a lot more flexibility to really dig down into the weeds in the data to look at the populations that they're interested in. Abt will continue work on this tool as part of a new award with BJS to expand its capabilities for users.