Community Courts and Recidivism (Whatever that means!)

Hello everyone! My name is Juwon and I am on the Community Courts team for the DSSG-ATL intern program. This week has been a lot of brainstorming and getting our ideas together and figuring any issues we’re going to have to tackle in the coming weeks.


Midtown Community Court, NYC

What is a community court?

A community court is a more community-focused court in an area’s judicial system that typically tries to use more progressive approaches to criminal justice. Most community courts try to rehabilitate criminals and make them not want to be criminals anymore through various programs, rather than simply throwing them in a cell and calling it problem solved. These courts recognize that the problem is not solved after a criminal has been given a sentence. A criminal will typically commit crimes multiple times. This phenomenon is called “recidivism”.


What are you doing to help Atlanta’s community court?

Currently, the court doesn’t have a better way to store data than Excel spreadsheets. Our first order of business is going to be making the spreadsheet data useful and come up with different ways to map crime, predict criminal behavior, predict recidivism, etc. Right now it seems like recidivism is something the court is very interested in and our goal is to come up with some sort of model to help predict who will slip back into a life a of crime and who won’t. Hopefully we’ll also be able to help the court come up with a better way to keep all their data. Excel spreadsheets are easy and readily available but not really ideal. It could be that they just don’t know what’s available, can’t afford it, or need a piece of software that doesn’t exist yet. Whatever the issue, we’ll help them.


What does recidivism have to do with data?

Recidivism is the problem that arises from ineffectual approaches to criminal justice. If we can predict which criminals are prone to recidivism, that is, committing another crime, we can also figure out which programs are actually ineffective, and use an effective program that would steer the criminal away from a life of crime. Some controversy can arise from this prediction, though. Pointing your finger at someone based on some data run through a computer and stating that they are likely to commit another crime is just wrong on so many levels. Our team will definitely need to be careful with our approach to any predictive models regarding recidivism.


I’m really excited to see what this project brings and I’ll be keeping y’all posted on our findings!

Leave a Reply