Thousands of people walk through the doors of the Cuyahoga County Justice Center each year to have their felony cases heard by judges. Most of them are Black, and most of them have been there before.
The Cleveland community has been blindfolded in trying to understand how systemic unfairness stacks up — including arrests and charging decisions, access to bail, plea bargains, and sentencing. With our ground-breaking local investigation, “Testify,” The Marshall Project is uncovering the inner workings of justice in the historically opaque Cuyahoga County Justice system.
Like many places, felony court data in Cleveland is both completely open — anybody can look up a case on the docket — and extremely opaque because the court refuses to release the data it collects. The Marshall Project built a scraper to collect 70,000 court records and counting and instead of hoarding the data for its own stories, journalists created a tool where residents and community members could ask questions that the journalists would answer.
We held a virtual community office hours event where we connected with Clevelanders to get feedback about our upcoming tool answering data questions about the courts. Cleveland Documenter Seanna Denee said she wanted us to provide data that “proves the theory on the street” — evidence that helps us understand our own senses.
Our on-the-ground engagement reporting and collaboration with Cleveland Documenters (soon to be part of a new publication called Signal Cleveland) built trust in the community and reflected the community’s questions and concerns about the court system and judges across many platforms.
The Marshall Project provided the community with a first-of-its-kind analysis of voting patterns that highlighted the fact that mostly white, suburban voters were electing the judges presiding over the cases of mostly Black Cleveland residents. The analysis, which was also shared in an accessible, one-page visual explainer, sparked community conversation. As [Ibram X. Kendi](https://twitter.com/DrIbram/status/1486822436653969416) tweeted, “This is the power of journalism.”
We republished the work with a slate of local partners — Cleveland Documenters, WOVU, Real Deal Press, The Cleveland Observer (digital and print), The Cleveland Scene (digital and print), The Land, and Ideastream — to ensure broad local distribution, and to support Cleveland-based media. We also created [this page](https://www.themarshallproject.org/2022/01/27/reprint-testify) to make the process of reprinting our work as easy as possible, and translated our visual explainer into Spanish.
We made extensive use of Python, Scrapy, Docker, and Amazon Elastic Cluster Service to acquire the data and load it into PostgreSQL. We used Hasura to provide a GraphQL API, and Observable notebooks to load and analyze the data.
The website we scraped had many anti-scraping measures and the court has denied direct access to bulk records for years.
To get around them we launched thousands of individual scrapers in the cloud. That also made things slow. Patience was a virtue because it took months of continuous operation to scrape just a few years of case records.
Solving the technical problem was only part of the battle. We still needed to validate whether our scraped data was a comprehensive record of cases with limited cooperation from the court. Through dogged reporting efforts of the entire team, we were able to obtain a list of case numbers from court administrators. While the court would not confirm or deny many of our findings, comparing the list of cases known to the court to the cases we scraped gave us full confidence we had captured the full universe of cases successfully.
Context about the project:
While Cuyahoga County allows anyone with access to the internet to look up a person’s case records, there’s no way to use those individual records to assess the records of individual judges, and the court has not released that data publicly. We were able to access and analyze that data. This is what we found:
* Court outcomes worsen existing racial disparities. Though Black people make up only about 30% of the county’s residents, almost two-thirds of the people who are arrested by police and charged with felonies by prosecutors are Black. Then, after judges impose sentences, state records show three-quarters of people in state prisons convicted in Cuyahoga County are Black.
* Individual judges make a big difference — for example, some judges almost never send defendants to prison for common charges like theft and low-level felony drug possession, while others incarcerate 30% or more.
* While Cleveland residents make up two-thirds of defendants in the court, votes from the city account for less than a quarter of those cast in judges’ races. That means the vote in the predominantly White suburbs in judges’ races effectively carries three times the power of the vote in the majority Black city.
* Voters have more power than they may think. If everyone who showed up to vote had cast ballots for judges as well, that could have swung the outcome in 9 of 15 contested judicial races since 2016 — without turning out a single additional voter.
Judges play a key role in the way justice is dispensed. Bringing transparency to how they do their jobs is essential to understanding flaws in the larger system. It’s also urgent for voters, who have the task of electing them or voting them off the bench. That’s what The Marshall Project heard from a chorus of academics, attorneys, people who have experienced the system firsthand, and more than 40 residents we worked with Cleveland Documenters to interview.
We also spent months gathering questions from the community, answering their questions and further trying to shed a light on the imbalance in power when it comes to electing judges.
Finally, our reporting on the court system continues: Our most recent investigation, by Wesley Lowery and Ilica Mahajan, found the majority of cases in the system feature defendants with at least one prior charge, and nearly a third have at least five prior criminal cases. The crimes are mostly not violent. A review of hundreds of cases found that most included a defendant who cited drug addiction, mental illness, or both as a factor in their crimes.
What can other journalists learn from this project?
1) Courts are worth investigating. After policing, courts are the most important gear in the machinery of mass incarceration in the United States. They are also one of the most opaque.
2) Analyze geographic concentrations of power. When our understanding of the data was still limited, we were able to rely on easier-to-use public voting datasets combined with geocoded defendant addresses.
3) Make the work a two-way street. Through our community “office hours” virtual meetings and our FAQ product, we’ve built channels for engaging with the community.
4) Target tons of platforms and audiences. We’ve shared our reporting in printed fliers, as an illustrated explainer, as an interactive chart, as a page of audio voices, on social platforms like Instagram, in Spanish, at local civic data events, and more.