Over more than a decade of education reporting at the Milwaukee Journal Sentinel, reporter Erin Richards often found herself hearing variations of the same comment from teachers when she asked them about why it is so hard to improve educational outcomes for children:
“You don’t understand,” they’d say. “I start the year with 30 kids, and I end the year with 30 kids, but it’s not the same 30 kids.”
Richards set out to understand.
Her effort launched a massive undertaking by a team at the Milwaukee Journal Sentinel to obtain and analyze a never-before-released database that tracked student-by-student movement among Wisconsin schools and to illustrate the causes and consequences of churn through the stories of individual families and schools.
Their reporting showed student churn drives a hidden crisis that is ravaging prospects for educational improvement. Each year, 22,000 students shift schools, often mid-year. That amounts to 1 in 4 students in the Milwaukee school system. In some individual schools, the annual turnover tops 40 percent.
While school choice proponents tout the virtues of easy movement, the vast majority of students are cycling from one low-performing school to another. What’s more, even if a student in a high-churn school stays put, their performance can suffer.
Their work is groundbreaking in every sense of the word.
Simply getting access to the data required ingenuity on our part. Analyzing the data – millions upon millions of individual records – presented its own challenges.
Among them: Accounting for the many ways a student could leave a school naturally (ie. moving from elementary school to middle school) and distinguishing those movements from unnatural ones (ie. mid-year transfers). What’s more, they had to account for changes within the schools themselves – for instance, changing from a middle school to a K-8 school. In other cases, schools closed, opened or were reconfigured by the district. This was especially problematic when it comes to independent charter schools and – for the most recent years – voucher-funded students in private schools.
We were able to obtain and analyze more than 12 million lines of student-level data for all of Wisconsin’s students over 13 years. We combined that with state test scores and school report card scores and showed the negative impact on achievement for students who moved – and how the churn is concentrated among the lowest performing schools in Milwaukee.
While the data work was extensive and challenging, we were determined to present it in a way that is accessible to readers, with data visualizations, narrative stories, searchable tables and more. Getting to the final product took sophisticated investigative reporting techniques, skilled data analysis and old-fashioned shoe-leather reporting.