Oral Histories of Surveillance

Text Exploration

In this interview, Pfefferkorn, interviewed by Spencer-Notabortolo, describes their experience working in Minnesota to create the Coalition to Stop the Cradle to Prison Algorithm—an effort to eliminate predictive analytics and data tracking in lower-socioeconomic, predominantly black and brown educational environments.

Pfefferkorn describes how a joint powers agreement was developed in Ramsey County that would merge data between the city of St. Paul and the St. Paul Public Schools. This agreement was made to create “an early warning system through predictive analytics to identify a risk factor for a student.” When questioned, the creators of the agreement struggled to articulate what would happen if a student was actually identified as being at-risk, and they simply reverted to what they were already doing as interventions (that weren’t effective). Pfefferkorn calls this a “deficit-based approach,” one that works towards correction rather than empowerment and growth; this particular deficit-based approach was one that would create a school-to-prison pipeline.

In addition to the actual content of the agreement, Pfefferkorn argues that the entire process leading up to the agreement was flawed. First, it was based on a presumption that data and algorithmic predictive analytics were good and desired by the community; when questioned, leaders in support of the agreement pushed back on criticism by claiming critics did not support innovation. The agreement was also based on a presumption that any community involvement in the discussion of the technology was informed, and this was false: residents didn’t understand the technology, and even those supporting the technology didn’t really understand it, either. Pfefferkorn describes that “we found out that the unanimous decision to approve this JPA, most of the folks that signed off on it did not understand what they were signing off on. They did not know what predictive analytics was.”

Pragmatically, the agreement had no plan for continual funding. The initial efforts were provided for free by a technology provider, but when asked, the advocates for the agreement were unable to identify a budget to support the program after the trial.

Pfefferkorn describes a strange irony to the experience—that those in support of the program were those that were often on the “same side” as the residents: democrats, who in theory were advocates for equal rights and equal educational experiences. Yet they were so enamored with the promise of innovation and the magic of an “algorithm” that the agreement had inertia, and until the Cradle to Prison group raised awareness of the potential risks of the program, they were unable to see beyond the promise of technology as empowerment.

Pfefferkorn ends by describing the ongoing political context in which discussions like this occur; they note that it’s become important to question what’s behind an invitation to give a presentation or to engage in a discussion about technology, because technological “advancement” is wrapped up in competing interests and perspectives.