In the United States, financial institutions leverage personal data for countless decisions impacting individual wellbeing, ranging from managing access to existing accounts to deciding who to offer credit and on what terms. The legitimacy, and arguably the efficacy, of those decision-making processes therefore hinges partly on the accuracy of the underlying personal data. But what does it mean for personal financial data to be “accurate?” Technical and legal definitions often define data accuracy as an objective quality of the information—it either reflects the correct value, or it does not—thereby merely replacing accuracy with other ambiguous terms like correctness. We are left with the questions: what counts as “correct,” and how are such determinations made? In this paper, a phenomenological account of data accuracy is offered.

Jordan Brensinger is a postdoctoral research associate in the Center for Information Technology Policy at Princeton University.