"It's Reducing a Human Being to a Percentage": Perceptions of Justice in Algorithmic Decisions

Abstract

Examines how people perceive fairness when algorithms make consequential decisions about them. Participants expressed deep discomfort with being "reduced to a number," revealing psychological tensions between efficiency and dignity in algorithmic governance.

Key Findings

• People feel dehumanized when reduced to statistical predictions
• Different explanation styles (input, demographic, case-based) affect justice perceptions differently
• Concerns about algorithmic decisions are rooted in identity and autonomy, not just accuracy

Source