Perceived Fairness in Networks

2025-10-13

AI Summary

Traditional fairness definitions focus on group-level statistical metrics (e.g., demographic parity), yet individuals’ subjective fairness perceptions often arise from local social comparisons rather than global statistics. This work identifies a critical gap: even when algorithms satisfy classical fairness criteria, network homophily can induce systemic *subjective* discrimination perceptions. To address this, we formally define *perceived fairness*—a novel construct that embeds individual local observations within network topology, thereby bridging objective fairness guarantees and subjective cognitive realities. Leveraging graph-theoretic analysis, game-theoretic modeling, and multi-agent simulation, we quantify how network structure amplifies fairness perception bias. Results demonstrate that higher homophily correlates strongly with increased perceived unfairness. Our framework provides the first structure-aware, cognition-grounded methodology for evaluating algorithmic fairness—particularly relevant for applications such as financial inclusion and peer-based insurance. (149 words)

Technology Category

• Philosophy and Ethics of AI: Bias, Fairness & Equity
• Computer Vision: Bias, Fairness & Privacy
• Game Theory and Economic Paradigms: Fair Division

Application Category

• Social Networks and Social Media: Fairness and bias in social network and social media analysis
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web
• Economics, Online Markets and Human Computation: Fairness, privacy, and diversity in economic environments

Abstract

The usual definitions of algorithmic fairness focus on population-level statistics, such as demographic parity or equal opportunity. However, in many social or economic contexts, fairness is not perceived globally, but locally, through an individual's peer network and comparisons. We propose a theoretical model of perceived fairness networks, in which each individual's sense of discrimination depends on the local topology of interactions. We show that even if a decision rule satisfies standard criteria of fairness, perceived discrimination can persist or even increase in the presence of homophily or assortative mixing. We propose a formalism for the concept of fairness perception, linking network structure, local observation, and social perception. Analytical and simulation results highlight how network topology affects the divergence between objective fairness and perceived fairness, with implications for algorithmic governance and applications in finance and collaborative insurance.

Problem

• Modeling perceived fairness through local network interactions and comparisons
• Analyzing divergence between objective fairness metrics and subjective perceptions
• Investigating how network topology affects discrimination perception in algorithms

Innovation

• Modeling fairness perception through local network topology
• Linking network structure to social perception of fairness
• Analyzing divergence between objective and perceived fairness