Belief in Authority: Impact of Authority in Multi-Agent Evaluation Framework

TL;DR

This paper systematically analyzes how authority roles influence multi-agent evaluations, revealing that expert and referent roles exert stronger influence than legitimate roles, primarily through consistent positioning rather than active conformity.

Contribution

It introduces a novel analysis of authority bias in multi-agent systems using French and Raven's theory, highlighting the importance of role clarity and position statements.

Findings

Expert and Referent roles exert stronger influence than Legitimate roles.

Authority bias arises from role stability, not active conformity.

Clear position statements are necessary to generate authority bias.

Abstract

Multi-agent systems utilizing large language models often assign authoritative roles to improve performance, yet the impact of authority bias on agent interactions remains underexplored. We present the first systematic analysis of role-based authority bias in free-form multi-agent evaluation using ChatEval. Applying French and Raven's power-based theory, we classify authoritative roles into legitimate, referent, and expert types and analyze their influence across 12-turn conversations. Experiments with GPT-4o and DeepSeek R1 reveal that Expert and Referent power roles exert stronger influence than Legitimate power roles. Crucially, authority bias emerges not through active conformity by general agents, but through authoritative roles consistently maintaining their positions while general agents demonstrate flexibility. Furthermore, authority influence requires clear position statements, as neutral responses fail to generate bias. These findings provide key insights for designing multi-agent frameworks with asymmetric interaction patterns.