Hierarchical oversight lowers clarity and raises cost in llm teams
Loop-Back Authority in LLM Agent Teams: A Paired Experiment on Flat and Hierarchical Coordination
Multiagent SystemsArtificial IntelligenceComputation and Language
Summary
Many systems that use teams of AI language models have a manager role that checks workers' answers and can ask for changes. The authors tested whether this manager role actually improves quality by comparing setups with and without it, using the same AI models and prompts. They found that having a manager who requests revisions didn’t improve accuracy but did reduce how clear the writing was and increased the cost in computing resources. This suggests that managers help only when they can verify answers, not just give opinions.
What this means in practice
- •For ai system designers: Optimize multi-agent AI workflows by skipping manager oversight when verification isn't possible, saving costs and improving output clarity.
- •For content generation teams: Use flat coordination among language models to produce clearer reports for business tasks without extra revision overhead.
Authors
Burak Agachan, Max van Duijn, Amirhossein Zohrehvand
Abstract
Hierarchical orchestration, in which a Manager agent reviews worker output and can send it back for revision, is the default coordination pattern in production multi-agent LLM frameworks. Classical organizational theory predicts that the authority link speeds convergence on decisive output; work on sycophancy and Degeneration-of-Thought predicts that authoritative critique makes LLM output worse. Prior comparisons vary whole frameworks on tasks with checkable answers, leaving the authority link untested on open-ended work. We present a paired experiment that holds five LLM agents, their roles, prompts, tools, models, and data fixed and varies one link: whether the Manager may reject a worker's output and oblige a revision. Across 43 paired products and 86 runs of a business-intelligence reporting task, a five-model judge panel and a deterministic specification check score every report. The flat organization scores higher on Utility (d = 0.42, p = 0.009) and on Writing Clarity (d = 0.34, p = 0.030); the classical prediction fails. The reports are the same length, but hierarchical reports hedge 53% more, each revision loop is associated with a 0.14-point drop in Writing Clarity, and the hierarchical Writer's first draft is indistinguishable from the flat report: the gap opens inside the revision loop. Specification accuracy is at ceiling in both organizations, and the supervisory tier costs 51.5% more tokens for no quality gain. A supervisor pays for itself when it can verify and becomes a liability when it can only opine.