Network Heterogeneity and Value of Information

TL;DR

This paper investigates how network structure influences the value of public information, revealing that transparency can sometimes harm welfare in tiered networks due to overreaction by central agents.

Contribution

It uncovers the non-monotonic effects of public disclosure on welfare in different network topologies, especially core-periphery structures.

Findings

Public disclosure improves welfare in uniform networks.

In core-periphery networks, transparency can reduce welfare.

Core agents over-respond to noisy signals, increasing volatility.

Abstract

Does greater connectivity enhance the value of public information? I study a networked beauty contest game where agents balance adaptation to the fundamental with local coordination. The analysis reveals a stark non-monotonicity: while public disclosure improves welfare when interactions are uniform, regardless of their intensity, it can be detrimental in core-periphery structures. This welfare loss stems from a distortion driven by the core, where core agents over-respond to a noisy public signal, forcing peripheral neighbors to absorb this volatility to maintain alignment. These findings suggest that standard transparency policies can backfire in tiered markets where dominant hubs propagate excess volatility.