[Paper Review] Spread of decisions in the corporate board network

Stefano Battiston, Gérard Weisbuch+1 more|arXiv (Cornell University)|Dec 9, 2003

Corporate Finance and Governance10 references3 citations

TL;DR

This paper models decision spread in corporate board networks using a herding mechanism on empirical interlock data. It finds that information about interlocked boards' decisions—rather than imitation or opinion bias—drives rapid, widespread consensus on macroeconomic strategies, explaining synchronized corporate decisions without external shocks.

ABSTRACT

Boards of large corporations sharing some of their directors are connected in complex networks. Boards are responsible for corporations' long-term strategy and are often involved in decisions about a common topic related to the belief in economical growth or recession. We are interested in understanding under which conditions a large majority of boards making a same decision can emerge in the network. We present a model where board directors are engaged in a decision making dynamics based on "herd behavior". Boards influence each other through shared directors. We find that imitation of colleagues and opinion bias due to the interlock do not trigger an avalanche of identical decisions over the board network, whereas the information about interlocked boards' decisions does. There is no need to invoke global public information, nor external driving forces. This model provides a simple endogenous mechanism to explain the fact that boards of the largest corporations of a country can, in the span of a few months, take the same decisions about general topics.

Motivation & Objective

• Understand how decisions on macroeconomic topics (e.g., investment, advertising) spread across interlocked corporate boards.
• Investigate whether herd behavior or opinion bias from shared directors can trigger large-scale consensus.
• Determine whether endogenous information flow—rather than external signals—can explain synchronized decision-making in large corporate networks.
• Model the dynamics of decision adoption on a real-world, heterogeneous network of Fortune 1000 board interlocks.
• Assess the role of meeting order and information access in enabling global consensus on binary decisions.

Proposed method

• Construct a network of corporate boards connected by shared directors using real data from Fortune 1000 firms (Davis, 2001).
• Model directors as agents with binary opinions (+1 or -1) representing decisions on general economic topics.
• Implement two models: Model 1 assumes initial opinion bias from prior board decisions; Model 2 includes information about interlocked boards’ decisions during deliberation.
• Simulate decision dynamics using a threshold-based update rule where directors adopt the opinion of the majority of their connected peers.
• Apply a time-scheduled meeting order based on distance from the central board to test the impact of information propagation timing.
• Use the Ising model framework for opinion dynamics, with parameters for influence strength from colleagues (α₁) and from interlocked board decisions (α₂).

Experimental results

Research questions

• RQ1 Can herd behavior among directors and opinion bias from interlocks alone trigger widespread consensus on corporate decisions?
• RQ2 Does information about decisions made by interlocked boards significantly increase the probability of consensus across the network?
• RQ3 How does the scheduling of board meetings affect the emergence of a global consensus?
• RQ4 Is public information or external shocks necessary to explain synchronized decision-making in corporate networks?
• RQ5 What structural and dynamic features of the board network enable rapid convergence on a single decision?

Key findings

• Model 1, which relies only on imitation and opinion bias from interlocks, results in only a 60% chance of adopting the central board’s decision when γ = 0.1 (realistic interlock level), indicating weak influence.
• Model 2 shows that information about interlocked boards’ decisions (α₂ > 0) dramatically increases consensus, achieving P⁺ = 0.9 for both CMB Net 1 and CMB Net 2 when α₂ ≈ 0.1.
• The probability of consensus is highly sensitive to meeting order: a schedule respecting network distance from the center is necessary for high consensus rates.
• Information about interlocked boards’ decisions has a stronger effect than imitation or opinion bias, even at moderate influence strength.
• The model demonstrates that global consensus can emerge endogenously through local information sharing, without requiring public announcements or external shocks.
• The radiation-from-the-center hypothesis—where peripheral boards learn from central boards via interlocked directors—explains real-world diffusion patterns observed in corporate governance.