Old-boy network's power exposed

PHILIP BALL

Cliques make it hard to keep things above board. © GettyImages

Cliques of well-connected businessmen can easily corrupt or distort corporate board decisions, but now a team of scientists say they can assess how much power old-boy networks have over boardroom meetings.

"A well-connected lobby of a minority of directors can drive the decision of the board," say Stefano Battiston of the Ecole Normale Supérieure in Paris and co-workers¹. But they think that it is possible to predict the chances that a board will agree with the opinion of such a lobby.

If this is so, the power of the lobby can be assessed and the changes needed to break its dominance identified. It all depends, say the researchers, on how many members of a board also sit together on other boards.

It's a common situation. Almost 100 years ago Louis Brandeis, a judge in the US Supreme Court, spoke of a "financial oligarchy controlling the business of the country". Not much has changed. Directors and board members of large companies form a kind of social network in which many individuals sit with at least one other person on more than one corporate board.

Battiston's research reveals that four of the directors of the Bank of America Corporation, for example, each sit with one of the other three on at least one outside board. Researchers argue that board members who are 'interlocked' in this way are more likely to influence one another's opinion than those who never see one another outside a single boardroom.

It goes without saying that cliques such as this, if they are large enough, can engineer a board decision in their favour. But just how great is this power, if the clique doesn't have a majority? And how much does that depend on the way decisions are reached?

Battiston and colleagues have devised a model that describes the voting behaviour when boards have to make some decision. Typically, this entails the chief executive officer (CEO) proposing a certain strategy and the other board members deciding whether or not to support it. Each member is assumed to be influenced by the decisions of others if they sit on another board together; otherwise, they decide for or against independently.

By considering various possible networks in a board of ten members (a typical size for a corporation), the researchers show that each network can be assigned a 'force' which measures the likelihood that the board will approve the CEO's proposal. If the board has, say, three links between members (a link indicates that they sit on an outside board), then 40% of the possible networks have at least a 75% chance of approving the proposal. For a non-interlocked board - one with no old-boy network - this chance is close to 50:50, as it should be. In other words, even a small amount of cliquiness can make a big difference.

Awareness of the power of cliques could help to offset their influence, Battiston's team suggests. But busting up old-boys' networks might not be a solution. The way in which people get onto boards in the first place may propagate the trend: an existing board member may often recommend candidates he (rarely she) already knows through joint positions on another board - and probably ones that share his views.

1. Battiston, S., Bonabeau, E. & Weisbuch, G. Decision making dynamics in corporate boards. Preprint, (2002). |Article|

© Nature News Service / Macmillan Magazines Ltd 2002