Board independence

Why does a board appointed to keep watch over the managers so often end up agreeing with them?

The standard story is that shareholders cannot run the company themselves, so they appoint a board to watch the people who do. The board hires and fires the chief executive, approves the strategy, signs off the accounts. On paper the managers work for it.

Then look at what boards actually do: the strategy management proposed is approved, the pay committee awards roughly what management's consultants suggested, and the chief executive is usually replaced only after the damage is public. The instinct is to call this weakness or cronyism. But if the failure were about character, we would expect it to be scattered. It is not. So what constraint makes a conscientious director behave this way?

Start with the appointment. Who chooses a director? Formally, shareholders elect them. In practice, in most listed companies, the board proposes a slate and shareholders ratify it, and the incumbent chief executive has substantial influence over who is put forward. So the monitor is, in the ordinary case, selected with the consent of the person being monitored. Notice this needs no corruption to bite: even a rigorous chair recruits people they expect to work well with, and "works well with" and "will not obstruct" are hard to separate before the fact.

Now the information. A non-executive director attends perhaps eight meetings a year. Almost everything they know about the company arrives in a pack assembled by management, on an agenda largely set by management, presented by managers who chose which alternatives to describe as realistic. This is not deception; it is the only feasible arrangement, since nobody else has the knowledge. But it means the board's picture of the firm is drawn by the party the board exists to assess. Ask yourself what it takes to disagree with such a pack. You need a rival account, and there is nowhere to get one.

Then the payoff. Suppose a director suspects the acquisition on the table is a mistake. What does pressing it cost them? Time they do not have, the reputation of being difficult, and a real chance of not being renominated. What does it earn them? A share of the value preserved — which, for a director holding a fee and a modest shareholding, is close to nothing. The gains from monitoring go almost entirely to dispersed shareholders; the costs fall on the individual who monitors. That is the whole asymmetry in one line, and it is the same arithmetic that makes a diffuse public lose to a concentrated industry in a regulatory hearing, transplanted inside the firm.

The last ingredient is social rather than economic. Boards operate by consensus, and a norm develops that disagreement is voiced privately and votes are unanimous. The norm is useful — it lets a board present one face — but it converts a single dissent from an argument into an act of disloyalty.

Put the four together and the well-behaved board falls out without anyone behaving badly: selected with management's blessing, informed by management, rewarded almost nothing for catching an error, and operating under a norm that makes catching one socially expensive.

Which suggests where the exceptions should be. If the mechanism is selection plus information, then boards where a director is put there by someone with a large stake and an independent line of sight — a block-holder, a private-equity sponsor, a creditor in a restructuring — should behave visibly differently: sharper questions, faster removal of a failing chief executive. That is the falsifiable part. If turnover after poor performance and the incidence of genuine dissent turned out to be the same on such boards as on boards of formally independent but management-nominated directors, this account would be wrong, and something like director quality or courage would have to carry the explanation instead.

The word "independent" is doing something misleading here. In listing rules and codes it means a relational test — not an employee, no material commercial tie, not a close relative. That is a proxy, chosen because it is cheap to verify, for the things that actually matter: whether the director is informed by anyone other than management, whether they have a stake worth defending, and whether they can be removed for asking. A person can pass the formal test completely and be dependent in all three senses that count. This is why the empirical literature on whether independent boards improve firm performance is, as I recall it, decidedly mixed — not because oversight does not matter, but because the variable being measured is a form, not the substance.

It is worth being blunt about one part. Some of the independence apparatus is best understood not as constraint but as legitimation: an independent pay committee, advised by consultants benchmarking against a peer group, produces awards that are defensible precisely because the procedure was followed, and benchmarking to a peer median ratchets upward whenever any firm below the median moves toward it. That is a machine for raising pay wearing the costume of a check on it.

The cures that touch the mechanism attack selection, information or exposure directly rather than the relational test: a chair who is not the chief executive, sessions held without management in the room, a board budget for its own advisers, directors nominated by parties with real money at stake. None makes a board an adversary of its managers, and it should not be — a board that cannot be told anything is as useless as one that believes everything.

• Why activist investors can dislodge a chief executive that the sitting board could not.
• How employee representation on boards, as in German codetermination, changes the information reaching the room.

Every term the collection defines is gathered in the glossary.