In its obituary of him, the Times says that Peter Roebuck grew disenchanted with England because he thought it lacked the right combination of egalitarianism and authority. This is a really useful way for thinking about the structure of organizations - be they sports teams, companies or whatever - because there are costs and benefits associated with both*.

The benefits of authority are:

A1. It allows for speedy, decisive decisions. Imagine how much time would be wasted if every decision in your firm required a meeting at which every whinger and windbag could vent endlessly. (You might not have to imagine.)

A2. Unilateral authority allows for conflicts of interest to be resolved quickly. It can actually prevent such conflicts, insofar as potential disputants anticipate the boss's ruling.

A3. It economizes on information -gathering. Where there's a single decision-maker, only one person need bother preparing for it. If everyone had to take decisions, huge amounts of time would be devoted to preparing papers for meetings.

A4. It allows for clear feedback, which in turn maximizes opportunities for learning. If a single boss is responsible for a decision, then if it goes wrong he learns something. If, however, a group is responsible, the error leads to buck-passing and arse-covering and so no-one learns. A similar thing is true for underlings. Feedback from a boss is clearer than that from a group; a man cannot serve two masters.

A5. Decision-making skills are like any other - they are not equally distributed. Some folk are better at managing than others, so the division of labour requires that they specialize.

Against this, there are benefits of egalitarianism:

E1. It can curb rent-seeking by opportunistic bosses.

E2. Where authoritarian monitoring is imperfect, it can improve discipline as workers police themselves. Some people have hated working for John Lewis because it's harder to skive when everyone thinks they're your boss.

E3. It can raise productivity as people work harder when they feel they have a stake (pdf) in the outcome or a say in the decision. As Tocqueville wrote, democracy 'spreads throughout the body social a restless activity, superabundant force, and energy never found elsewhere.'

E4. It can improve decision-making by making better use of the dispersed, fragmentary knowledge of individuals. Also, in authoritarian regimes - be it Fred Goodwin's RBS or Hitler's running of Germany's war effort - leaders are often told what they want to hear rather than what they need to, with sometimes catastrophic effects. Equality prevents this.

The right balance between these will vary from organization to organization. For example, football teams are run on authoritarian lines because A1 and A2 are important whilst E2 is insignificant and E3 - because there are other sources of motivation - is less relevant. A similar thing is true for the army.

I stress the word 'balance' here because there needn't be - and rarely are - pure authority and pure egalitarian structures. In authoritarian structures bosses might pay more or less attention to workers or trades unions, thus having varying degrees of equality. And in egalitarian structures (say jury rooms), some individuals will carry more weight than others.

So, how can we tell what is the right balance? Often, it'll be hard. For one thing, simple metrics such as the number of levels in the hierarchy from CEO to shopfloor won't necessarily help; a few rigid levels might carry more costs of authority than more levels accompanied by an egalitarian ethos.

And for another thing, the balance will depend on individuals. If a boss is a genius, the organization can get away with what would otherwise be an inefficient balance of authority and equality; think of Steve Jobs' bullying at Apple.

As you know, I share Roebuck's hunch, that the balance we have is badly skewed, as we exaggerate the importance of A1-A5 and under-estimate that of E1-E4.

* I'm drawing here on Henry Hansmann's The Ownership of Enterprise and Oliver Williamson's Markets and Hierarchies.