I feel this suggestion is more or less the same as that of Penrose when she notes the limits to the growth of the firm are set by the management team’s ability to learn how to negotiate the disjunction between resources and the services they provide.

So I’m agreeing with Steve. But his use of the term ‘complexity’ may throw a lot of people off. Complexity may have little to do with ‘genuine’ Knightian uncertainty.

One point it makes is that upon Jobs’s return to the helm the company drastically cut down its product portfolio so as to get management focused on making a few great products rather than a host of mediocre ones. (They contrast this with what they incorrectly allege is a tendency of business schools to preach diversification, but what the heck.)

When you read in an interview with Jobs that he a) wants his managers to have a lot of autonomy (with “kibitzing” from him), b) he wants each manager to know all about the entire Apple business so that they internalize tradeoffs properly, and that c) every Monday the top management team spends hours reviewing the entire business from top to bottom, you can see that reducing complexity may be a necessity for Apple to carry on its current style.

According to Jobs, every Monday the management team reviews the previous week’s sales, every single product under development, every product they sell, goes over development snags, capacity problems, and so on. You can see how that style of management would not work if the product portfolio were really diverse and big–the complexity would overwhelm the process. (You can also see how product-oriented Jobs and the whole company is.) So it’s a pretty stark example of complexity limits on managerial oversight.

SRP provides an excellent illustration of ’situated logic’. Then adds that even when this is allowed there is surely some point at which ‘the situation’ becomes so large/complex/multifaceted (?) that the individual’s practice-based logic/judgment is stretched beyond the limits of its ‘bounded-ness’ - to use Simon’s term.

I’m interested in the increasing number of executives who try, in their various ways - sometimes in court, sometimes as they attempt to justify their pay-packets - to explain the modern corporation has reached this critical degree of complexity. No human mind can any longer grasp its globalized multi-cultural multi-technology inter-relations - which likewise reminds me of Perrow’s early work on systems failures. It is not so much that the anti-missile system is bound to fail - along the lines of a column exceeding its Euler buckling height - as the firm, as an anti-poverty system, also fails at a certain point. Which then reminds me of Rumelt’s classic PhD work on related and unrelated diversification.

The problem with this as a critique of Hayek is that Hayek doesn’t make the knowledge problem strictly about action sets. It’s also about preferences and behavior. There is a very good question about how big the size, scope, and diversity of an enterprise has to become before this particular Hayek critique sets in. (But you also have to bear in mind the difference between an organization and an order in terms of shared purpose, which is a separate issue.)

But if we admit judgment into the analysis the question then arises, how is the judgment of others to be ‘controlled’? Knight’s suggestion, that one depends on others who have better knowledge and therefore do not need to apply the same kind of ‘judgment’ is hiding the contradiction - how do you know who to delegate to? This implies a different kind of certain knowledge - the overlap to which Nicolai refers, thereby invoking what Hayek points out is simply NOT available.

Notwithstanding, Knight is one of a handful of people - Simon being another - who ever asked this question in the context of theorizing firms, and is deservedly honored for the impact he made in so doing.

What is needed, of course, are some viable answers. They seem likely to fall into two groups (a) theoretical and (b) matters of practice. By theoretical I mean answers that follow from a more substantial theory of mind than that, or those, which currently dominate our literature. In this sense behavioral economics is clearly a small but significant step forward. Other theories of mind, some based on recent neurological work, are floating around and have encourage some to think the answer is at hand. By practice I imply something closer to ’situated logics’, and there is an expanding neo-Vygotskian literature here.

I would be exceptionally interested in others’ views of what might be done at this stage.