Ronald Coase famously pointed out that the reason we have firms is that there are costs to using the market, transaction costs as we call them now. The existence of transaction costs means that firms can carry out some activities at a lower cost than if we were to transact across the market. But this raises an obvious question, If firms are so great why are there markets? Or, in other words, why isn't everything done in one huge firm with no recourse to markets?
Coase's answer to this question is that there are organisational costs as well as transaction costs. In his 1937 paper Coase asked the question as to why, if by creating a firm, the costs of production can be reduced, are there are any market transactions at all? “Why is not all production carried on by one big firm?” (Coase 1937: 394). The answer according to Coase is, first, that as a firm gets bigger there may be decreasing returns to entrepreneurial activity. That is, the cost of an additional transaction being organised within the firm may rise. Second, as the number of transactions which are organised in-house increases, the entrepreneur may fail to place the factors of production in the uses where their value is maximised. This means that the entrepreneur fails to make the best use of the available factors of production. Finally, the supply price of one or more of inputs to production may increase because the ‘other advantages’ of a small firm are greater than those of a large firm. As a result of these factors a firm will grow to the point where the costs of organising an additional transaction within the firm become equal to either the costs of carrying out the same transaction via the market or the cost of organising the transaction in another firm (Coase 1937: 395).
In more modern terms it would be argued that firms have to cope with, for example, problems such as shirking, which can be covered up in situations of team production; informational complexity and asymmetries; monetary calculation, as the socialist calculation debate made clear; rent seeking and an increase in principal-agent problems.
As firms get larger, teams get bigger and monitoring costs go up. As teams get larger the possibility of shirking increases as the monitor has increased difficulty in detecting that shirking is occurring and who is doing it when it does. When a firm does more in-house appropriable quasi-rents are likely to grow making rent seeking more profitable and more likely. While rent seeking is "personally" profitable it is "socially" destructive and thus decreases the efficiency of an organisation. The larger the is organisation, the larger the asymmetries in information are and therefore the costlier it is to provide the correct incentives for mangers and workers to use their knowledge efficiently and the harder it is to design hierarchies so that the right people get the information they need and have the expertise to use it and to get people to share their information. The costs of communicating, the number of communication channels, delays in information transmission etc all raise as the firm gets bigger. As an organisation become larger with a greater number of levels of management there is an increase in the principal-agent problems that arise. More incentive mechanisms have to be developed, implemented, and monitored, in an effort to ensure that an efficiently working organisation will be created. Ultimately firms will be limited in size by the fact that as they grow and do more in-house they eliminate capital goods and this increases the costs of internal corporate planning and the eventually such costs become prohibitive. This is essential an application of the arguments in the socialist calculation debate to the firm.
So Coase gave us a framework, albeit one that he didn't develop as far as we may have wished, which can be utilised to explain why central planning is limited in the range over which it will work. Central planning only works over a limited range in terms of a firm's size. A firm can only ever be so big. The market will at some point be a cheaper way to manage transactions.
Coase, Ronald Harry (1937). ‘The Nature of the Firm’, Economica, n.s. 4(16) November: 386–405.