Printer Friendly

The nature of the firm reconsidered: information synthesis and entrepreneurial organisation.

Introduction

The theory of the firm has been one of the most exciting fields of economic research over the last twenty years. Yet despite a vast amount of work, no consensus has been reached on the nature of the firm. Theories of contracts, transaction costs and entrepreneurship vie with each other to form the foundation for a truly comprehensive theory of the firm. Each of these theories has its own idea of what the key issue is, and naturally claims that it alone addresses this particular issue head on. This paper sets out a general framework within which all the key questions in the theory of the firm can be brought together and discussed at once. The range of different issues discussed is summarized in Figure 1. The pattern of the figure reflects the nature of the synthesis attempted in this paper.

The framework is based upon the concept of volatility. It is supposed that the economic environment is continuously disturbed by shocks of both a persistent and transitory nature. Persistent shocks provide a stimulus to the formation of new firms or the radical restructuring of existing firms. Persistent shocks are intermittent and diverse, and are usually dealt with by improvisation. This improvisation is effected by the entrepreneur who founds the firm. By contrast, transitory shocks are repetitive and conform to a more limited number of types. They are dealt with routinely using procedures devised by the entrepreneur. Applying a division of labour to the implementation of these procedures creates the organisation of the firm.

Information on shocks is costly to collect and communicate. The initial impact of shocks is localised and dispersed, which means that some people get to learn of them before others. Those 'in the know' buy up resources which have become more valuable as a result of the shock, in order to make a speculative gain. The gain is realised when the resources are deployed to a more profitable use. The most significant type of shock, so far as firms are concerned, is one which creates a new market opportunity. The shock could be a change in tastes, factor costs, technology, social values, or indeed anything that impinges on the gains from a particular type of trade. A firm is created whenever an entrepreneur speculates that certain resources - including labour time - should be acquired in order to create a new market of some kind.

Market opportunities normally have to be realised through intermediation. Intermediation reduces transaction costs which would otherwise inhibit trade. It is also a profit-extraction mechanism for the entrepreneur. The organisation of intermediation has many different aspects, which are explained below; it is these different aspects that are followed up in different theories of the firm. Embedding all these issues in a common framework clarifies the connections between them. It also reveals a common theme - namely the importance of synthesising information and of ensuring that the information that is synthesised is true. It is the synthesis of different kinds of persistent information that brings firms into being, and it is their skill in subsequently synthesising transitory information which, together with the quality of the initial synthesis, governs their subsequent success.

There are two main reasons, it seems, why this general framework has not been presented before. One is that economic theories of the firm have placed too much emphasis on production and too little on market-making. This in turn reflects an excessive concern with material flow in conventional economics, and too little concern with information flow instead. The second is that theories concerned with the formation of firms - such as theories of entrepreneurship - have remained somewhat divorced from theories of the routine operation of firms found in the literature on organisational behaviour and operational research. This has meant that the synthesis of persistent information discussed in the former theoretical tradition has not been integrated with the synthesis of transitory information considered in the latter. While the principle of synthesis is quite explicit in the former, it is only implicit in the latter and so the connection has provided difficult to make.

Defining the Firm

An unnecessary source of confusion in the theory of the firm stems from the lack of consensus on the definition of the firm. Because of the sheer diversity of legal forms that a firm can take, a purely inductive approach to defining the firm is of little use. It is necessary to identify 'firm-like' qualities independently of what any particular real world firm happens to be like. From an economic point of view the most useful way of defining an institution such as the firm is in terms of the function it performs. 'The firm is what the firm does' in other words. But what exactly does the firm do?

According to Coase (1937), the firm is a co-ordinator. The market is a co-ordinator too, but unlike the market, where responsibility is dispersed through negotiations, the firm concentrates this responsibility upon itself. The firm may therefore be defined as an institution which specialises in co-ordination using a single locus of responsibility. Firms exist as 'islands of conscious power' in an 'ocean of unconscious cooperation' because, for some sets of activities, this approach to coordination is superior to others (Robertson 1923, p. 85).

This definition needs to be amplified, however. The essence of co-ordination is decision-making. The firm is therefore a specialised decision-making unit. Specialisation is effected when the firm takes decisions about resources which it does not itself supply and about resources which it does not itself consume. Unlike an individual worker, the firm does not decide how to allocate its own labour time; rather it decides how to allocate labour time purchased from others. In the same way the firm, unlike an individual consumer, does not decide how best to consume certain goods, but how best to allocate the goods it owns between the consumers who would like to purchase them.

But what is the point in an entity that does not work itself acquiring the rights to labour, and an entity that does not consume itself acquiring products whose value ultimately derives from consumption by other people? The answer lies in the improvement in the quality of the decisions that results when the firm rather than the worker decides how labour time shall be allocated and the firm rather than the consumer decides on the ultimate sources from which goods shall be procured.

But how does this improvement in the quality of decisions occur? It arises because the firm has information at its disposal that other people lack. Even if it does not have the information immediately to hand, it has the capacity to acquire this information when needed at lower cost than other people.

A methodological problem arises here, however, because decisions are actually taken by people and not by impersonal entities such as firms. What the firm can do, however, is to structure the activities of the different people who participate in the decision-making process, so that their individual contributions to the decision-making process are made in the most effective way. The firm, in other words, is essentially a structure designed to harmonise the decision-making efforts of a group of people who are focused on a single issue or set of related issues.

Where an individual is confronted with a one-off decision it may not be worthwhile to transfer the responsibility for the decision to a firm. The set up cost that is involved may outweigh any benefit to the individual from the support of the firm. Where recurrent decisions are required, however, the value of the prospective stream of benefits may well outweigh the set up cost. It becomes economic to vest decision making responsibilities in a firm. Because the set up cost requires time to pay back, continuity of operation is one of the hallmarks of the firm.

The firm is more than just a structure, however. From a legal point of view the firm is a kind of fictional person which can acquire, hold and dispose of property in its own right. This legal fiction is a rational institutional response to the function of the firm. When the firm needs to acquire command of labour services in order to control their use it can do so by purchasing labour time. Likewise, when the firm needs to dispose of the product on which labour has done its work it can do so by selling it. In this way it generates revenue which can be used to cover the cost of the labour it hires.

But is this fiction really necessary, it may be asked. If the decisions are taken by a single individual, for example, can this individual not buy labour services and sell the product on his own account? Why does he need to operate inside the 'legal shell' of the firm? The answer lies in three legal privileges normally conferred by the corporate form, namely limited liability, indefinite life, and the right to set off purchases of inputs against revenues from output when assessing liability for tax. The advantages of limited liability to a risk-averse individual are fairly obvious, although there are disadvantages too: shifting risk to creditors could conceivably damage the financial reputation of the firm. Since limited liability is only an option, however, it is available to those who wish to take advantage of it without detriment to those who do not. The unlimited life of the firm allows its contractual rights and obligations to survive the death of its owner and so permits the structure to be perpetuated by his heirs or his trustees. So far as taxation is concerned, most tax systems aim to tax consumption. If an individual employed a worker in a personal capacity, the expenditure could be interpreted as a form of consumption. By employing the worker through a firm the employer makes clear that the expenditure must be set off against revenue so that only the profit is liable for tax.

In the light of this discussion, a firm may be defined as a specialised decision-making unit, whose function is to improve coordination by structuring information flow, and which is normally endowed with legal privileges, including indefinite life. There are, of course, many other common characteristics of firms, but these are best regarded as following from the implications of this definition rather than being elements of the definition itself.

The Firm as a User of Information

A person who takes a decision must be motivated to collect the information that is needed, and this means they must bear some of the consequences of the decision. Thus the right to take decisions about a resource is normally ascribed to the owner of the resource. Thus people who have the relevant information already, or can collect it more cheaply than others, have an advantage as owners of a resource. They can afford to bid away the resource from others who lack this information or would find it more costly to collect. The advantage conferred on the firm by its structuring of information flow allows it to out-bid ordinary individuals for ownership of certain types of resource, and that is how these resources come to be within the control of the firm.

But why should certain people have better access to information than others? Information has the property of a public good, in the sense that supplying information to someone else does not reduce the supplier's access to it. Information is easy to share, in other words. In this case, why does everyone not have access to the same information?

The sharing of information is restricted in two ways. The first is by the cost of communication. When sources of information are localised and costs of communication are high, those who are closest to these sources can obtain information more cheaply than others (Hayek 1937). The second is a contractual problem. Sources of information may be costly to discover. In a private enterprise economy people can only specialise in the discovery of information if they can generate an income from this information. Selling information is extremely difficult, however. Transactions costs are very high for the kinds of reasons discussed later on. This obliges those who discover information to exploit it for themselves (Casson 1982).

But just as information incurs costs of communication, so ordinary resources incur transport costs when they are moved from the custody of one owner to the custody of another. Ordinary resources encounter contractual problems too. So why transfer the ownership of a resource to the person who has the relevant information rather than transfer the relevant information to the person who has the resource? In other words, why is the acquisition of resources explained by the distribution of information amongst prospective owners rather than the acquisition of information explained by the distribution of resource ownership instead?

The answer is that information is more costly to trade than most other resources. Resource ownership therefore moves to the information source, rather than the other way round. This is a special case of the more general proposition that ownership of resources is acquired by people who have a complementary non-tradable resource. This non-tradable resource is normally information-based; if it is not pure information then it is usually a related resource such as technological know-how.

In the short run this complementary resource may appear as a competitive advantage (Porter 1980) or absolute advantage (Hymer 1960, Dunning 1977) possessed by the owner of the tradeable resource. In the long run, however, the advantage possessed by the owner is best construed as a comparative advantage instead. Sustained competitive advantage or absolute advantage is achieved by investments of particular kinds. Anyone can undertake investments of this kind, but in the long run only those with a comparative advantage in making such investments will find that it pays them to do so. In the long run, therefore, the ownership of resources is acquired by people who have a comparative advantage in investing in non-tradable resources. The prime example of such people are those who have a comparative advantage in collecting and processing information. These people will usually exploit their comparative advantage through the institutional framework of a firm.

Not all information is equally costly to trade, however. Communication costs are greatest for information of a tacit nature (Polanyi 1964, Winter 1988). Contractual problems are greatest for information that is difficult to patent and whose quality is difficult to assess (Buckley and Casson 1976). Sustainable competitive advantage therefore rests on a comparative advantage in handling tacit information of uncertain quality which is difficult to patent.

A comparative advantage in scientific research satisfies some of these conditions, but not necessarily all of them: some kinds of technology can be patented, and thus their information content can in principle be sold to those who already possess the resources required for their exploitation. Information about opportunities for trade, on the other hand, normally satisfies all the conditions for non-tradability. Opportunities for trade hinge on consumers' preferences and producers' opportunity costs. Information on such subjects is usually impressionistic, and therefore of a tacit nature. Unlike scientific inventions, such information cannot be patented, and because objective evidence to support it is difficult to obtain, its quality is uncertain. There is, therefore, no way in which such information can be adequately conveyed to the people who own the resources required for its exploitation; the costs of communication and the threat to appropriability are just too great. The resources required for exploitation must be purchased by the possessor of the information instead. it is, therefore, a capacity for acquiring trade-related information which is pre-eminent as the basis for the long-run comparative advantage of the firm.

Optimism and Competence

Given that information is costly to trade, the exploitation of information is effected by allocating resources in a two stage process. In the first stage the resource is allocated to the appropriate owner and then the owner exercises his power of control to allocate the resource to a particular use. The first stage is based upon trade and the second upon the exercise of control. Trade is a voluntary process which involves the consent of both parties whereas control is a more autocratic process in which the will of the owner prevails.

Not all ownership confers the same degree of control, though. If the resource can be used in only one way then the owner really has no choice to make. Only a versatile resource affords the owner control over how it is to be used. It is typical of firms that the resources they own are very versatile indeed. It is the importance of exploiting this versatility effectively that creates the demand for specialised decision-making that the firm is designed to meet. Labour time is the preeminent example of a versatile resource. Capital equipment is another example: though capital is less versatile than labour, the manager of a capital-intensive firm still has important decisions to make about the scheduling of equipment use. The third main factor of production, land, is versatile too - but only in the long run.

The information used by firms to control their versatile resources is of two main kinds: long-run information on permanent factors and short-run information on transitory factors. Permanent factors are exemplified by the long-run opportunities for trade alluded to the above. Transitory factors are exemplified by the impact of fashion on consumer demand.

Consider, for example, a unique physical asset that can be used to produce different varieties of a given consumer good. The permanent factor governs whether there is a market for the good. The transitory factor governs which variety is in demand in a given period. No individual can observe the permanent factor directly. Individual judgements about the state of the market are highly subjective, it is assumed. Some individuals - the optimists - believe that general conditions are favourable, but others - the pessimists - believe that they are not. The transitory factor can be investigated more objectively, but people differ in their competence. Some people can investigate it more easily than others.

In the course of the negotiations each person signals something of their own valuation of the asset to the other party. This valuation will be high if the individual is an optimist and low if they are a pessimist. It will also tend to be high if they are competent - i.e. they can observe the transitory factor easily - and low if they are not. Trade will proceed, if negotiations permit, when the buyer is more optimistic or more competent than the seller. If the seller is more optimistic or more competent than the buyer then no trade will occur. As a result, ownership is conferred on the more optimistic and/or more competent party.

The transitory factor can be investigated either before or after a bid is made. When the transitory factor is highly volatile it pays to investigate afterwards because this ensures that the information governing the use of the resource is most up to date. If the value of the resource in its best available use is independent of transitory conditions then it also pays to defer investigation, since investigating beforehand does not significantly improve the quality of the bid. A person who is optimistic about the permanent factor also has more incentive to investigate first, since they are more likely than a pessimistic person to finish up acquiring the resource. Thus optimistic people are likely to be better informed than pessimistic people when bidding for the resource. This explains why optimistic people do not make more mistakes on average than pessimistic people do - it pays them to invest more in avoiding them.

It is still possible for optimists to make a mistake, of course. Long run success requires an individual to know when it is right to be optimistic, and when it is appropriate to be pessimistic instead. Because of the highly subjective nature of trade-related information, ownership of resources may be acquired by people who mistakenly think that they have a comparative advantage in handling such information, as well as by people who really do. Those who think that they have a comparative advantage, when they do not, squander resources through over confidence. Conversely, those who have a comparative advantage, but do not realise it, miss out on the profits they could make because they have too little confidence in themselves. It is only those who have a comparative advantage, and have sufficient self-knowledge to be confident of it, who benefit themselves (and society) in the long run.

This analysis can be extended in a straightforward way to include other attributes of the owner, such as their degree of risk aversion. Given the uncertainty that surrounds the permanent factor, risk-averse individuals will be reluctant to become the owners of resources, just as pessimistic or incompetent people are. Risk aversion may not affect owners so much as pessimism or incompetence does, however. It may simply provide them with a greater incentive to investigate the transitory factor before they make a bid. It is quite possible that a risk-averse individual, provided that he is optimistic and competent, may become the owner of a resource because, as a result of his commitment to investigation, he already knows before the negotiations begin that the transitory factor is favourable.

The implications of these results for the firm are seen by identifying the firm with the legal shell used by the owner of the resource who makes the decision on how it is to be used. The analysis identifies the owner of the firm as the optimistic and competent individual with low aversion to risk who acquires resources from other people because he believes that in the light of his better information he can put them to better use. He normally puts them to use by producing goods for sale. If his beliefs about trading opportunities are warranted then his venture in employing the resources will be a success.

The Firm as Employer

The preceding analysis unites three qualities in the owner of the firm. But it may be a tall order for all three qualities to coexist in the same person. Is it possible to organise a division of labour in which different qualities are supplied by different people? It is indeed possible, and the examination of this possibility provides further insight into the nature of the firm. In particular, it elucidates the concept of 'employment' introduced above.

The basic idea is that a pessimistic but competent person may undertake to investigate the transitory factor on behalf of an optimistic but incompetent one. The incompetent optimist acquires the ownership of the resource and then hires the competent pessimist to carry out the investigation. The competent pessimist may even have been the person who sold the resource to the owner in the first place. The relevant resource may even be the pessimist's own labour time. The resulting arrangement resembles the employment relation - though not in every respect. The competent pessimist becomes the 'employee' of an incompetent but optimistic 'employer'. He investigates the transitory factor and is then directed to produce the appropriate product - i.e. he is directed to the appropriate use of his own labour time.

A further division of labour may be effected, however, which separates the manual and cerebral aspects of this work. Two 'employees' may be hired. One is hired to investigate the transitory factor and the other is hired to produce the appropriate output. The former is a pessimist with low information cost while the latter is a pessimist with high information cost. This scenario may be elaborated by supposing that, to simplify communications, the owner tells the investigator what the implications of each observation are for product choice, and leaves the investigator to tell the manual worker what to do. This creates a simple hierarchy in which the owner prescribes a rule for the 'managerial employee' who observes the transitory factor and passes on the appropriate instruction to the 'manual employee'.

This analysis assumes that the investigation of the transitory factor occurs only after the manual worker has been employed. If the investigation is carried out beforehand then a somewhat different picture emerges. The role of the manager is now to advise the employer on his wage bid. More significantly, the employer is now in a position to state what particular use he intends to put the manual labour to.

In many cases the seller of a resource, like the manual worker, does not care to what use his resource is put because his alienation of ownership is complete. This is not generally true of labour time, though. The manual employee may care a great deal about how his time is used - he may very much prefer to produce one kind of product than another, for example. If the 'manual employee' is averse to the risks connected with the job to which he may be assigned then on average it will be cheaper for the owner to recruit labour to a specified job rather than to make a general bid for his labour time. Defining a specific job means that the worker is no longer under the control of the manager in the way he was before (Simon 1957).

The disadvantages of specifying the job in advance through prior observation of the transitory factor have already been alluded to, though. In particular, the volatility of the transitory factor favours carrying out the observation at the last possible moment, after negotiations are complete. The manual worker must therefore already be in place at the time the observation is carried out. This in turn means that the subordination of the manual worker to the manager's decisions is a necessary part of achieving a prompt response to news about the latest conditions.

The same kind of argument applies to a managerial employee who cares about the kind of instructions he is required to issue. Such a manager may wish to know the rules and procedures he must follow before he joins the firm. It may be difficult, though, for the owner to divulge this procedure if it is one that he has invented himself and that rival owners would like to imitate. He may wish to protect it through secrecy. To advertise the procedure as part of the job specification would be most unwise. It is far more prudent to build a requirement of confidentiality into the contract of employment which is already binding on the manager at the time he is told the procedure.

Furthermore, the procedure relating the production decision to the transitory factor may be complicated. Potential managers may decide that, despite their concern about the nature of the procedure, it is simply not worth the effort of assimilating it until they have already agreed to implement it. Because the complexity of the procedure raises the cost of communication, it is uneconomic to evaluate this aspect of the job in advance.

It can be seen that communication costs once again combine with the need for secrecy to create contractual problems. As noted above they created problems in the market for information by making the information itself difficult to divulge. Now they create problems in the labour market by making the procedure that the manager will be required to implement difficult to divulge. The confidentiality about the procedure has implications for the manual employees as well. Although the employee may not know to which job he will be assigned, he needs to form some expectation on the subject in order to evaluate the contract of employment he is offered. As explained in Casson (1995 b), a knowledge of the procedure would allow him to infer, from his own probability estimates of the transitory factor, the probability that he will be assigned to any given job. If he cannot know the procedure, however, then he must consider all the possible procedures that might be applied and attach a subjective probability to each. He can then rework his probability estimates, but the resulting degree of uncertainty about his job will be much greater than before. It is not just the manager, therefore, who is disadvantaged by the costs of divulging the procedure, but the worker too.

Another feature of the employment relation is its long-term nature. Because the costs of negotiation are mainly fixed costs, independent of the value of the contract, it is normally advantageous to substitute a long-term contract for a sequence of short-term ones. Fixed costs are important in other ways too; the manager may have to learn a complex procedure at the outset, while the manual worker may have to invest in habit formation to speed up his repetitive tasks. Screening costs must be taken into account as well (see further below).

A final feature of labour contracts is that they are often of an open-ended nature, allowing either party to terminate them by giving suitable notice. This has the unfortunate effect of allowing the party that terminates the contract unexpectedly to inflict considerable damage on the other. Where the damage would be great, open-endedness may discourage the parties from making appropriate commitments. There will be too little expenditure on non-recoverable set up costs, in other words. It is instructive to note that open-ended contracts are much less common where the hiring of machinery and other non-human resources is concerned. This peculiarity of labour contracts may be best explained, not in terms of efficiency, but as a custom which reflects a moral judgement about the inalienability of certain rights to labour. Further implications of moral judgements about the alienability of labour are considered below.

The Firm as Intermediator

It was suggested above that trade-related information is just as important as technological know-how - indeed, more important, perhaps - as a source of competitive advantage to a firm. While technological know-how is typically exploited through production, trade-related information is most naturally exploited through intermediation instead. Rather than sell the information directly, the possessor of it extracts rents from it by intervening in the market process - by buying goods cheap and selling them dear. Speculation and arbitrage are terms often used to describe such intervention, but these suggest, quite wrongly, that it is usually an intermittent rather than a continuous process. Trade-related information is usually exploited on a systematic basis by organising a market that did not exist before. The intermediator does not intervene in an already existing market so much as set up the market from scratch himself.

Markets need to be specially set up because the conduct of trade is fraught with difficulties. It may be difficult for prospective purchasers to get in touch with existing owners. Discovering the exact specification of what is on offer is another task. Then there is the problem of bluffing in negotiations. The essence of the market process is that each individual keeps his beliefs to himself. People do not share their beliefs, and come to a consensus view. Quite the contrary, indeed. Beliefs are encoded in price quotations and these quotations are intended to mislead other people as much as to inform them. The price quotations made at the outset of the negotiation process are likely to give a highly distorted view of the subjective valuations which underpin them. Even when a deal is eventually agreed, default is always a possibility.

Intermediation can reduce transaction costs of this kind. Contact can be made more easily if a middleman provides well-advertised and conveniently situated retail premises. Specification costs are also reduced if buyers can inspect samples of each product. Most importantly, the process of negotiation can be accelerated by having an intermediator intervene. The skill of the intermediator resides, not in knowing exactly how the buyer will use the good he acquires from the seller, but in identifying the appropriate buyer and assessing the maximum he can afford to pay. Complementing this is a skill in identifying and making contact with the present owners of goods for which ready buyers can be found. The intermediator, in other words, trades on his special knowledge of valuations to reallocate goods to the people who can put them to the best possible use.

The intermediator exploits his own knowledge in the same way that the buyer and seller do - through negotiations. The essence of his strategy is to out-bluff both the buyer and the seller by claiming to the seller that the buyer can afford no more than the intermediator is offering to pay, and persuading the buyer that the seller will not accept less than what the intermediator is asking for it. (The intermediator needs to keep the buyer and seller apart, of course, to do this successfully). By building a reputation for taking a hard line in negotiations the intermediator can effectively discourage haggling. At the same time, by setting realistic prices he can encourage buyers and sellers to regularly channel transactions through him.

A reputable intermediator can also eliminate default by creating a chain of trust in cases where the direct link between buyer and seller is very weak. Because the intermediator is a specialist, making his living by trade, he has an incentive to build up a reputation for integrity. Once he has acquired this reputation, he can require the seller to supply goods in advance of payment in a way that the buyer could not. Similarly he can require the buyer to pay in advance of delivery in a way that the seller could not. In this way he can guard himself against default by the parties lacking reputation, whilst passing on the good between them.

Unlike the producer described in the previous section, the intermediator buys resources for resale rather than for use. His comparative advantage resides not in his knowledge of the use to which the good will be put, but his knowledge of who is the best person to put it to that use. Although the final users know this too, they do not know the sources of supply as well. The intermediator can obtain both the items of information required for coordination more cheaply that other people. He has the optimism that demand will be buoyant for the goods he has bought, and is not so risk-averse, or devoid of confidence, that he fears for the consequences of his judgement being wrong. That is why he becomes the owner of the goods that he re-sells.

The Firm as an Organisation: A Four-factor Theory of the Firm

A distinction was drawn in between the permanent and transitory information used by the market-making firm (see above). It was suggested that the firm typically used just one item of each kind. This is not generally correct: there are at least two items of each type. Both the long-run profitability of the market, and the short run equilibrium of it, depend upon both supply and demand. It is a synthesis of information that the firm must make in each case. Other factors, such as transport costs, are important too, but including these would unduly complicate the analysis.

Permanent information on demand and supply is synthesised on a once-for-all basis. This is the synthesis that underpins the formation of the firm. Transitory information must be synthesised on a recurrent basis, however, Each period the intermediator faces the same problem of how to synthesise information on the demand factor and the supply factor in order to decide what output to order and what prices to quote. Because the problem is always the same from one period to the next it is advantageous for him to devise a procedure which routinely collects and combines the different items of information. Because information is costly to collect, however, it is not always advantageous to commit in advance to observing both demand and supply factors in every period. It is normally advantageous to adopt a sequential procedure in which the most volatile factor is investigated first.

It can also be advantageous to effect a division of labour in the implementation of the procedure. Two managers may be appointed: a marketing manager who observes demand and a purchasing manager who observes supply, each making observations that are a natural byproduct of their other duties. The choice of procedure then governs the distribution of managerial power within the firm.

If demand is more volatile than supply then the procedure will normally begin by investigating demand, and proceed to investigate supply only if the observation on demand is indecisive. This makes the manager who monitors demand more powerful than the manager who monitors supply in the sense that the demand information collected by this manager is used to decide whether to consult the other manager or not. The firm is therefore demand-driven (Casson 1994, Carter 1995). Conversely, if supply is more volatile than demand then the manager who monitors supply becomes the more powerful one, and the firm becomes supply-driven instead.

There are many possible procedures by which information can be synthesised in order to take price and production decisions. These different procedures are analogous to the different techniques that may be used in an ordinary production plant. While the efficient choice of technique in a plant is governed by the available technology and by the relative prices of different material inputs, the efficient choice of procedure in a market-making organisation is governed by the pattern of volatility in the market environment and by the different components of information cost - observation cost, communication cost, memory cost, and so on. Emphasising the market-making role of the firm therefore gives a quite distinctive perspective on the issue of choice of technique.

The Firm as a Producer of Market-making Services: Highly Specific Factors and the Homeostasis of the Firm

The analogy between intermediation and ordinary production may be pushed even further. Just as ordinary production transforms goods physically, so intermediation transforms their ownership instead. Intermediation is effected in a number of stages, just like the ordinary production of a good. Intermediation involves two activities - buying, and then re-selling - with possibly a third activity - such as quality control - in between. What is more, the different stages are complementary to each other. Like the different stages in the production of a good, they operate in fixed proportions.

Intermediation is produced, like ordinary goods are produced, by the utilisation of durable assets such as capital and labour. The role of managerial labour has been alluded to already. Physical capital is exemplified by office equipment - fax machines, photocopiers, computers and filing systems, for example. Given that different stages of intermediation are complementary, the main threat to coordination comes from breakdowns - mechanical breakages, power failures, illnesses and so on.

Breakdowns threaten the homeostasis of the firm. The incidence of a breakdown is a transitory factor - but it is a highly specific transitory factor relative to the other transitory factors considered above. These other factors were specific only to the market for a particular type of product, whilst a breakdown factor is normally specific to a particular asset. A firm which employs a complex of assets therefore faces a large set of different breakdown factors. The problems created by breakdowns can be controlled in two main ways. The first is by preventing breakdowns, or ensuring that when they happen it is not the firm's responsibility to put them right. The second is that when they do happen they are dealt with promptly and that 'knock on' disruption to other activities is minimised.

Every asset has certain hidden characteristics which govern whether it is prone to breakdown or not. Some of these qualities may be revealed through a screening process. Intermediators who have a comparative advantage in screening such assets will normally prefer to purchase them outright, whilst those who have a comparative disadvantage in screening may prefer to purchase the specific services of these assets instead. Thus a technologically sophisticated firm may prefer to own its equipment outright (and even buy it secondhand, which requires considerable skill), a less sophisticated firm may prefer to rent equipment on rolling short-term contracts, whilst a completely unsophisticated firm may prefer to purchase specific services under a contract which makes the supplier responsible when breakdowns occur. This is another example of the equilibrium described above. Ownership of the asset, and the attendant responsibility for breakdowns, is vested in the person most competent to assess the risk, and most confident that the risk is low.

Where the asset concerned is labour, further complications arise. Because labour is inalienable (see above), labour cannot be bought and resold like other commodities. Intermediation cannot occur in the labour market in the same way that it does in product markets (Spence 1973). Speculators intermediate in capital and land; in the labour market, alone among the factor markets, intermediation can only be effected on a fee basis. Fee-charging intermediators, such as employment agencies, cannot be trusted to the same extent as normal intermediators because they do not directly bear the risks of their mistaken decisions. Transaction costs are therefore higher in the labour market than they would otherwise be. This discourages market-making firms from adjusting the composition of their labour force in response to shocks, and encourages them to adjust their use of other factors (notably capital) instead. Transaction costs are particularly high where skilled labour is concerned, since accurate screening is crucial in this case. For this reason many firms treat skilled labour (including managers) as a fixed factor of production so far as transitory fluctuations are concerned.

The second arm of strategy is to optimise the response to breakdowns. Sophisticated procedures are required. But the number of different combinations in which breakdowns can simultaneously occur makes it uneconomic to pre-plan the optimal response in every conceivable case.

To economise on information costs, slack may be deliberately built into the system. Holding precautionary inventories and maintaining spare capacity are obvious strategies for promoting flexibility. For example, ageing office machinery with high operating costs may be 'mothballed' for emergency use. Labour contracts may be written to allow the employer to insist on people working overtime if a crisis develops. This facilitates substituting labour for capital in the event of a mechanical failure.

Another opportunity for flexibility arises in the deployment of managerial labour between different jobs. This is a subject of considerable contemporary interest in view of recent Japanese successes in flexible production management (Geanakopoulos and Milgrom 1991). Consider, for example, two employees on the same premises, each of whom has both a specialist skill and a general skill. The specialist skill is required by the activity to which the employee is allocated on a regular basis. The general skill relates to helping out the other employee when a breakdown occurs. A breakdown creates a crisis situation for the specialist, which calls for an immediate response from the other employee. The need for rapid communication suggests that 'horizontal' communication from one employee to another may be more effective than 'vertical' communication - i.e. communication intermediated by another manager (Aoki 1986). Moreover the difficulty of guaranteeing a pre-planned response requires good understanding between the employees themselves. It requires the employer to invest in employees who have general as well as specific skills and are willing to use these skills in order to help each other out. It also suggests that the employees should be drawn from a culturally homogenous group, or that some process of assimilation should be organised, in order to reduce the costs of communication between them (see further below). In other words, an intermediating firm works best when it functions more as a family or as a social unit than purely as a collection of individual people.

The Firm as the Producer of the Product It Sells: The Vertical Integration of Marketing and Production

The advantage of the intermediator is that he has a wider vision of the situation than the buyer and seller with whom he trades. If the intermediator finds that the same buyers and sellers are dealing with him on a recurrent basis then it may be advantageous for him to acquire their resources and to become the permanent owner of them. Thus the buyers and sellers cease to be owners in their own right and become his employees. As his employees they no longer encode their information in the form of decisions whether or not to trade. They share their information directly with each other. The intermediator structures their information flow in order to economise on the cost of communication. He replaces negotiations with a system for synthesising information on a more centralised basis. The pure intermediator expands his activities into those of a fully integrated firm.

A typical intermediator deals with fewer sources of supply than of demand. Sources of supply also tend to be more homogenous and more geographically concentrated than sources of demand - customers are often widely distributed and use a good in different ways. Because of the more limited diversity of supply, backward integration into production is often more practical than forward integration into demand.

The theory of vertical integration is, of course, already well established, but it is most commonly applied to successive stages of ordinary production (see for example Bernhardt (1977) and Carlton (1979)). Introducing market-making into the picture considerably strengthens the theory's relevance. The integration of production and market-making explains a prominent feature of many firms, namely the coexistence of a strong marketing department and a strong production department within the same firm. Internal conflicts between these two departments are a natural consequence of the internalisation of the market for wholesale supply, which locks the two departments into each other.

Conversely, the possibility of disintegrating such a firm through the subcontracting of production sheds considerable light on recent debates over the 'hollow firm'. More specifically, it shows that the hollowing out of a firm typically returns it from an integrated form to a pure intermediating role. It is only the mistaken belief that it is ordinary production rather than market-making that is crucial to the firm that leads people to perceive this as a paradoxical development.

What then governs the integration and disintegration decisions of the market-making firm? The structure of the wholesale market is one factor. If the market for supplies is basically competitive then an intermediator who is dissatisfied can always switch to a different producer. If switching is difficult, however, then the producer may realise that he enjoys a monopolistic position, and so the supply price may be set above the marginal cost of production. As a result the intermediator may wish to take steps to reduce the supplier's market power (Waterson 1982).

Acquiring the seller is one approach, but this has the disadvantage that the monopoly rents accuring to the seller will be capitalised in the purchase price. In so far as the exercise of monopoly power is distorting allocation decisions at the margin, however, there will be efficiency savings even if the supplier is acquired at an 'inflated' price. The exercise of monopoly power by the seller only causes distortions of this kind, though, when the seller cannot extract his monopoly rents through price discrimination. Where the same seller is supplying a number of different customers, and resale between them is difficult to control because of information costs or legal constraints, then discrimination is difficult and distortion does indeed occur. Under these circumstances, therefore, the acquisition of a monopolistic producer may be effected purely to improve the co-ordination between production and final sale.

An alternative to acquisition is for the intermediator to build his own production plant in competition with his existing supplier. The disadvantage of this is that it may add unnecessarily to capacity in the industry. In any case, if extra capacity really is needed then it may be questioned why others have not entered the industry to build additional capacity already. One explanation may be that producers are less optimistic than the intermediator about the prospects for demand (for the reasons given above), and so are reluctant to invest. In this case the intermediator may invest as a speculation that his forecast of demand is better than theirs. If extra capacity is built then competitive forces will encourage the existing producer to cut his prices, which means that the intermediator may still wish to patronise that producer even though he now owns his own plant as well. Indeed if he delegates the management of his plant to people he does not really trust then he may find it advantageous to threaten them with switching production to an outsider unless they can produce more cheaply themselves.

There are three main reasons why, notwithstanding this, the intermediator may wish not only to invest in a plant of his own but also to rely on it exclusively for his supplies. Though logically distinct, they are related because they all apply with greatest force to the supply of nearly innovated products. This makes the impact of the separate factors difficult to distinguish in practice, since they tend to appear jointly where innovative products are concerned.

The first concerns quality control (Casson 1987, Chapter 4). The intermediator may be unsure whether he can trust the producer to match the specification. He may not believe in the supplier's integrity, as noted above. But even if he trusts the supplier's integrity, he may not trust in his competence. With a new product whose production requires a modification of existing processes, the supplier may just not be up to the job.

To assure quality, the intermediator could, of course, insist on supervising the production operation. An independent producer is likely to object to this, however, because the information obtained as a result of supervision is likely to come as a package which contains some items of a confidential nature. A by-product of observing the production process may be a fairly accurate assessment of the production costs, for example. Use of this information in subsequent negotiations could improve the intermediator's bargaining power to the detriment of the producer. Again, the producer may have been selected for a special skill he has, which he needs to protect as a trade secret. Allowing supervision would divulge the secret to the intermediator and further encourage him to set up rival production of his own.

The second factor is sunk costs. The intermediator may require the producer to invest in specific equipment, or incur other forms of set up cost, in order to customise the product to his requirements (Klein, Crawford and Alchian 1978). Because of legal shortcomings, the producer cannot sell forward sufficient supplies to cover his costs before they are sunk. His consequent reluctance to customise the product obliges the intermediator to customise it for himself.

The third factor concerns the intermediator's desire to fully appropriate the value of his profit opportunity. In subcontracting the production of a new design the intermediator risks building up a competitor. Unless the design is patented, the intermediator's only method of excluding competitors is through secrecy. Even with a patent, the subcontractor may use his experience of production to improve the design and thereby render the original patent obsolete. For all these reasons, therefore, an intermediator launching an innovative product is likely to integrate backwards into its production.

Marketing, Technology Transfer and Horizontal Integration

The integration of production and marketing need not be confined to a single market. The intermediator may have an idea so general that it has much wider implications than this. The design for the product may, for example, be relevant not to just a single market but to a whole set of markets. The limiting case involves a marketing concept of global relevance.

The exploitation of a general idea leads to a distinctive pattern of horizontal integration. Because the global market is spatially segmented by transport costs and tariffs, and by non-tariff barriers such as national safety standards, intermediation has to be replicated in different locations. In each town there may be a shop, in each country a warehouse and in each trading bloc or supranational region a production facility. Within this framework, substitution possibilities tend to emerge. While each production facility normally sources its nearest markets, it is also possible for it to top up more distant markets if shortages develop there. Even within its own market area the firm can switch consignments between warehouses, and warehouses can switch consignments from one shop to another.

These internal substitution possibilities have important implications for the organisation of the firm. The logic of synthesising information is driven by the need to source each retail unit from the minimum cost location. Under these circumstances it may be advantageous for each production unit, and each warehouse, to encode its report on local supply conditions as a price quotation. Price information is more explicit than most other forms of information, and is therefore cheaper to communicate; moreover it feeds naturally into an algorithm geared to identifying the cheapest source of supply. Internal markets with substitution possibilities may therefore be co-ordinated using transfer prices. Where complementarities continue to dominate, though, conventional procedures based on budgets matched to target quantities are likely to be used instead. Internal substitution possibilities are characteristic of large multinational enterprises, and it is in firms of this type that internal prices are most likely to be used.

Some of the most powerful general ideas relate not merely to market-making, but to the commercial application of technology as well. Schumpeterian innovation, for example, combines market-making intelligence with an assessment of the industrial potential of new technology (Schumpeter 1934, 1939). Indeed, because technology is an internationally transferable public good, ideas for technological innovation are naturally global in a way that many ordinary market-making ideas are not. Technology transfer is thus an important aspect of the international operations of many firms.

Just as the production of a good, and the making of a market, involve a number of stages, so too does the development of a technology. Besides the 'value-chain' encompassing production and the distribution channel, there is the 'value chain' of research and development (R & D) as well. This involves, for example, basic research in the laboratory, scaling up to pilot production, and the replication of plants with suitable adaptations to the local environment.

The issue of vertical integration arises in connection with this R & D value chain as well. The importance of technology is such that there is a strong case for effecting a division of responsibility between its discovery and its exploitation. The natural way of effecting this separation is to create a market in which patented technology is licensed out for use. If licences were priced appropriately, and agreements were costless to enforce, then charging for use would not inhibit diffusion in any socially harmful way. This is because no-one would be asked to pay more than what they valued the information at, provided that this covered the cost of communicating it to them. If the licensing approach were generally applied, it would create two distinct types of firm: one that specialised in R & D, and another that specialised in exploiting the knowledge so produced (Buckley and Casson 1976). The first type of firm would typically license a given idea to several different firms of the second type, each of which would make a market in a particular location.

Apart from the difficulties of licensing, noted earlier, there is a serious objection to this approach, however. Different licensees could easily dissipate the rents from the idea by 'dumping' in each other's markets. Export restrictions in the licence agreement can attempt to control this, but the most efficient method of rent extraction is to organise the licensees as a cartel. Covert price-cutting then becomes a major problem, however. If the licensees do not believe that this problem can be solved then they will reduce their bids for the licence. The sum total of the bids then becomes much lower than the maximum rent available and so, even allowing for his comparative disadvantage in exploitation, it pays the owner to integrate forward into the exploitation of the idea himself.

Obstacles to licensing thus encourage the internalisation of the market for technology. This, in turn, creates further complexities for internal co-ordination. Besides the flow of materials from production through to distribution, there is now the flow of technical information that needs to be co-ordinated too. Experience from production and from customer use must be fed back to the research laboratories, while improvements effected in these laboratories must be fed back to production and explained to the customers. Additional flows of co-ordinating information are therefore required to co-ordinate the flows of technical information with each other and with material flow. This obviously adds further complications to the organisational procedures of the firm.

The Firm as a Social and Cultural Unit

The scope of the firm is ultimately constrained only by the scope of the owner's vision. Even if the owner lacks knowledge of crucial details required for implementation, he may be able to hire the expertise necessary to make good his own deficiencies. Multi-stage vertical integration may emerge if the owner has a vision of how an entire production sequence could be rationalised, with intermediate product flows being restructured and plants expanded or shut down. Conglomerate diversification may result from the owner having a vision of a new kind of lifestyle that consumers wish to pursue, involving the purchase of a complementary collection of novel household goods. In principle, horizontal, vertical and conglomerate diversification can all occur together in the same firm.

Actual instances of this are fairly rare, however. The practical limit on the scope of the firm's operations stems from the increasing problems the owner encounters in recruiting staff with the appropriate competencies and in making sure they behave in an honest way. Skills in screening people for senior management posts are indispensable in the owner of a large firm. Indeed, it may be said that in managing a very large firm it is the skill of recognising other people who have the skill to recruit their own subordinates that is the most vital thing (Knight 1921).

The larger is the firm the more complicated are likely to be the procedures that it uses. Wide-ranging integration means that many specific sources of environmental volatility impinge on the firm's operations. Managers must synthesise information from increasingly diverse sources in order to support decision-making. Although every member of the organisation has some interest in its overall performance, the larger the organisation becomes, the more remote the connection between overall performance and individual reward tends to be. Increasing complexity creates more and more opportunities for managers to pervert procedures in order to promote their personal interests. (This 'agency problem' has become an obsession with some modern writers on the theory of the firm; see Milgrom and Roberts 1992). With a fixed span of control, monitoring procedures involve an increasing number of administrative layers, and motivating people to monitor each other attentively becomes more difficult as a result.

The basic problem is that the decision-making procedures used in any organisation are premised on people telling the truth. In a highly centralised organisation this means quite simply that everyone reports their observations of transitory factors honestly to their superiors. In a decentralised organisation, where different people are given their own particular rules to apply to the data they collect, it means that what they decide is based on applying the rule they have been given to the data as actually collected. The decision rules only make sense if they are correctly implemented and applied to honest data.

It is sometimes possible to check up on people after the event to see if their report was honest. But such monitoring procedures depend in turn on the supervisors following the procedures that they have been prescribed. 'Who will supervise the supervisors?' encapsulates the dilemma of the large firm. The answer is that only those who are really trusted can be permitted to supervise without being supervised themselves.

This points to an alternative approach to eliciting truth, which is to make everyone trustworthy as far as possible. Instead of having a small core of trustworthy people doing little else but supervising untrustworthy people, the untrustworthy elements are eliminated by making everyone trustworthy instead. The trustworthy people can then concentrate on taking decisions, since there are no untrustworthy people left for them to supervise.

But how can this desirable state of affairs be engineered? Trustworthy people, from the owner's point of view, are essentially self-monitoring. They check up on themselves and punish themselves with a bad conscience if they default (Casson 1991). They are motivated to check up on themselves because they believe in the objectives of the firm. These objectives enjoy a moral legitimacy which the pursuit of their own narrow material interests lacks. If the corporate objective is purely to maximise the owner's profit, however, then it is difficult to see why they should subscribe to it on ethical grounds. But if the maximisation of profit is represented as something instrumental to the pursuit of a morally higher goal then individual self-interest may well be suppressed. More precisely, when self-interest is construed more broadly in terms of the satisfaction of helping to achieve this higher goal, material self-interest carries much smaller weight.

Engineering beliefs of this kind is difficult, however. Indeed, it is difficult to sustain such beliefs in the long run unless they really are true. How can a rational self-interested owner, who really wants to maximise profit, cope with the dilemma that he can only maximise profit by wanting to maximise something else instead? The answer is to appoint someone as a symbolic leader of the firm who really does want to maximise something else. It may be cheaper to allow this person to extract certain rents from the firm for the cause in which he believes than to do without this person and have numerous employees extract rents for themselves instead. There is, therefore, scope for a contract between an owner and a moral leader, allowing both to pursue the objectives - material and moral respectively - in which they believe (Casson 1995 a, Chapter 6).

The Foundation of the Firm

Judgements about market prospects are essentially subjective. It was noted above that opinions differ about the scope for profitable intermediation. Some people take an optimistic view and others a pessimistic one. This diversity of opinion about the permanent factors may be explained by the fact that the underlying market situation cannot be observed directly, but only through certain symptoms. Some people interpret the situation in terms of one symptom and others in terms of another. Each symptom is correlated with the true situation. Accurate symptoms are highly correlated with the true situation and the most successful owners use these. Others are more weakly correlated. Some symptoms may not be used at all because people realise that they are, on balance, too misleading.

Some symptoms may have a bias to optimism in the sense that while picking out good conditions very accurately they are quite likely to report bad conditions as good as well. Others may have a bias to pessimism in that they pick up bad conditions very accurately but are likely to report good conditions as bad too. Wrongly accepting good conditions as bad is less hazardous than wrongly accepting bad conditions as good, and therefore risk-averse people are likely to avoid using symptoms which are biased to optimism. Although the distinction between pessimism and risk aversion made earlier remains valid, it is also true that when people have a choice between alternative symptoms, risk-aversion is likely to lead to pessimism because of symptom choice.

In principle, of course, people could reduce risks still further by combining information on different symptoms. The principle of synthesis enunciated earlier in connection with transitory factors, would then apply to different symptoms of the permanent factors too. The subjectivity of opinion about permanent factors is almost certainly explained, however, by the fact that this cannot be easily done. Symptoms may be identified as a highly tacit form of information, each obtained from a quite distinctive source. No-one who has not consulted a particular source directly can fully understand what the symptom shows, and the sources are too diverse for anyone to consult more than one symptom.

It is possible, for example, that the identification and interpretation of symptoms operates at a subconscious level. Each individual uses symptoms which their accumulated experience has taught them is useful in a particular type of situation. They may not know why they use this symptom or even be aware that they do so. The use of symptoms may be a highly pragmatic affair.

People can choose which symptoms to use only in the sense that by choosing a particular occupation and lifestyle they determine what kind of experiences they accumulate. Sophisticated symptoms are likely to be the result of a broad and varied lifetime experience. Because the use of symptoms is subconscious, people cannot share their expertise in any formal way. It is possible, though, to exchange life experiences through conversation, and quite a bit of business wisdom may be transmitted in this way. People may also find it easier to share their expertise when confronted with a specific situation. Each person examines particular symptoms in the situation and gives their opinion on the matter. Because these opinions may be difficult to justify by explicit reasoned argument, however, the weight that other people attach to a given person's opinion may depend more on their reputation and general social standing than on their ability to justify their coice.

The accuracy of the symptoms used is not the ultimate determinant of successful intermediation, however. If two or more people use similar symptoms to arrive at the same conclusions then rents from the exploitation of the general idea will be dissipated through competition between them. The loss is substantial because the costs of discovery are sunk coss and the marginal costs of exploitation are relatively low.

The combination of symptoms used to identify an idea therefore needs to be unique. This in turn suggests that the individual or group that develops the idea needs to be unusual - even eccentric - in some particular way. This makes them alert to opportunities which others simply fail to discover (Kirzner 1973). Alternatively, it makes them aware that possibilities rejected by others really are opportunities, despite the fact that the others believe that they are foolish to go ahead with exploiting them. They are therefore able to preempt the opportunity and to invest in barriers to entry which will keep competitors out once they realise their mistake. Few barriers succeed in the long run, of course, but they may still be useful in prolonging the otherwise transitory period for which monopoly rents can be earned.

If an individual believes that he has recognised a change that others have not, and realises that this is because of a special symptom that he has used, then he may well ask himself whether he may not be in the wrong and others in the right. He requires self-confidence to back his own judgement. Without self-confidence, moreover, he will probably be unable to get financial backing from others, as discussed below. Whether the individual is successful or not depends upon whether his confidence is well-placed (as noted above). People with good self-knowledge, who can assess their own competencies successfully, are more likely to achieve success because their degree of confidence will be matched to their true capability. People who lack confidence will never found a firm, while those who are over-confident are likely to found firms that fail.

Maintaining Flexibility: the Firm as an Entrepreneurial Unit

The opportunity for intermediation has hitherto been described as governed by permanent factors, but this is not strictly correct. The factors are certainly more durable than the transitory ones, but they are better described as persistent rather than permanent, as indicated at the outset. Opportunities for intermediation are certainly liable to change them from time to time. This raises two issues, both of which are important for the nature of the firm.

The first is whether innovation in market-making is better effected by new entrants or established firms. Should the owner of an established firm continue to monitor the persistent factors by regularly updating his estimate of the gains from intermediation, or is it more efficient for him to leave this to others instead? If he leaves it to others then it is possible that his firm may not survive any change in long-run market conditions. Competitors who are more alert than the owner will emerge when a change occurs, and if the owner fails to recognise the competitive threat then the firm may fail. It may still be worth bearing this risk, to economise on information costs, if the owner considers that the market situation is basically very stable. Alternatively, the owner may believe that he has the flexibility to respond rapidly to any threat. In this case, the firm may be able to survive by free-riding on the information signalled to it by the behaviour of potential competitors. It relies on its established position to 'see off' any entrants.

If an established firm decides to retain an innovative capability of this kind then a second issue arises. This is whether the monitoring of market-making opportunities should be governed by a procedure. The handling of information on transitory factors is normally procedure-driven, as noted earlier. Should the handling of information on persistent factors be procedure-driven too?

Changes in the transitory factors are liable to occur each period. Information is frequently updated, and the costs of developing the relevant procedure are consequently spread over many successive applications. Moreover, the range of variation in conditions may be fairly narrow. So far as the persistent factors are concerned, however, change is more intermittent and the optimal frequency of observation is therefore low. The range of possible situations that can emerge is very wide and hence the probability that any one of them will ever occur - at least within the owner's lifetime - is very small. This argues against planning in advance how to respond to every conceivable change. Plans that are developed may never need to be implemented, and even if they are eventually implemented the cost of remembering them until they are required may be quite considerable. It is therefore more efficient, in terms of information costs, to wait till a particular situation reveals itself and then improvise a response in the light of the symptoms observed (Casson 1996).

A change in the nature of the market-making opportunity implies a change in the procedures by which transitory information is synthesised. Indeed, if no change in these procedures is required then there may be little point in identifying the change in the first place. The implication is that changes in procedures are normally improvised - such changes are not themselves procedurally driven.

The responsibility for improvising procedural change may be set apart and allocated to a particular individual - typically the owner of the firm, but possibly a trusted employee instead. This responsibility for improvised decisions is the defining characteristic of the entrepreneur (Casson 1982). The exercise of entrepreneurship reflects a fundamental division of labour within the firm between routine procedure-driven decision-making on the one hand and the improvisation of changes in procedures on the other.

This division of labour is not absolute, of course. Because changes in market-making opportunities are only intermittent, the act of improvisation is intermittent too. Although symptoms of the persistent factor need to be regularly monitored, this may not be sufficient to keep the entrepreneur fully occupied. It may therefore be advantageous for the entrepreneur to fill out his spare time by helping with routine work. This explains why entrepreneurship does not always correspond to a specific full-time job within the firm.

The kind of improvised response that is made to changes in the persistent factor is analogous to the decision made about the foundation of the firm. This is based on preempting the use of information about the persistent factor - information which would almost certainly have been exploited by someone else if the entrepreneur had not responded first. The same kind of entrepreneurship that is required to found the firm is also required to give it the continuing flexibility needed to survive. The firm is effectively 're-founded' each time its procedures change in response to a change in fundamental change in market conditions recognised by the entrepreneur.

The Firm as a Political Institution

Given that the entrepreneur reserves the right to change the procedures of the firm at any time, it is open to employees to object that they did not anticipate such changes when they joined the firm. Notwithstanding the formal legal power that the owner of a firm possesses, employees might claim that he had reneged on a implicit component of the employment contract. It might be alleged that the procedures had been changed in a wholly unreasonable way.

It is, of course, possible to dismiss the entire problem on the grounds that only the explicit part of the contract of employment counts. But implicit contracts can often be useful in economising on transaction costs, and cannot be dismissed altogether. It could be argued, however, that implicit contracts cannot emerge simply from custom, and that an employer who follows the same procedure for a long time does not morally validate employees' expectations that it will continue for ever.

One way of clarifying the issue for employees is to frame a constitution for the firm which sets out the range of permissible procedures. Given the enormous breadth of the range, it may be easiest to specify this set by exclusion - i.e. to specify what is not permitted rather than what is. Another possibility is to give the employee a right to be consulted over any change, or even to allow him to participate formally in the decision by giving him a vote that he can exercise along with other recognised 'stakeholders' in the firm. This is somewhat akin to a worker-controlled firm, or at least to a firm where management is overseen by a worker-controlled supervisory board. In this case the entrepreneur comes to resemble a political leader within his own firm, persuading employees to vote in favour of his recommended change of procedure.

The amount of protection afforded to the employee under such arrangements depends upon the nature of the voting system - thus majority rule affords the employee less protection than does a consensus rule, which confers a right of veto on any employee who believes he would lose by the change. The consensual approach has the desirable property - in theory at least - that it obliges the employer to compensate all losers and so ensures that the new procedure is a Pareto-improvement on the old one. It has the practical disadvantage, though, that it allows any one employee to hold up the change in return for extortionate compensation, and may precipitate the growth of factions, each determined to get a bigger payoff for its members than the others do.

A more satisfactory solution from the standpoint of the entrepreneur is for him to constrain himself by the application of certain ethical principles. If these principles command the assent of the employees, and they trust the entrepreneur to act by them, then formal consultation may prove unnecessary.

It is obvious that the arbitrary exercise of power is an issue which arises not only in connection with the governance of the firm but also in conjunction with the state. There is, however, a difference between these two cases, in the sense that employees have more freedom to move between firms than citizens do between states. Competition between employers, even though it may be somewhat restricted, is normally sufficiently intense that those made discontented by a change in procedures can find alternative employment with another firm. In Hirschman's (1958) terms, because the 'exit option' is more attractive there is less need for the firm to invest in mechanisms for 'voice'.

Where key employees with job-specific skills are concerned, however, the exit option may be costly to the firm as well as to the employee. Costs sunk in training employees for their specific roles may be lost for good if discontented employees quit their jobs. The implicit contract between employer and employee is normally a reciprocal one in which the firm receives loyalty and hard work from the employee in return for steady career progression. An entrepreneur who instigates an adverse change in procedures without full agreement therefore risks undermining the loyalty and commitment of his employees.

In the light of all this, it may be asked why firms do not equip themselves with a formal constitution regulating changes of procedure. The answer is, quite simply, that the constitution would be costly to specify, and probably too costly to explain to employees as well. The probability of a change being required that, because of the absence of a constitutional process, will alienate key employees is quite small. From the employees' perspective, the opportunities for changing jobs are normally sufficiently good that refusing to work for a firm which lacks a formal constitution is hardly worthwhile. Nevertheless, the costs and benefits of a constitution are very different when the constitution is standardised across all firms. The fixed costs of specification and assimilation are then spread more widely. Moreover, if established firms become more entrepreneurial - in response, for example, to the globalisation of competition - then procedural changes may occur more frequently within any given firm. Under these circumstances it may be efficient for nation states to establish a standardised corporate constitution as a response to the greater volatility now characteristic of the world economic environment.

The Firm as a Partnership of Risk-bearers

It may seem obvious that the owner of a firm is a risk-bearer, but this is not inevitably the case. It is possible to imagine a firm which buys up all its supplies in advance and sells all its output forward too. It owns no resources, except those it can fully screen, but hires specific services instead. Flexibility is achieved by making forward contracts contingent on all uncertain events, or by negotiating separate insurance to cover these possibilities. Revenues and costs are carefully matched under each possible situation. It might, it is true, be advantageous to sink costs before negotiations are complete in order to discourage competitive entry into the market, but contingent contracts can, in principle, be extended to deal with entry threats as well, provided potential entrants can be identified in advance.

This scenario is not so far-fetched as it may seem. Many subcontractors do indeed supply their product to an intermediator under long-term arrangements, as indicated earlier. Furthermore, some sales, such as those in the defence industry, we made on a cost-plus basis, which effectively combines forward sale with insurance arrangements. The defence industry is unusual, however, in having a single dominant institutional customer with ready access to capital on account of its powers of taxation. In consumer goods industries, by contrast, forward sales are not really practical because the very large number of small orders involved means that transactions costs would be very high.

It is because forward contracts cannot be entered into that the profitability of intermediation becomes so uncertain. There are, of course, ways in which, in the absence of forward contracts, the set up costs of intermediation - and hence the down-side risks - can be reduced. For example, trade credit can be obtained by paying for supplies after they have been received. Indeed, if the entrepreneur is highly reputable then he may be able to obtain prepayment from buyers sufficiently early that he can pay for this supplies using the revenues from his sales. It is unusual, however, for a new enterprise to enjoy a reputation of this kind.

If the set up costs of the firm are exceeded by the owner's personal wealth then there is no financial obstacle to starting the business. Otherwise the owner will have to dilute his role by borrowing from other people. Fixed interest loans may be obtained where suitable collateral is available - inventory for short-term loans and buildings for long-term loans, for example. But the main source of funds will normally be equity.

Since the ultimate power of control normally resides where the principal financial responsibility lies (Knight 1921) equity holders will normally demand a right of veto on any proposed use of their funds. As a result the founder of the firm is likely to become part-owner and part-employee of his firm - delegated to use his discretion under the general supervision of a board comprising the major shareholders. The shareholders count on the fact that the arrangement does not significantly affect the motivation of the entrepreneur. In practice the major problem is likely to be, not any dilution of effort by the entrepreneur, but rather his concern with other objectives - like growth - which may be a variance with the narrower interests of the other shareholders (Marris 1964).

It is not necessary that the entrepreneurial function is performed by a single individual, though. The advantage of consultation in pooling relevant life-experiences was mentioned above. Major shareholders may wish to contribute their expertise in this way. In this case the deliberations of the board may actually strengthen the entrepreneurial role within the firm.

The willingness to share experience of this kind, and the need for honesty about what is shared, raise difficult incentive problems, however. The value of what a person contributes when they share their experiences is difficult to measure. The influence of a person's contribution to the emergence of a consensus on a particular issue - and their consequent responsibility for any mistakes - is also difficult to assess. For this reason, most firms use very simple rules to decide such issues. Each equity share, for example, carries a single vote, and conflicts are resolved by applying majority rule. In small partnerships shares may not be freely tradeable, and the right to additional shares may be dependent on the senior partner's judgement of a junior partner's contribution to the firm. It is very difficult to justify these rules except as norms and customs designed to avoid the information costs which would be incurred if every aspect of joint ownership became the subject of negotiation.

No single rule can resolve all the incentive problems satisfactorily. Ultimately there is no substitute for a high degree of trust between the major partners in a firm (Buckley and Casson 1988). Just as their rights in the firm are general rights of ownership rather than claims on specific services, so their obligations to the firm are general rather than specific too - to contribute all they know that is relevant to the firm, should there be a need for them to do so. It is therefore advantageous if they are drawn from a reasonably homogenous cultural group. They may well know each other personally, or know the entrepreneur. They must have sufficient in common to share a similar vision, and to have the same degree of confidence in it as does the entrepreneur.

Not every shareholder may wish to hold a substantial equity stake, however. The advantages of 'portfolio' diversification mean that some shareholders may simply wish to take a small stake as part of a policy of spreading their risks. Such share holdings can be encouraged by floating the firm's equity on a stock exchange. By reducing the transaction costs of both buying and re-selling equity, this not only markets the equity to a wider group of investors, but provides additional liquidity for the existing investors too.

The smaller shareholders will generally not wish to play an active role in managing the firm. This is because the costs of involvement in management are fixed costs, independent of the size of the shareholding. The small shareholders are in effect free-riding on the efforts of the larger shareholders to manage the firm (Grossman and Hart 1980). The larger shareholders may have ways of extracting rents to compensate for their costs, however: management fees, capital gains from insider-trading, directing trade to other companies with which they are financially involved, and so on. The scope for abuse is obvious. There is, however, the sanction that once abuses are discovered the small shareholders may sell out to a takeover raider who disciplines the larger shareholders involved.

The potentially 'footloose' nature of small investments can damage the loyalty and effort of ordinary employees. It is one thing to call upon an employee to match the commitment of an owner who has sunk all his personal wealth in building up a firm, and another to call for the same degree of commitment on behalf of a diverse collection of anonymous speculative investors. The discussion suggests that the links between stock-exchange financing, ownership structure and company performance, which are currently receiving much public attention, have an important ethical dimension which has not been emphasised as much as it should (see above).

Reflections on the 'Ownership' Theory of the Firm

According to the theory presented above, ownership of a firm is an essentially speculative activity. The aim of the owner is to appropriate rents from what he believes to be privileged information. The information may relate to the hidden quality of some asset, but the most important case concerns an opportunity for intermediation. This opportunity may be associated with technological innovation; in various sections above though this is by no means always the case. The owner cannot readily appropriate the rent from information by licensing its use to other people because property rights in ideas of this kind are insecure. The idea must be kept a secret until the opportunity has been preempted and, since it is difficult to sell a secret, he must preempt the opportunity himself.

Another way of appropriating rent from information is by placing a bet upon the subject with other people who are believed to be not so well informed. The rent is extracted when the truth is revealed and the bet is won. This is, in effect, what happens when a successful firm acquires resources from other people. The owner of the firm bets that his judgement of the situation is better that the judgement of those who sell their resources to the firm. The profit of the firm is his reward for being right.

Differences in judgement exist because different people use different symptoms to interpret the environment. People with wide-ranging but unusual experiences have the kind of entrepreneurial judgement which is at once both 'deviant' in terms of received opinion, and yet often correct. Other people who have confidence in the judgement of a person of this kind may contribute risk capital to the firm he founds. This helps to found the indivisible sunk cost required to establish a new market. Major investors may also contribute their opinions on how the founder's idea is best exploited. Minor investors may be involved too; they essentially free-ride on the judgement of the principal entrepreneur and the major investors. All these investors count on the entrepreneur to exercise vigilance even though he may own only a modest portion of the equity.

The exploitation of an opportunity requires that once the relevant resources are acquired they are utilised in an appropriate way. Versatile resources confer power of control on the owner. An intermediator owns two main types of resource: the assets he uses to operate his business - office equipment and the labour-time of his employees - and the goods that he 'turns over' as a result of these operations. To fix price and output on the market that he makes, the intermediator requires a constant flow of information. This information is routinely collected by the office staff. Their role is to synthesise two types of transitory information, concerned with demand and supply. Information processed by stocks of resources of the first type is thus used to control the flow of resources of the second type. Control of the flows is effected using a procedure devised by the intermediator which is intended to achieve the appropriate degree of accuracy in decision-making at minimum cost.

Once he has devised this procedure, the entrepreneur must keep it under constant review. Should circumstances change, and another procedure become appropriate, he must beat potential competitors on speed of response. He has a natural advantage in his ability to feed back lessons from the use of the existing procedure, but vigilance is required to do this properly. To change procedures when circumstances require it is a prerogative of ownership, but there are also constraints on the extent to which this can be done without consulting, and if necessary, compensating, loyal employees who are adversely affected by the change. While the formal contract of employment may seem seriously incomplete, it is 'completed' by the procedures, the constitution and the culture of the firm.

Grossman and Hart (1986) offer a very different account of ownership, however. In their account, the emphasis on speculation is missing. According to them the owner seeks control, not in order to specify a decision-making procedure appropriate to the intermediation opportunity he has identified, but simply to gain the upper hand in a game of strategy played out with other people who contribute resources to the firm. Grossman and Hart lay great emphasis on the incompleteness of the employment contract. They argue that this creates an ambiguity which is resolved by allocating residual control - i.e. control over all the decisions not fully specified in the contract - to one particular party, namely the owner of the firm. Such an account of ownership is misleading because it overstates the incompleteness of contracts and thereby suggests much more scope for strategic behaviour than is possible in practice. It ignores the fact that the owner has committed himself to certain procedures and that although he can change these procedures, the probability that he will do in any given period is actually very small. Moreover, his right to change these procedures in a manner which seriously damages the interests of other parties is constrained by the constitution of the firm. Should the owner breach the constitution he will lose the loyalty of key employees, who can inflict considerable damage on the firm if they quit.

The incompleteness of contracts certainly exists as a rational response to information costs. It does not, however, create a major ambiguity which strategic behaviour by a specialised residual claimant is required to control. Rather it creates a need for employees to acquaint themselves with the constitution and the procedures of the firm and to form rational expectations of how their conduct is likely to be regulated by these procedures on a day-to-day basis.

The only substantial way in which contracts are incomplete is that the owner retains the right to change the procedures in the light of changes he believes he has recognised in the environment of the firm. This is simply the entrepreneurial function, which Grossman and Hart apparently fail to recognise when they see it. The owner is far too concerned with maintaining procedures that efficiently exploit his intermediation opportunity to be bothered, to any significant extent, with the strategic fine-tuning that Grossman and Hart discuss. Such fine-tuning may well occur, but it is the speculative factor that Grossman and Hart omit that is the principal factor governing of the ownership of the firm.

Reflections on the New Institutional Economics

Williamson's theory of the firm is based on three key assumptions of opportunism, bounded rationality and asset-specificity (Williamson 1975, 1985). Opportunism is relatively clear and uncontroversial; it describes a phenomenon well known to all economists - the pursuit of material self-interest under conditions of asymmetric information. Bounded rationality is more controversial - to economists at least - and it is also more obscure. While it is easy to see what bounded rationality denies - namely the substantive rationality of full optimisation - it is not clear exactly what it substitutes for it. Simon's (1947) alternative is procedural rationality, which is akin to the idea of using optimal procedures in a world of costly information. Williamson does not, however, develop a calculus of optimal procedures of this kind, as described above. Instead, bounded rationality is simply invoked at key points of the argument to explain why contracts are incomplete.

As noted before, however, the significance of contractual incompleteness has been rather exaggerated in recent literature. Moreover, at other points in his argument Williamson side-lines bounded rationality altogether. He does not, for example, consider bounded rationality to be a serious obstacle to the selection of an efficient organisational form. It is because the concept of bounded rationality is never properly defined that Williamson is able to deploy it in such a discretionary manner.

Williamson has accorded asset-specificity a central place in his analysis ever since Klein, Crawford and Alchian (1978) highlighted its implications for vertical integration. Asset specificity is another concept that exhibits great versatility in Williamson's hands. An asset becomes specific when costs are sunk to customise the asset to the requirements of another party. Rendering an asset specific weakens the owner's bargaining position and encourages the other party to renegotiate the contract. Forward contracts are legally too fragile to prevent this, and so pooling the ownership of the assets is the natural response. The resulting concentration of ownership generates the vertically integrated firm. The key point in this argument is that asset-specificity reduces substitution opportunities and weakens competitive threats. The reduction in competition causes problems with bilateral market power.

Now bilateral monopoly is ubiquitous. No market is perfectly competitive; even in a relatively atomistic market there is almost always one buyer who is closest to any given seller, and vice versa, which means that with positive transport costs there is invariably an element of bilateral market power. In this sense asset-specificity emerges any time anyone fixes a location for their activity. Asset-specificity is so widespread that it is hardly surprising that almost any kind of vertical integration can be explained in these terms if desired.

Asset-specificity is thus a very versatile concept; it is so versatile, in fact, that even assets that seem demonstrably general can emerge as highly specific in Williamson's terms. Consider technological know-how, which was mentioned earlier as a prime example of general information because it can be used anywhere in the world. Such versatility does not prevent it being highly specific in Williamson's terms (Kay 1993). This is because the R & D expenditures are already sunk before the know-how can be licensed, and so the knowledge is highly-specific to the requirements of local licensees. Although the problems of licensing know-how are in practice caused by weaknesses in patent rights, and the difficulty of selling secrets, the rejection of the licensing option in favour of formal integration into production can therefore be explained superficially in terms of asset-specificity too. The problem with asset-specificity is that it is just too successful in accounting for vertical integration. The real challenge is to explain why vertical integration does not occur in the many cases where asset-specificity is present, rather than to account for why it does occur in a few of the many cases where asset-specificity can be found.

Williamson is undoubtedly correct in his contention that a comprehensive theory of the firm must draw upon the insights of Simon (1947), Commons (1934) and Coase (1937). His own analysis hardly qualifies as comprehensive, though. Williamson has little to say about intermediation, and almost nothing to say about the role of entrepreneurship in the formation and the growth of the firm. A sense of the truly global dimension to the exploitation of general ideas is missing too. The problem seems to be that bounded rationality is too vague, and asset-specificity too ubiquitous, to build an entire theory on just these foundations. It has been suggested in this paper that a better approach is to analyse the firm as a specialised intermediator created by an entrepreneur to routinely synthesise information about different sources of volatility. Information costs take the place of bounded rationality in this approach. Asset-specificity is supplemented by a number of other factors to provide a full account of vertical integration. The result is a theory that is entirely consistent with the principle of rational action employed in mainstream economics. It is these concepts of intermediation, synthesis, volatility and information cost - rather than opportunism, bounded rationality and asset-specificity - which turn out to be most crucial when economic principles are properly applied to the theory of the firm.

Summary and Conclusions

This paper has provided a definition of the firm as a specialist decision-maker and then employed a small number of key concepts to examine a number of related facets of the firm. It has been argued that previous writing on the firm has been essentially partial, with different writers focusing on different issues. A synthesis has been proposed which leads towards a theory of the firm centred on the entrepreneur as the founder and prime-mover within it.

The basis of this theory is that an entrepreneur monitors the environment in order to detect symptoms of change. This leads him first to found the firm and subsequently to effect intermittent changes in its procedures. These procedures synthesise information from two main sources of volatility - namely the demand factors and the supply factors in the market that the entrepreneur has set up. While shocks in the general environment tend to persist for a considerable time, shocks to the market, once it is set up, are of a more transitory nature. That is why routine procedures are an efficient way of dealing with them.

The managers who are delegated by the entrepreneur to monitor the volatile factors also have responsibility for implementing decisions. They are supported by administrative staff, who have to be in post before the decisions are made. The work of these staff is in turn disturbed by volatile factors of very high frequency. These disturbances arise from breakdowns, illnesses, and so on; they may be interpreted as very short-run production shocks, since they affect the production of intermediating services. Efficient response to these shocks calls for spontaneous cooperation amongst the administrative staff involved. This in turn calls for 'horizontal' information flow within a high-trust environment.

There is an asymmetry of information in the supply of the product to the intermediating firm. This raises problems of quality control which, together with other considerations, such as missing forward markets, may lead to backward integration in production.

General changes in tastes, and advances in technology, may require a global response. If the entrepreneur has the vision to recognise this, then the scope of the firm will extend well beyond a single market. In this case multiple sources of volatility will be encountered within the firm. Internal substitution possibilities are likely to emerge which require rather different kinds of procedures to those of a single market. In a single market complementarity between purchases and sales is the dominant consideration and information is most naturally encoded in quantity form. With multiple markets it may be advantageous to encode information in the form of internal prices instead, since the synthesis of price information is a natural means of coordinating interrelated substitution decisions.

The basic point about the firm is that all the relevant information on transitory factors is, in principle, shared within it. Some may be suppressed in the interests of economy, and some may be communicated in the form of orders, but at all times the main decision-making procedures are premised on the information being true. These procedures may be supplemented by monitoring procedures, or by moral rhetoric, both of which aim to validate this assumption of truth. The market alternative to the firm is very different to this. The market approach accepts that bluffing is likely and attempts to constrain it by competition. Factual information is simply encoded in price quotations, and coordination is then effected through negotiation. Competition and the law replace monitoring and moral rhetoric in the transition from the firm to the market.

Overall, the nature of the firm is most naturally explained in terms of the qualities required of a successful entrepreneur. The entrepreneur processes information on both persistent and transitory factors. He normally processes the former himself, albeit on an intermittent basis, and delegates the processing of the latter. Delegation is effected using procedures, some of which impose decision rules on managers and some of which allow improvisation instead. Where improvisation is allowed, entrepreneurship is partially diffused within the firm. The key to the firm's success lies not in specific business strategies, as suggested by Porter (1980), nor in specific ownership advantages, such as technology, as suggested by Dunning (1977). These are both short run consequences of long run decisions about organisation and corporate culture taken previously by the entrepreneur. It is the quality of entrepreneurial judgement, as reflected in the correctness of these decisions, which holds the key to long-run success.

Good judgement involves correct interpretation of the symptoms of changes in the market environment. Some of these concern tastes (in respect of consumer demand) and some concern resource availability (in respect of product supply). Others concern technology; production technology is important in respect of product design and choice of technique, but information technology is just as important too. Changes in information technology - from the telegraph through to the internet - alter information costs and so impact directly on the organisation of the firm. Value change - in morals and social customs - must be considered too. The successful entrepreneur invests in building networks of trust between his employees, and the amount of investment required will depend on the degree of trust that already exists in the society in which the firm operates. Interpreting symptoms of value change is one of the most challenging areas of entrepreneurship because of the inherent subjectivity of the phenomenon.

As a firm grows, the range of symptoms that need to be interpreted grows as well. The entrepeneur must delegate more, and the screening of delegates then becomes a major task. Interpreting symptoms of personal competence and integrity in potential employees, and sustaining motivation amongst those recruited, become increasingly important skills. To take advantage of all the local opportunities offered by a truly global market-making concept, the firm may have to grow faster than the reinvestment of profits will allow. This requires the entrepreneur to take on equity partners too. In looking for partners who can also play an executive role, the entrepreneur must screen for a wide range of knowledge and experience complementary to his own, together with a willingness to share it without reserve. The selection of business partners raises issues similar to the selection of managerial delegates, but in even more acute form.

From a knowledge of markets, through to a knowledge of technology and a knowledge of people, it is the ability to interpret symptoms and synthesise results that emerges as the hall-mark of the successful entrepreneur. The only thing to add is that the successful entrepreneur requires self-knowledge as well. When backing his own judgement against that of others, he needs to be sure that his confidence is well-placed. Conversely, he must frankly acknowledge his own shortcomings, because it is these that dictate the nature of the complementary expertise that he must recruit. 'Know thyself' is not only a moral injunction, the theory suggests, but a key to entrepreneurial success as well.

Note

1 I am grateful to Peter Buckley, John Cantwell, Eric Jones and Animesh Shrivastava for comments on an earlier draft.

References

Aoki, M., Horizontal vs. Vertical Information Structure of the Firm, American Economic Review, 76, 1986, pp. 971-983.

Bernhardt, I., Vertical Integration and Demand Variability, Journal of Industrial Economics, 25, 1977, pp. 213-229.

Buckley, P. J. and M. C. Casson, The Future of the Multinational Enterprise, London: Macmillan 1976.

Buckley, P. J. and M. C. Casson, A Theory of Cooperation in International Business, in F.J. Contractor and P. Lorange (eds), Cooperative Strategies in International Business, Lexington MA: Lexington Books, 1988, pp. 31-53.

Carlton, D. W., Vertical Integration in Competitive Markets under Uncertainty, Journal of Industrial Economics, 28, 1979, pp. 189-209.

Carter, M. J., Information and the Division of Labour: Implications for the Firm's Choice of Organisation, Economic Journal, 105, 1995, pp. 385-397.

Casson, M. C., The Entrepreneur: An Economic Theory, Oxford: Martin Robertson 1982, reprinted Aldershot: Gregg Revivals 1991.

Casson, M. C., The Firm and the Market: Studies in Multinational Enterprise and the Scope of the Firm, Cambridge MA: MIT Press 1987.

Casson, M. C., Economics of Business Culture: Game Theory, Transaction Costs and Economic Performance, Oxford: Clarendon Press 1991.

Casson, M. C., Why are Firms Hierarchical? International Journal of the Economics of Business, 1, 1994, pp. 47-76.

Casson, M. C., Entrepreneurship and Business Culture, Aldershot: Edward Elgar 1995 a.

Casson, M. C., Information and Economic Organisation, University of Reading Discussion Papers in Economics, No. 317, 1995 b.

Casson, M. C., Comparative Organisation of Large and Small Firms: An Information Cost Approach, Small Business Economics, 8, 1996, pp. 1-17.

Coase, R. H., The Nature of the Firm, Economica (New series), 4, 1937, pp. 386-405.

Commons, J. R., The Legal Foundations of Capitalism, New York: Macmillan 1934.

Dunning, J. H., Trade, Location of Economic Activity and the Multinational Enterprise: A Search for an Eclectic Approach, in B. Ohlin, P. O. Hesselbom and P. M. Wijkman (eds), The International Allocation of Economic Activity, London: Macmillan, 1977, pp. 395-418.

Geanakopolous, J. and P. Milgrom, A Theory of Hierarchies Based upon Limited Managerial Attention, Journal of Japanese and International Economics, 5, 3, 1991, pp. 205-225.

Grossman, S. J. and O. Hart, Take-over Bids, the Free-rider Problem and the Theory of the Corporation, Bell Journal of Economics, 11, 1980, pp. 42-64.

Grossman, S. J. and O. Hart, The Costs and Benefits of Ownership: A Theory of Vertical and Lateral Integration, Journal of Political Economy, 94, 1986, pp. 691-719.

Hajek, F. A. von, Economics and Knowledge, Economia (New Series), 4, 1937, pp. 33-54, reprinted in F. A. von Hayek, Individualism and Economic Order, London: Routledge and Kegan Paul 1959, pp. 33-56.

Hirschman, A. O., Exit, Voice and Loyalty, Princeton, NJ: Princeton University Press 1958.

Hymer, S. H., The International Operations of National Firms: A Study of Direct Investment, PhD thesis, MIT 1960, published Cambridge MA: MIT Press 1976.

Kay, N. M., Markets, False Hierarchies and the Role of Asset Specificity, in C. Pitelis (ed), Transaction Costs, Markets and Hierarchies, Oxford: Blackwell, 1993, pp. 242-261.

Kirzner, I. M., Competition and Entrepreneurship, Chicago: University of Chicago Press 1973.

Klein, B., R. G. Crawford and A. A. Alchian, Vertical Integration, Appropriable Rents and the Competitive Contracting Process, Journal of Law and Economics, 21, 1978, pp. 297-326.

Knight, E. H., Risk Uncertainty and Profit, Boston: Houghton Mifflin 1921.

Marris, R. L., Economic Theory of Managerial Capitalism, London: Macmillan 1964.

Milgrom, P. R. and J. Roberts, Economics of Organisation and Management, Englewood Cliffs NJ.: Prentice Hall 1992.

Polanyi, M., Science, Faith and Society, Chicago; University of Chicago Press 1964.

Porter, M. E., Competitive Strategy, New York: Free Press 1980.

Robertson, D. H., The Control of Industry, London: Nisbet 1923.

Sah, R. K. and J. E. Stiglitz, The Architecture of Economic Systems: Hierarchies and Polyarchies, American Economic Review, 76, 1986, pp. 716-727.

Schumpeter, J. A., The Theory of Economic Development (trans. R. Opie), Cambridge MA; Harvard University Press 1934.

Schumpeter, J. A., Business Cycles: A Theoretical, Historical and Statistical Analysis of the Capitalist Process, New York: McGraw-Hill 1939.

Simon, H. A., Administrative Behaviour, New York: Macmillan 1947.

Simon, H. A., A Formal Theory of the Employment Relation, in H. A. Simon, Models of Man: Social and Rational, New York: John Wiley & Sons 1957.

Spence, A. M., Job Market Signalling, Quarterly Journal of Economics, 87, 1973, pp. 355-374.

Waterson, M., Vertical Integration, Variable Proportions and Oligopoly, Economic Journal, 92, 1982, pp. 129- 144.

Williamson, O. E., Markets and Hierarchies: Analysis and Anti-trust Implications, New York: Free Press 1975.

Williamson, O. E., The Economic Institutions of Capitalism, New York: Free Press 1985.

Winter, S. G., On Coase, Competence and the Corporation, Journal of Law, Economics and Organisation, 4, 1988, pp. 163-180.

Author

Mark Casson, Professor of Economics, Department of Economics, Faculty of Letters and Social Sciences, University of Reading, Reading, U.K.
COPYRIGHT 1996 Gabler Verlag
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1996 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:International Business Theory: The Nature of the Firm and the Role of Management
Author:Casson, Mark
Publication:Management International Review
Date:Jan 1, 1996
Words:17979
Previous Article:The role of management in international business theory: a meta-analysis and integration of the literature on international business and...
Next Article:Wise before the event: the creation of corporate fulfilment.
Topics:


Related Articles
Reconceptualizing entrepreneurship: an input-output perspective.
High technology and globalization challenges facing overseas Chinese entrepreneurs.
Entrepreneurship.
Entrepreneurship in multinational enterprises: a Penrosean perspective.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |