1. ‘Ownership’ of the Firm
The firm in transactions cost analysis is a ‘nexus of contracts’ as explored in Chapter 2. These ‘contracts’ represent promises between participants in the firm. The ability and incentive of the participants to fulfil their contractual obligations will depend upon their assignments of property rights. Within the firm, property rights in resources are assigned to the various participants. Different assignments of property rights can then be seen as characterising different types of organisation (for example, the proprietorship, the partnership and the joint-stock enterprise) as discussed in sections 5 and 6. If transactors successfully pursue efficiency gains, property rights will be assigned in such a way that the gains from trade net of all costs of transacting are maximised. This type of analysis of property rights is thus complementary with transactions cost theory.
An interesting aspect of the transactions cost theory as summarised above, however, is that the concept of ‘ownership’ is not well defined. Assets seem not to have ‘owners’. Rights to use assets and to claim income flows are simply divided up between cooperating groups of people in ways which facilitate mutual gain. The quote from Demsetz (1967) at the end of section 1 simply mentions ‘the owner of property rights’, suggesting that the ‘owner’ is simply a ‘possessor’ or a ‘holder’ of rights. In so far as a more specific idea of ‘ownership’ has so far appeared in the analysis, it has implicitly been associated with residual claims. References, for example, to the ‘division of ownership from control’ suggest that the residual claimants (shareholders) are the ‘owners’ of the firm but no longer are really in a position to exercise control over the use of the assets.
Given the complexity of property rights assignments – the fact that many people may be contractually permitted to use an asset for specified purposes and may claim portions of the resulting profit – is it possible to identify an ‘owner’ from amongst all those who can be seen as holding property entitlements? The answer to this question has led to a view of the firm which differs from the pure ‘nexus of contracts’ approach. In modern property rights theory, the ‘owner’ of an asset possesses ‘residual rights of control’. The word ‘residual’ here, does not refer to the right to receive profit (the residual). It is used as an adjective, not as a noun. It refers to those rights of control which have not explicitly been ceded to others by means of contract. They are residual rights in the sense that they are ‘left over’ and remain with the owner. The most important thing about these ‘residual control rights’ is that they cannot be exhaustively listed and written down. If they could be so listed, they would not be ‘residual’ rights. They would simply be one subset of specified property rights – no different in principle from anyone else’s non-overlapping subset of specified rights which, when summed together, exhaust all possibilities. If every contingency can be contracted for, the ‘owner’ disappears. ‘Ownership’ is important because all contingencies cannot be contracted for. When contingencies occur for which there is no contractual provision, it is the ‘owner’ who has the right to determine how an asset should be used.
Ownership is thus intimately associated with contractual incompleteness. In subsection 2.1 we saw that ‘private rights’ in housing could be divided between landlord and tenant. What distinguishes these subsets of rights? In principle, the tenant’s rights are delineated in a contract. The landlord retains all those rights which are not explicitly contracted away for a specified length of time. Thus, the landlord is the ‘owner’ of the housing even if he or she cannot live in it for the time being. It might be objected that even the tenant’s rights cannot be exhaustively written down and that interpretation and ‘gap filling’ by the courts will be important. To this, the response is simply that, in their attempts to fill gaps in contracts, the courts are trying to discover what comes within the domain of the contract and what is outside it. Providing there is an acceptance that some matters are not, and were never intended to be, contractually determined (the tenant’s subset of rights is in some sense ‘bounded’), the ‘owner’ with his or her ‘unbounded’ set of residual rights has a distinct and important role.
Recognition of the distinct function of ‘ownership’ has resulted in the development of two rather different (though complementary) types of theoretical analysis. In the first, associated with the work of Hansmann (1996), the ‘nexus of contracts’ view of the firm is maintained and extended. In the second, associated with the work of Grossman and Hart (1986), attention is directed at the problem of determining the circumstances under which there will be advantages to different assets coming under the control of a single ‘owner’ (integration within the firm) and the circumstances when it will be better to leave the assets under the control of different owners. The ‘firm’ for Grossman and Hart is defined by the common ownership of a set of physical assets, not by the nature of the contracts which bind it to its suppliers. In subsections 7.2 and 7.3 we outline these two approaches to ‘ownership’.
2. Hansmann and the Costs of Ownership
For Hansmann, the firm is the hub of a set of contracts with its ‘patrons’. Patrons are simply any people who find it useful to deal with the firm. They might be workers, providers of capital, consumers of the firm’s output, or suppliers of other inputs such as raw materials, professional services or intermediate goods. Hansmann is not concerned with the nature of these contracts. Whether they are durable and relational or arm’s length and conducted on spot markets is not the focus of attention. The main feature of concern to Hansmann is that some patrons will be ‘owners’ and others will not. ‘Owners’ are those patrons who possess ‘residual rights of control’. They are also likely to possess rights to the residual; that is, to the profit. It is usually efficient that control rights and profit rights are held together, although they might, in principle, be held by separate people.
What determines the allocation of ownership rights among the patrons? Efficiency requires that they are assigned to maximise the gains to trade after allowing for all costs of transacting and the costs of ownership. Transactions costs have been discussed in detail in Chapter 2 – the costs of search, bargaining and contractual opportunism. Ownership costs are new. In so far as ownership is associated with claims to profit, one cost is the bearing of Knight’s ‘uncertainty’ – uninsurable risk. The owners have to accept fluctuations in the value of the firm’s ‘equity’, which in the joint- stock enterprise is represented by the market value of shares but in other types of enterprise will not take such a clear-cut form. Another cost given a central role by Hansmann is the cost of exercising residual control rights. Again, it is possible to see this cost as bound up with entrepreneurship. At certain points, where contract is silent, owners determine the use of assets. If we envisage such owners as single individuals, they are entrepreneurs of the variety discussed by Casson – making judgemental decisions. If, as is common, ownership rights are shared between many patrons, they are the group which must determine overall priorities and appoint and monitor the managers and agents who will act on their behalf.
The cost of controlling managers and agents is of central importance in the theory of the firm and is discussed in detail in Chapters 8 and 9. We have already seen in section 5.3.2 that the incentive to monitor in a partnership is undermined as claims to the profit are shared more widely. Hansmann’s distinctive contribution, however, is to emphasise the costs of collective decision-making as a category of ‘ownership cost’. If ownership is shared between patrons, they must exercise their control rights by means of some ‘collective choice mechanism’. The individual preferences of the owners have to be aggregated and somehow transformed into a decision which can be said to ‘represent’ them as a group. The ownership of enterprise is thus closely connected to the problems of ‘public choice’ – not at the level of the state, but at the microeconomic level of the firm. Each firm has a ‘constitution’ which will specify how the ‘owners’ are to exercise their collective rights. Collective choice processes are not costless, however. Each person faces the costs of becoming informed and deciding where their own interests lie. In this function, as in the monitoring function, each owner might try to free-ride on the work of others. They might calculate that their individual influence on the outcome was likely to be so slight that extensive information gathering and analysis would not be individually worthwhile (the problem of ‘rational ignorance’). They might not even bother to register their preferences and take part in whatever voting processes are used (the problem of ‘rational abstention’). Even where informed owners register their preferences, the choices of the group as a whole might be paradoxical and unrelated to these individual preferences (the problem of ‘preference aggregation’ or ‘the paradox of voting’).
The simplest and most famous demonstration of the ‘paradox of voting’ is as follows. Suppose there are three voters who must decide collectively which of three available mutually exclusive choices they should make. Voter 1 ranks the options in order of preference ABC. Voter 2 ranks them in the order BCA and voter 3 in the order CAB. If the three people take each pair of options in turn and vote according to a majority rule, they will find that, collectively, they prefer A to B, B to C, and C to A. This is the classic ‘voting cycle’. Assuming the final choice to be made by elimination, the outcome would be arbitrary and depend entirely upon the order in which the options were put to the vote.22 It is not true, of course, that a voting cycle will always be generated. If the rankings of the voters were ‘sufficiently similar’ the problem would not necessarily arise. Collective choice will obviously be a great deal easier in circumstances where there is a fair amount of agreement between the voters. This leads to the conclusion that, where ownership is shared between many patrons, collective decision-making costs will be lower if these patrons comprise a reasonably homogeneous group.
Hansmann argues, therefore, that minimising the costs of the firm’s transactions will imply the allocation of ownership rights to two groups:
- to those who would otherwise face relatively high costs of market transacting; and/or
- to those for whom the costs of ownership are relatively low.
The first point above derives from the fact that ownership can be a substitute for contract. If post-contractual opportunism is a serious problem, for example, the patrons fearing ‘hold-up’ might become owners. They will now have control and will therefore avoid the dangers involved in ‘contract renegotiation’. Similarly, if another group of patrons faces severe adverse selection or moral hazard problems, ownership might mitigate the difficulty, providing that monitoring costs and other costs of ownership are not too great. In Chapters 10 and 11, we will apply this type of reasoning to the worker-owned firm as well as to consumer cooperatives, mutual enterprises and non-profit firms.
The second point simply emphasises that it is no good avoiding the costs of contract if the costs of ownership are even greater. As will be seen later in Part 2, it is high costs of ownership that lie behind many of the problems of worker-owned firms and non-profits. Workers tend to have heterogeneous interests unless ownership is confined to a small group such as in a professional partnership. In contrast, a major advantage of the investor- owned joint-stock enterprise, notwithstanding the problem of managerial monitoring which will be discussed in Chapters 8 and 9, is the relatively homogeneous interests of the group of patrons who own it.
3. Grossman and Hart and the Property Rights Theory of the Firm
Hansmann’s approach to the ‘ownership’ of the firm is rooted in transactions cost analysis. In contrast, the property rights approach to the firm as developed by Grossman, Hart and others does not focus on transactions costs. It starts by asking the question: why does it matter who holds residual rights of control in an asset? What is the nature of the social benefit derived from an ‘efficient’ rather than an ‘inefficient’ allocation of control rights? Hansmann and other transactions cost theorists would answer that the social benefit takes the form of lower costs of transacting. Property rights theorists argue that this answer is not precise enough. They show this by imagining a world in which bargaining costs are actually zero but in which long-term contracts are incomplete. Some promises are simply ‘nonverifiable’. This does not necessarily mean that behaviour or results are unobservable. It means that it is impossible to construct a contract which a court or some third party can interpret accurately in order to tell whether a promise has or has not been met. One party might promise to publish a book of a particular word length. This would easily be verifiable. Once the nature and quality of the book become important, the terms of the contract become ‘non-verifiable’. The manuscript exists, its contents are observable, but is it what the publisher was expecting and had contracted for?
The property rights theory of the firm is based upon the idea that the assignment of residual control rights matters in a world where some issues concerning ‘quality’ are non-contractible. Property rights matter not because they determine the costs of transacting. ‘Quality’ is non-contractible, irrespective of how property rights are assigned, thus effectively rendering a long-term enforceable contract ‘infinitely costly’. Property rights matter because they influence the ‘power’ that transactors have in their (zero-cost) post-contractual bargaining when the provisions of a contract are nonverifiable. Why, though, should the distribution of such ‘power’ matter to society as a whole? Will not one person’s loss be another person’s gain when this post-contractual bargaining happens? The answer is that people are forward looking. They know that their contractual arrangements are incomplete and that they cannot rely on third parties to enforce all provisions. They know, therefore, that post-contractual bargaining will take place and that ‘hold-up’ will occur. This certain knowledge that they are vulnerable to holdup will influence behaviour. In particular, it will influence each party’s willingness to make ‘ex ante’ transaction-specific investments.23
As we saw in Chapter 2, transaction-specific investments are those which raise the return to a particular contractual relationship but which have a lower payoff outside. A supplier, for example, may spend time and resources learning how best to satisfy the requirements of a buyer. These are investments in ‘performance’. A buyer may expend resources preparing for the delivery of the supplier’s goods or services. A restaurateur, for example, may invest in publicity and in planning menus prior to the delivery of interior designs. These are investments in ‘reliance’. The essence of the property rights theory of the firm is that the combined incentive to invest in performance and reliance will depend upon how property rights in physical assets are assigned.
The contractors know that their ex ante investments are non-verifiable. They know that after they have made these investments they will bargain about the distribution of the ex post gains from trade. Each will want to be in as strong a position as possible when this bargaining takes place. A ‘strong’ position in this context is the ability to walk away from the agreement with as little penalty as possible. In other words, the person with the most valuable outside opportunities will be in a stronger position than a person who is very ‘dependent’ on the relationship and has few opportunities elsewhere. Each person, therefore, will have a ‘threat point’, a payoff below which he or she cannot be squeezed because they will be better off simply terminating their existing agreement and trading with someone else.
Holding residual control rights in physical assets is important because, in the event of a breakdown in an agreement, such rights might enable alternative possibilities to be pursued. Consider our example of the restaurant. Imagine that there is a single physical asset – the building itself. Who should be assigned the control rights in the building – the restaurateur (chef) or the interior designer? Each might be expected to make greater ex ante investments if they own the building than if they do not. If the designer owns the building and his or her contract with the restaurateur breaks down, he/she can at least open the restaurant using the services of some other chef hired in the market and so install his or her interior designs. Similarly, if the restaurateur owns the building s/he can, on the collapse of the original agreement, go ahead and open the restaurant even if the interior is not as attractive as s/he had hoped. The crucially important question is therefore the sensitivity of investments in performance and reliance to the possession of residual control rights. The greater the sensitivity of a contractor’s ex ante investments to control of a physical asset, the more powerful the case for the assignment of ownership to that contractor.
Without the availability of the building, the restaurateur will find advertising and planning to no avail. Perhaps a different location could be found and a lease arranged, but geographical position and exterior appearance are important especially as these might have featured prominently in marketing efforts. Failure of the contract with the building-owning interior designer would therefore be expected to be serious for the restaurateur. He or she might consider the whole investment effectively wasted. Control of the building would be likely to provide a powerful incentive to greater investment on the part of the restaurateur because there is the reassurance that, even without the interior designs, the investment will be productive. Is the position of the interior designer similar?
For the sake of argument, suppose that the designer has other outlets for his or her services. Although the designs for the restaurant are, to some degree, specific to the original restaurateur’s requirements, the ideas can be used, with some not too costly modifications, in restaurants, clubs or pubs elsewhere. This implies that, although his or her effort will be somewhat less if s/he does not own the building, the assignment of ownership to the restaurateur will not greatly diminish the designer’s ‘up front’ commitment. Changing the ownership of the building from the designer to the restaurateur will radically increase the latter’s ex ante investment and only marginally reduce that of the former. Thus, the restaurateur is the most efficient owner.
Another way of expressing this result is to note that, as this story has been set up, the restaurateur’s human capital is complementary to the physical asset to a much greater extent than is the designer’s human capital. Ownership of the physical asset does not help the designer much. S/he needs the chef’s human capital before the project becomes really successful. In the absence of the chef’s human capital (that is, if the contract breaks down and the agreement is terminated) the designer gains little from ownership of the building. The chef’s human capital is thus ‘essential’ in the technical sense that, without it, the designer gains nothing from ownership of the physical asset. If the human capital of one of the parties to a contract is ‘essential’, it is this party that should own the asset.
Although the threat to trade elsewhere and renounce a non-verifiable agreement plays a central role in property rights theory, it would be wrong to conclude that agreements are likely to break down. On the contrary, because bargaining costs are assumed to be zero, ex post bargaining is always successful and the ex ante specific investments in human capital undertaken by the contracting parties are never (theoretically) wasted. It is this feature that points up the highly distinctive nature of modern property rights theory. A transactions-cost approach would tend to emphasise the use of governance mechanisms to make the breakdown of contract less likely. The property rights theorising sketched here does not investigate such issues because (perhaps somewhat paradoxically) contracts always survive ex post bargaining. In fact, it is usual to invoke the Nash solution to a cooperative two-person bargaining game which (assuming contractors have similar preferences) predicts that the gains from trade will be shared equally between the bargainers.24 Thus, our contractors not only predict that they will have to engage in ex post bargaining; they also predict what will be the outcome of this ex post bargaining. They each receive their ‘threat value’ plus one half of the gains available from a successful agreement. Their ex ante investments are then based upon these accurate forecasts, and the optimum assignment of ownership rights maximises the resulting total investment in performance and reliance. An algebraic presentation of the theory based upon the work of Hart (1995a) is presented in the notes at the end of this chapter.
Thus far the analysis has focused on the problem of deciding which of two contractors should hold the residual control rights to a single physical asset. If we now imagine that there are two assets involved, further conclusions can be drawn about how the rights in these will be distributed. The physical assets, to follow our example further, might be the building and kitchen equipment. It seems natural to expect the restaurateur will control both the building and the kitchen equipment. We would not anticipate the designer owning the kitchen equipment and the restaurateur the building. Why do we not think such an arrangement would be reasonable? The formal answer is once more based upon the willingness of the contractors to make ex ante transaction specific investments. The restaurateur will undoubtedly conclude that the control of the restaurant, in the event of termination of the agreement with the designer, will be of no avail unless he or she also has access to the kitchen equipment. Willingness to invest in reliance is not increased by control of just one of these assets; s/he requires control of both. Given that the designer will not increase his or her investment in performance if s/he controls the kitchen equipment, it is obvious that the chef should control both assets. They can be regarded as a single ‘composite’ asset, to be allocated on the principles discussed above. The two assets in this example are ‘strictly complementary’ and strictly complementary assets should be held together.
In contrast, we might imagine that the second asset consists of some specialised computer equipment used in design work. No one would expect the restaurateur rather than the designer to own this equipment, but can we explain this commonsense response more formally? In the event of the termination of their agreement, the restaurateur gains nothing from his or her control of design equipment. On the other hand, the designer might find it impossible to undertake the modifications necessary to take advantage of alternative opportunities without it. His or her ex ante investment in performance will be adversely affected if s/he does not control the design equipment. The chef’s ex ante investment in reliance is not affected either way. Thus, the designer should hold the residual control rights in the design equipment. The two assets are now ‘independent’ and independent assets should be held separately.
Coase’s theory of the firm saw internalisation as a means of reducing the transactions costs of using markets. Given, however, that incentives are still a problem within the firm and that moral hazard and other difficulties do not go away just because a transaction is conducted within the firm rather than outside, the precise source of any advantage to integration is left unspecified in the Coasian analysis. Durable, long-term agreements supported by reputation or indeed supported by a degree of monitoring are possible across ‘markets’ as well as in ‘firms’. As was seen in Chapter 2, Williamson (1985) emphasises the firm as a system of ‘governance’. In a world of specific assets and opportunism, the firm establishes structures for the governance of long-term incomplete contracts. For Williamson, these governance arrangements lower the cost of long-term contracting. Property rights theory also starts from the acceptance of contractual incompleteness, asset specificity and opportunistic post-contractual bargaining. The essence of the firm, however, is not that it is an effective governor of contracts but that it owns a set of physical assets which induce optimal ex ante investment in performance and reliance between contractual agents.
Source: Ricketts Martin (2002), The Economics of Business Enterprise: An Introduction to Economic Organisation and the Theory of the Firm, Edward Elgar Pub; 3rd edition.