In the last section we focused attention on the structure of property rights characteristic of the single proprietorship. Common observation tells us, however, that, numerous and economically important though such arrangements are (for example, Storey, 1982; Bolton, 1971), the modern economy has developed institutions of far greater complexity. An explanation of these more complex institutions must ultimately reflect the idea that concentrating property rights in a single holder is not necessarily the most efficient structure. Sometimes the sharing of rights between people or the apportioning of different private rights between people may be efficient. Thus the full package of property rights held by a proprietor may instead be shared between two or more people in a ‘partnership’, or alternatively the right to claim the residual may be shared between one group of people and the right to monitor the inputs may be held by another as in a ‘joint stock’ company.
At first sight such a statement appears to contradict flatly all that was argued in the previous section concerning the necessity of overcoming the moral hazard problem posed by shirking and the resulting concentration of property rights. It is evident, however, that the complete avoidance of all moral hazard problems is neither feasible nor efficient. If this were not so
it would be a simple matter to circumvent moral hazard by refraining from all contractual relations with other people and forgoing the benefits of division of labour and exchange. The maximum concern for fire prevention and theft prevention can be achieved no doubt by abolishing insurance markets, but few people would advocate such a move or claim that economic efficiency would be enhanced. It may be worth while tolerating reduced incentives if the benefits of a more efficient distribution of risk taking are sufficiently great. Conversely (and this is something we shall discuss in more detail in Chapter 5 on principal and agent) if a perfectly efficient distribution of risk taking involves severe problems of moral hazard it may be worth while sacrificing risk-sharing benefits in the interests of providing incentives. In other words we might expect observed contractual relations to reflect the available trade-off between risk-sharing benefits and effort incentives, a trade-off which will be affected in any given case by the costs of monitoring.
1. The Single Proprietor
Consider once more the single proprietor. A team endeavour requires a monitor if it is to operate effectively. The problem of providing incentives to the monitor is then encountered. This can be viewed as a classic ‘agency’ problem. If members of the team cannot tell whether the monitor is performing the promised services, they have to devise some incentive structure based not on unobservable behaviour but on observable outcomes. The one thing that, by assumption, is observable by everyone is the final output of the entire team. Instead of a system in which this output is shared between all members of the team therefore, a form of organisation evolves in which team members receive a fixed wage (irrespective of overall team performance) and the monitor receives the residual. In Chapter 5 we shall investigate more carefully the circumstances in which we expect an agent to promise a principal a fixed sum in this way and thus to relieve the latter of all risk. However, it is not offensive to the intuition to learn that if the agent is risk neutral and the principal is risk averse the efficient contract between them will involve the agent bearing the entire risk.
It is very important to understand the nature of the ‘thought experiment’ conducted in the above paragraph. The single proprietor is clearly not an agent in a legal sense. As we saw in an earlier section, the proprietor must be the employer, not merely the agent of the team, if he or she is to have the authority to influence team behaviour. However, in piecing together a rationale of the proprietorship, it is defensible and indeed enlightening as a first approximation to regard the structure of incentives embodied in this form of organisation as a solution to an agency problem. In a similar spirit, we will later on consider how far and in what circumstances an employee might be considered an agent of the manager or the manager an agent of the shareholder.
Let us suppose now that the moral hazard problem is so severe that in the absence of a monitor-employer the team could not survive in any shape or form. On the other hand, assume that the returns to monitoring activities are substantial so that a proprietorship is viable. If the proprietor is risk neutral and the employees are risk averse, the traditional structure with the proprietor receiving the residual and the employees a given wage will be efficient. Notice how akin to Knight’s is this conception with the ‘confident and venturesome’ providing insurance for the ‘doubtful and timid’. Alchian and Demsetz, however, explicitly reject the Knightian risk-sharing approach to the firm and prefer to concentrate instead on the advantages of team production. Yet any final organisational form, with its structure of property rights held by the members of the organisation, will presumably reflect all the forces which we have so far discussed:
- The potential advantages of further specialisation (division of labour) or team production.
- The extent to which the exchange relationships involved in point 1 above give rise to problems of moral hazard.
- The returns to monitoring.
- The trade-off between risk-sharing benefits and incentives.
The single proprietorship is a ‘solution’ to the problem of organisational form under rather special circumstances:
- The potential advantages of specialisation or of team operations are limited to groups sufficiently small to be efficiently monitored by a single person.
- Moral hazard problems are severe, but . . .
- The returns to monitoring are such that at least over a certain range the extra output of the monitored team is more than sufficient to compensate the monitor for his or her effort.
- Either (a) the returns to monitoring effort are certain, or (b) monitoring effort favourably affects the probability distribution of the residual by increasing expected output net of monitoring and contractual costs, and the monitor is risk neutral.
By setting out the conditions most favourable to the establishment of a single proprietorship in this way we begin to perceive the circumstances in which alternative organisational forms might be observed.
2. The Partnership
2.1. Monitoring costs
We start by assuming that returns to monitoring effort are certain. In Figure 4.2, the curve labelled MCA = MCB represents the marginal costs to monitors A and B of different levels of monitoring effort. It measures the extra monetary payment required to induce them to exert one more unit of effort. Curve MB1 indicates the marginal returns to extra effort. We suppose that after a certain point (E1) the marginal cost of monitoring effort rises and eventually becomes vertical as the limits of human endurance are reached. The returns to extra monitoring of the team decline throughout. A single proprietor (say person A) facing curve MB1 would put in effort level E1 where the extra returns from marginal effort are just equal to the compensation required. The final reward of the monitor will depend upon the level of effort required for the team to break even. Suppose for example, monitoring effort E0 is required if contractual payments to other team members are to be met from the value of output. All further monitoring effort will produce a residual which can be claimed by the monitor. Since we have assumed that returns to monitoring effort are certain, we would expect this residual will just be sufficient to compensate the monitor for the costs he or she incurs (including wages forgone as a team member).12 Thus, the additional benefit to monitoring effort above E0 (the value of the residual, areas A + C) will be just equal to the costs incurred in being a monitor (areas B+ C). Hence, under conditions of a certain return to monitoring effort, area A = area B.
Now suppose that the returns to monitoring effort are given by MB2 in Figure 4.3. Under these conditions a partnership of two people monitoring together will exert effort level 2E1 =E2. The marginal cost of effort at this point is the same for each partner, but we might ask why a single proprietor should not monitor the team and exert effort level E1‘? The answer is that competition from the partnership form of organisation will undermine the single proprietorship. Once more, assume that effort level E0 is required for the team to break even. It follows, as before, that where the returns to monitoring effort are certain, both partners will just be compensated for the costs of monitoring the team. Thus area A'(abc) = area B'(of aE0). But if area A’ equals area B’, a single proprietor exerting effort E1‘ would not be able to earn enough to compensate for the work involved. Clearly, the residual of the single proprietor will be bdE1‘E0 which will fall short of total monitoring costs (the total area under MCA up to Ej’) since area aedb<B’. The essential point is extremely obvious.
If a partnership can monitor at the same marginal cost as a proprietorship or lower up to E1‘ and can further afford to monitor to a higher level of intensity (E2), the partnership form of enterprise will take over from the proprietorship and the efficiency gains represented by area edc will be the prize.
The assumption that the monitoring costs faced by each individual are uninfluenced by the forms of organisation is, of course, crucial to this piece of analysis. A partnership involves an agreement between two or more people to perform certain monitoring services in exchange for a specified share in the residual. Even when the returns to monitoring effort are assumed to be certain, therefore, contractual difficulties are likely to be encountered. If the monitoring effort of each partner is perfectly and costlessly observable by all the others, partnership arrangements will be a predictable response to the increasing productivity of team effort and hence will permit larger team sizes than would otherwise be possible. Where, however, the behaviour of each monitor is costly to observe, Alchian and Demsetz’s moral hazard problem reasserts itself. The effort of one monitor (partner) confers benefits on the others and the result will be an incentive to shirk. We therefore deduce that partnerships are more likely to evolve where the process of monitoring is itself routine and susceptible to a degree of accountability than where the effort of each partner is almost impossible to observe or deduce. Where the returns to monitoring are certain, two partners should each be able to police the activity of the other since each will know their own effort and can deduce the effort of the other from the final team output.13 As the number of partners increases, however, the incentive to shirk will rise, since assigning individual responsibility for poor team performance may become impossible, and the effort of any individual partner will have a smaller and smaller effect on the value of his or her share.
Our discussion of the partnership, thus far, leads to the conclusion that sharing the right to the residual may hold out the possibility of potential efficiency gains by reducing the marginal cost of monitoring large teams. Against these potential gains we must set the extra problems of moral hazard which may arise when rights are shared in this way. If monitors begin to shirk, the efficiency gain (area edc) in Figure 4.3 may be dissipated and the single proprietorship will continue as the most effective organisational form. At any rate, partnerships are likely to be fairly small and, given the trust which must exist between partners if they are to avoid the losses involved in opportunistic behaviour, it is expected that the use of close family connections and those amenable to peer-group pressure will be frequent. A more extended discussion of profit sharing as an institutional form appears in Chapter 10.
2.2. Risk sharing
The assumption that returns to monitoring are certain is a convenient simplification when discussing the trade-off between monitoring costs and the hazards encountered when the residual is shared. As was noted earlier, however, the residual received by the monitor is unlikely to be deterministically related to monitoring effort. By accepting a residual reward, the monitor is exposed to risk. The fact that the residual may vary for reasons unrelated to monitoring effort has important implications.
- In the first place the suppliers of ‘contractual’ resources to the team effort will want assurances that, if the residual turns out to be negative, they will still receive the promised payment for services rendered. This implies that a monitor will require some personal wealth to act as ‘collateral security’. We discussed at some length in Chapter 3 the possibility that an entrepreneur with no wealth might persuade others to provide finance, but, accepting that this may occur, there are clear limits to the resource inputs which can be acquired in this way. Risk implies that the size of a single proprietorship will be limited by personal wealth and that partnerships will be necessary if firms are to grow beyond these limits.
It is worth emphasising that the above argument relies on the premise that the risk faced by the monitor is ‘uninsurable’. Even if the residual could be represented by a probability distribution conditional upon effort, and was thus ‘risky’ in Knight’s strict sense, insurance markets would succumb to the moral hazard problem if the monitor’s effort were not observable. If the monitor were certain of achieving a given residual through an insurance contract, the incentive to exert effort would be entirely lost, and with it the whole point of giving the monitor the residual claim.
- Where the monitor is risk neutral, the concentration of risk is efficient as we have seen. Where, however, the monitor is risk averse along with other members of the team it makes no more sense to concentrate all the risk on such a person than to insist that someone confronting rapidly rising marginal costs of effort should do all the work. By taking a partner, the risks of the enterprise are shared, and where both partners are risk averse, total risk-bearing costs will decline. Thus, just as we showed in subsection 5.2.1 that partnerships might permit lower monitoring costs, they might also permit lower risk-bearing costs. It is then the sum total of these two possible efficiency gains which must be set against the moral hazard problems arising from sharing a property right in the residual.
The history of the partnership form of enterprise illustrates the operation of these conflicting forces. Each partner, in addition to sharing in the residual, has rights to use and manage the resources of the team. The decisions of each can therefore bind the others, and the partners are responsible for all debts, whether or not as individuals they were personally involved in incurring them. Indeed, the English law of partnership developed the rule that each partner was liable ‘to his last shilling and acre’. In the face of this stringent legal background of unlimited liability, it is not surprising, as was noted earlier, that the property rights of each partner are not freely tradable. If a partner withdraws or dies, the partnership is broken and has to be reconstituted. Thus, it is very complicated and difficult for a partner to extricate his or her share of the resources from the business.
Because in the case of a partnership there is no single contractual agent but several agents capable of contracting on behalf of each other, the transactional difficulties involved in this type of enterprise are substantial. During the 1830s in the UK, when dissatisfaction with the law of partnership was growing, one of the major issues concerned the difficulty faced by a third party in suing a partnership or vice versa, and the difficulty involved in one partner suing another. Grievances between partners were particularly troublesome given the difficulties of acquiring information, and cases were reported ‘which were upwards of thirty years in the Court of Chancery’.14 The fact that the 1837 report on partnerships was particularly concerned ‘with regard to the difficulties which exist in suing and being sued where partners are numerous’ is indicative of the transactional problems encountered by large partnerships. These problems rapidly cancelled any risk-sharing or monitoring benefits available, and effectively placed a limit on the size to which a partnership could grow.15
3. The Joint-stock Company
Large-scale enterprises involving the cooperation of thousands and even hundreds of thousands of individuals would clearly not have evolved had the property rights structures characteristic of the single proprietorship and the partnership been the only possibilities available. The potential gains available from the monitoring of large teams – the ‘visible hand’ as Chandler (1977) has termed it – required a new structure of rights to emerge, a structure which did not expose the managers of large-scale enterprise to a degree of risk which they were not prepared to shoulder, and which permitted capital to be supplied by many people who would play no part in day-to-day business decisions. This particular combination of characteristics was impossible to achieve under the strict law of partnership. Capital could be borrowed, no doubt, at fixed interest from many people, but only at the cost of tolerating a very high ‘gearing’ or ‘leverage’ in the financial structure. Such high ratios of debt to proprietor’s or partners’ wealth would increase the risk of insolvency; a spectre made even more appalling by the provisions of unlimited liability which effectively meant personal ruin in the event of business failure on such a large scale. As we have seen, the alternative of growing through the addition of new partners was rendered unattractive because of the transactional difficulties involved and the enormous trust required in the integrity of other members of the partnership.
The joint-stock company developed as a response to these difficulties. For our purposes, there are three characteristics of great economic importance.
- A joint-stock company has a legal existence quite distinct from the people who comprise the company at any given point in time. People may come and go but, unlike the partnership, the company continues in existence. Further, as a separate legal entity, a joint-stock company can sue and be sued. This greatly simplifies contractual relations with third parties and helps to overcome some of the difficulties alluded to in the previous section.
- The shares of public companies are freely exchangeable. A market can therefore develop in these shared rights (the stock exchange) and it is a relatively costless exercise to buy or sell an interest in any particular company. Ekelund and Tollison (1980) argue that this ease of transferability was important in the early history of the development of the joint-stock form of enterprise. Lack of transferability would inhibit the most talented and qualified people gaining control of productive resources, which would instead remain in the same hands or in the same family for many years. Over the long run, the flexibility offered by joint- stock enterprises in reassigning property rights to more energetic people would give them an advantage over alternative institutional forms.16
- The third important characteristic of joint-stock enterprises is that the liability of shareholders is limited.17 With unlimited liability, people will naturally be chary of business associations involving people they do not know personally. With limited liability the prospect of subscribing relatively small amounts to an enterprise will be more tolerable, in the secure knowledge that the rest of a person’s fortune is not inevitably at hazard in the same enterprise. Perhaps a more important implication of limited liability than the effect on the willingness of people to supply finance (as we have seen they could always lend at fixed interest to other types of enterprise) is the willingness of managers to raise finance. For the directors of a joint-stock company are themselves liable only to the extent of the shares they hold in the company and indeed there is no legal requirement that they should hold any. With risk spread widely in this way, rising costs of riskbearing do not constrain the size of operations as severely as they do in a partnership or proprietorship.
It is sometimes said that the coming of limited liability and the joint-stock enterprise lowered the ‘cost’ of finance. This is a somewhat misleading way of thinking, however. When people supply finance, whether by loan or by buying shares, they are aware of the institutional arrangements prevailing and are unlikely to ask for or expect a lower return when dealing with a limited liability company than with other forms of enterprise. They will ‘pierce the veil of limited liability’ and may adjust upwards their required return to allow for any perceived adverse effect on managerial incentives.18 The advantage of limited liability is that even after these upward adjustments have been made, the possibilities opened up by large-scale operations may be more than adequate to compensate. Much depends here, however, on terminology. A single proprietor with unlimited liability would be expected to undertake fewer projects than would be the case after turning the enterprise into a limited company. Even if the available projects were identical in the two cases, the additional risks faced by a proprietor will induce him or her to apply a higher discount rate, and fewer of the projects will yield expected returns which exceed the ‘cost of capital’. Thus, ‘the cost of capital’ to the enterprise, interpreted in this way, is very likely to be lower in the limited company, but this is just another way of saying that riskbearing costs are lower to the decision-makers.
Although Ekelund and Tollison emphasise the transferability of shares as a crucial force in the origins of the corporation in the sixteenth and seventeenth centuries, by the mid-nineteenth century the risk-sharing characteristics of the corporate form with limited liability appear to be a more decisive consideration. Hannah (1983a, p. 23) reports that, in the UK, eighty per cent of joint-stock companies were private not public as late as 1914. Private companies are those which specifically restrict the right to transfer their shares while retaining the other characteristics of the joint- stock form, including limited liability. It is instructive to consider the possible reason for this popularity of the private company in the UK into the twentieth century.
In terms of property rights, the fundamental characteristic of the corporate form is that the ‘right to claim the residual’ is separated from the ‘right to monitor the inputs’. This ‘separation of ownership from control’ has certain transactional advantages, as we have seen, and it permits the development of a class of specialist managers, but it also confronts the problem of managerial incentives which was so central to our earlier discussion of team production. The suspicion that joint-stock enterprises would result in inefficient if not corrupt management has a long history. Adam Smith, for example, wrote that ‘negligence and profusion, therefore, must always prevail, more or less, in the management of the affairs of a joint stock company’.19 Such ‘negligence and profusion’ will not prevent the emergence of the joint-stock form if potential efficiency gains exist which are sufficient to compensate, but it is clear that the problem of managerial incentives is central to this form of enterprise. The use of managers from a restricted family circle, each with a considerable shareholding and with limited ability to dispose of their holding, as in a private company, can obviously be viewed as a response to the incentives problem. Even successful public companies at the turn of the twentieth century in the UK used management from the families which founded and built the firms during the nineteenth century; firms such as J. and P. Coats, Imperial Tobacco, and Watney Combe Reid. As Hannah (1983, p. 24) remarks, ‘while this solved a fundamental problem of the corporate economy – that of maintaining managerial efficiency while divorcing ownership from control – it did so more by avoiding the issue than by devising new techniques of incentive and control’. In the United States, on the other hand, the development of the corporate form occurred more rapidly than in the UK, and innovation in corporate structure was more advanced. Indeed Chandler (1977, 1990) attributes backwardness in the UK up to the 1940s to the influence of family management and the failure to develop sufficiently quickly a class of professional managers.20 How far the former was the cause of, rather than a rational response to, the latter is, however, a moot point.
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.