The Classical Capitalist Firm Revisited

Recall from chapter 4 that the classical capitalist firm is owned by one indi­vidual.⁵ The presumption is that the amount of capital involved is not so large that most of it must be raised in the capital markets.

The classical capitalist firm can be distinguished from other types of organizations by a distinctive pattern of interrelations among the set of functionally defined economic roles identified in the first section of chapter 4. Specifically, there is a single monitor (or a small and short hierarchy of monitors) who is the central contracting agent with laborers and all the other input providers (e.g., suppliers of raw materials and semifinished products). This presupposes that this monitor has the authority to negoti­ate the terms of these contracts, including the authority to terminate con­tractual relations between the firm and all input providers, including labor­ers. It also means that the monitor can determine, within a more or less broad zone of acceptance, how all inputs are to be used or combined in the production process.

In the classical capitalist firm, this monitor is also the sole residual claimant on the firm’s income stream and is responsible for determining the product to be produced, its characteristics, and the price at which it will be offered for sale. Because of his control over the output side, as well as the input side, of the firm-market interface, this monitor-residual claimant is the firm’s entrepreneur.

The Single Monitor

Recall that a monitor is needed when team production takes place. This is production in which several cooperating factors are used to produce a product that is not the sum of separable outputs of each factor (Alchian and Demsetz 1972, 778). As a result, the contribution of individual factors is dif­ficult to ascertain. This measurement problem make it possible for a factor provider who shirks or otherwise opportunistically reduces output to get more than the value of what he contributes.

To see why, suppose that there is no monitor and the provider of factor F₁ opportunistically reduces her contribution (input) and thus negatively affects the team’s output, whereas the providers of the other factors of pro­duction, F₂... F_n do not. For example, F₁ might be labor and its provider might be a shirker. Supposing that team output is priced at the approxi­mate marginal cost of production, it follows that the providers of these other factors are having some of the value of their assets appropriated by the shirking provider of F₁. Because the value of these other assets in non­team production is much less, they are specialized assets; thus, some of their value constitutes quasi-rents. Under these circumstances, the provider of F₁ can be said to be opportunistically appropriating some of the quasi­rents of other factors of production.⁶ Why would these other input providers put up with this arrangement? Presumably, they would not, if they had somewhere else to go.

Notice that the problem is stated in a completely general form so that it applies to all factor providers and not just to those who provide labor ser­vices. If those who provide capital or raw materials did so in exchange for a share of the output, they would also be subject to an expropriation hazard. And, depending on the nature of their contributions, they might also be in a position to reduce output by acting opportunistically (e.g., by short­weighting deliveries). One solution to the problem (perhaps the most obvi­ous one) is for one individual to take on the role of monitor of team pro­duction. The monitor checks up on each member’s provision of inputs to detect and deter shirking and other forms of opportunism. But it is, in fact, not so obvious that one individual should take on this role. Though some monitoring might be needed, it is not directly evident that one person should do it. There are other possibilities to be considered.

Clearly, it would be inefficient for everyone to monitor everyone else, since that would involve an enormous duplication of effort. But this is not the only way to involve everyone in monitoring. Each individual could devote part of his or her workday (or workweek) to monitoring the team. Or, team members could rotate in and out of the monitor’s job, so that everyone gets an opportunity to crack the monitor’s whip, so to speak. The first alternative is generally inefficient, because the activities involved in monitoring are, in general, more difficult to monitor than the other activi­ties involved in team production. If an agent’s job consists of two tasks, one of which is easier to monitor than the other, then, all else equal, she will devote more effort to the task that is easier to monitor (Holmstrom and Milgrom 1990).

The other way of involving everyone in monitoring would be to have one person to act as monitor in a full-time capacity but to rotate each mem­ber of the team in and out of this role on a regular basis. Generally speak­ing, however, this would be inferior to having only one person—or a small and short hierarchy of persons—specialize in the task of monitoring. To see why this is so, recall what the monitor does. According to Alchian and Demsetz, monitoring consists in “measuring output performance, appor­tioning rewards, observing the input behavior of inputs as means of detect­ing or estimating their marginal productivity and giving assignments or instructions in what to do and how to do it” (1972, 782). Doing well at these tasks is a matter of talent, tastes, and training; not everyone is equal in the first two and not everyone is equally disposed to undergo and profit from the third.

What the monitor does is prevent or minimize shirking and other forms of opportunism by other input providers. One way to do this in the case of labor services is to redesign the task (or even the product itself) to make the work more intrinsically rewarding, thereby making other forms of monitor­ing less necessary. There are, however, limits to what can be done to pre­vent shirking and other forms of opportunism in this way. Often, sterner measures are called for—in particular, the other elements of monitoring mentioned by Alchian and Demsetz: directly measuring output to the extent possible, metering inputs, and apportioning rewards and penalties as a result of other forms of monitoring.

These activities generally require those who do it to look upon others as potential shirkers and opportunists who have to be kept in line for their own good and the good of the team. Not surprisingly, there is a kind of adverse selection problem associated with this in that those who excel at these forms of monitoring often have other disagreeable characteristics. At the minimum, they tend to regard others as more given to shirking and opportunism than they really are. By contrast, kindly souls who are predis­posed to think well of their fellows tend not to do as well at metering inputs, apportioning rewards and penalties, and the like. They systematical­ly underestimate people’s penchant for opportunism, and are reluctant to impose serious penalties on those who act opportunistically. However, suc­cessful monitoring tends to result in people’s getting paid approximately the value of what they contribute. For this reason, good monitoring is essential to prevent the exploitation of some input providers by others. If everyone gets the opportunity to serve as monitor, the opportunists in the team will periodically get their golden opportunity to siphon off some of the value of other team members’ contributions.

For all these reasons, the most efficient arrangement is to have one indi­vidual specialize as monitor of team production.⁷ This arrangement is not limited to the classical capitalist firm. As Putterman says, “In the most egali­tarian of producer cooperatives, the kibbutz, each work branch has its head, who supervises and tries to assure the effective work performance of its members. A relatively conventional supervisory structure marks many col­lective enterprises, such as those of the Mondragon network of cooperatives in Northeastern Spain” (1984, 173). Within the broad parameters suggested by Alchian and Demsetz ^,s definition of monitoring, the rights and privi­leges associated with the monitor’s role may vary from one type of firm to another.

The Monitor as Central ContractingAgent

The central contracting agent is the person in the firm who enters into con­tracts with suppliers of inputs (laborers, suppliers of raw materials and semifinished products, etc.) on behalf of the ultimate decision maker. Why would the monitor occupy this position of central contracting agent? Notice that the various tasks the monitor must perform presuppose that he has the authority to decide how inputs are to be used or combined, at least within a broad zone of acceptance. In other words, though there may be side con­straints on what he may do with the assets to be used in production (labor, leased capital goods, and other inputs), the monitor has the authority to determine how inputs shall be used. The monitor cannot assign or design tasks, meter inputs, and so on without having that authority. The need for this basically open-ended authority comes from the fact that contracts with input providers governing every contingency cannot be written, and the costs of continually renegotiating contracts in response to unforeseen con­tingencies is prohibitive. This is perhaps the primary efficiency advantage of hierarchy, an advantage first noticed by Coase (1937). Grossman and Hart call this authority “residual rights of control” (1986, 717).⁸ Because of the monitor’s role in directing and regulating the flow of inputs, the con­tract between the input provider and the firm is effectively or essentially between the former and the monitor. Having these residual rights of con­trol is even more essential if (as shall be argued shortly) the monitor also decides on the product to be produced and its characteristics. The general point is that the role of monitor effectively presupposes that the monitor is also the central contracting agent with all the other input providers.

Must the monitor have the authority to fire input—providers, in partic­ular, the workers? It would seem so, if she is to exercise real authority in carrying out her tasks. This claim, as well as the monitor’s authority, is sub­ject to an important qualification, however. The zone of acceptance within which the monitor has the authority to fire input providers may be more or less narrowly defined. The state and collective bargaining agreements have narrowed that zone considerably over the past century. This erodes but does not usurp the monitor’s authority to fire. This can best be appreciated by contrast with the situation in earlier times. In the good old days of capi­talism—indeed, up until a couple of decades ago—the boss could fire workers for the most trivial and insubstantial reasons. Now many workers have protections so elaborate and extensive that they appear weak and ten­uous only to civil service bureaucrats and tenured professors. Nevertheless, these protections have the form of side constraints in that they detail what a person cannot be fired for. They do not reassign what might be called “the residual right to fire.” That right remains with the monitor, as it always has. The transfer of that right to the workers (as a collective) or to the state would be an important step toward socialism, which is why this measure is generally favored by socialists and opposed by defenders of a free enter­prise system.

The monitor-central contracting agent’s residual rights of control over inputs (whether labor or nonlabor) is what prevents input providers from forcing a renegotiation of their contracts when unforeseen contingencies arise. In other words, this authority relationship between the monitor-cen­tral contracting agent and the input providers prevents the latter from “holding up” the firm and/or its members by reopening negotiations in response to contingencies not explicitly covered in the contract (e.g., assigning one employee to do another’s job when the latter is sick).⁹

Transactions cost analysis uses an iterated version of this explanation to explain decision-making hierarchies within firms generally (and not just the classical capitalist firms). Hierarchies economize on decision-making costs, if only because the agreements required by more consensual arrange­ments take time and other resources to achieve. This is not to deny that there may be offsetting efficiencies to more consensual arrangements; in a free enterprise system one would expect to see—and one does, in fact, see—more democratic methods of decision making in some circumstances and over some range of decisions. The militarylike hierarchies of nine­teenth-century industrial capitalism have not disappeared entirely, but this is certainly not the norm in the late twentieth century, which has witnessed a proliferation of experimentation with alternative decision-making processes and structures.

Do transactions cost efficiencies wholly explain why hierarchies exist? Perhaps not. An alternative explanation, much favored by radical political economists, is that hierarchy serves the interests of power and that is why hierarchy in its many forms persists.¹⁰ This explanation is often inferred from the fact that there are no production cost (i.e., technological) efficien­cies to hierarchy (Marglin 1974, 46). This inference overlooks the possibili­ty that there might be transactions cost efficiencies that attend hierarchy. Moreover, power has also proved to be a difficult concept to define and operationalize. Those problems to one side, however, the main difficulty with citing power as the sole or predominant reason why hierarchies exist is that there is no evidence to suggest that less efficient modes of organiza­tion that concentrate power win out—or would win out—over more effi­cient modes that disperse power more evenly (Williamson 1985, 231 and chap. 10). That is exactly the sort of evidence that would be needed to dis­credit transactions cost efficiency explanations at the expense of power explanations.

Power might be part of the explanation for the origins or genesis of the classical capitalist firm. Given the concentration of roles in one individual, that has a certain measure of prima facie plausibility. However, absent the kind of evidence just alluded to, there is no good reason to believe that hierarchies exist in the classical capitalist firm primarily because they con­centrate power in the hands of some individuals (viz., classical capitalists) at the expense of others.

The Monitor-Central ContractingAgent as Residual Claimant

The discussion in the first subsection explained the need for a monitor to prevent the exploitation of input providers by other input providers and the advantages of having just one monitor (or a small and short hierarchy of monitors), and the preceding subsection explained the efficiency advan­tages of the monitor’s being the central contracting agent with other input suppliers. But what are the efficiency advantages of this individual’s being the residual claimant? As was pointed out in chapter 4, the explanation for this starts from the fact that monitoring is itself difficult to monitor. Task (re)design requires specialized knowledge of the production process and creativity; the deployment of both is hard to assess from the outside. The same is true of the design and implementation of more old-fashioned metering strategies. Not only are the inputs to the various monitoring modalities difficult for outsiders to judge, but so are the outputs. What does a well-monitored team look like? If the monitor lacks residual claimancy status, then she herself has an opportunity to shirk in the provision of mon­itoring services, which, all else equal, will lead to greater shirking and other forms of opportunism among all other input providers. In short, there will be a failure of leadership.

Another advantage to making the monitor-central contracting agent the residual claimant is that it gives her a better incentive not to expend too many resources on monitoring or on inputs generally. It is obvious that a central contracting agent will economize on production costs if she has residual claimant status. The same holds true of monitoring costs. Some forms of opportunism are simply not cost-effective to prevent. If the moni­tor-central contracting agent has residual claimancy status, she has to con­sider the costs of monitoring more carefully than she otherwise would. On the other hand, if this individual were to lack residual claimancy status, then she would have an incentive to expend more resources on monitoring activities (including task redesign) than is warranted, especially if those expenditures can be buried in such a way that whoever employs her has trouble identifying them.

By way of contrast, if monitoring were constituted by only intrinsically desirable tasks and consumed few resources, it would be unnecessary to make the monitor-central contracting agent the residual claimant. Though some elements of monitoring, (e.g., task design or redesign) might be inherently interesting, other aspects of monitoring (e.g., metering inputs and outputs, apportioning rewards) are not. The intrinsically interesting tasks could be farmed out to whoever is otherwise best suited to perform them. Indeed, this is done in many firms in advanced contemporary free enterprise systems (e.g., in Japan, the United States, and parts of Europe) where workers’ input is sought in the redesign of their tasks.

But the other tasks that the monitor performs are often positively unpleasant for most people, and carrying them out makes the monitor an unpopular figure. These are generally not farmed out in private firms in existing free enterprise systems—perhaps because the temptation to shirk in this role would be overwhelming for most people.¹¹ (By contrast, casual observation suggests that shirking in the provision of monitoring services is pandemic in public enterprises.) In addition, those who delight in these forms of monitoring are inclined to do too much of it, which is counterpro­ductive. Unlike most other tasks, one does not want the monitor to be someone who really enjoys her work. The best monitors are those who see much of what they do as necessary but disagreeable. Making the monitor a residual claimant eliminates the need for monitoring the monitor without at the same time creating an adverse selection problem that would attend hiring those who relish the opportunity to crack the whip.

The same sort of considerations explain why the central contracting agent should be the residual claimant. The central contracting agent exer­cises significant entrepreneurship in searching out the best and least costly input suppliers, considering alternative methods of organizing production, and so on. These activities, like the other forms of entrepreneurship to be discussed shortly, are difficult to monitor and are, for that reason, best han­dled by someone with residual claimant status.

A heretofore unnoticed consequence of making the monitor-central contracting agent the sole or primary residual claimant is that it facilitates the development of regular markets in other inputs. If every factor provider involved in team production were a residual claimant, the returns to a given factor of production would vary considerably across firms. Not only does this subject especially risk-averse input providers to risks that they would be willing to trade for a more definite income, but it also impedes the functioning of the market as a provider of information about the value of inputs. Returns to a factor owner would reflect not only the scarcity value of his factor of production but would also include an element traceable to his entrepreneurship, such as it is, as well as a purely stochastic element. This impedes the development of relatively stable prices in the factor markets, which, in turn, makes the payoffs of different production decisions even more uncertain than they would otherwise be.

The Monitor-Central ContractingAgent-Residual Claimant as Director of the Firm’s Product

The high cost of monitoring is also a reason why it more efficient for the individual who is the monitor, central contracting agent, and residual claimant to exercise the crucial entrepreneurial functions of choosing a product and its characteristics and making the strategic decisions on mar­keting and pricing. Once again, specialized and costly knowledge is required to evaluate the quality and amount of effort put forth. In addi­tion, as Barzel points out, luck has an imponderable influence on entrepre­neurial success, which further aggravates the difficulty of monitoring entre­preneurial activity (1987, 104). A salaried director of the firm’s product would have an opportunity to exploit her employer by shirking or by engaging in other forms of opportunism (e.g., entrepreneurship “on the side”). This opportunity can best be foreclosed by making her the residual claimant.

Finally, this arrangement allows for better coordination between the input and output firm-market interfaces. The central contracting agent is responsible for the input interface between the firm and the market; her task is easier to discharge if she is responsible for—or is directly in charge of whoever is responsible for—the output interface. The latter includes decisions on product, pricing, and marketing. In this way, the major entre­preneurial tasks are concentrated in one person.

The Monitor-Central ContractingAgent-Residual Claimant-Product Director as the Primary Provider of Capital

In chapter 4, it was noted that this arrangement, in contrast to an arrange­ment in which all capital is borrowed, serves the same function as good monitoring. Since the entrepreneur risks her own capital, she will pursue projects that maximize expected yield instead of excessively high-risk, high-yield ventures that she would be inclined to pursue if it were only other people’s capital at stake.

But how does this preclude exploitation? If the venture succeeds, out­side providers of capital would get the return on their investment that they were promised, and the entrepreneur would get the residuals. Supposing that they were promised the going rate for competitively efficient projects, they would be getting the value of what they contribute. No exploitation there: the problem comes if and when the venture fails. Not only can a salaried manager exploit her employer by doing less than she represents herself as doing in the provision of entrepreneurial services, but she is usu­ally in a position to appropriate more than just the quasi-rents of the assets if a project is failing. In the real world, by the time sheriffs with padlocks show up, often everything not bolted down has somehow disappeared. This is a natural consequence of the physical control of the firm’s assets exercised by the firm’s head, coupled with a penchant for opportunism on the part of the latter. On the other hand, if this individual provides most of the capital herself, this potential moral hazard is averted.

There may be other transactions cost efficiencies to the practice of this individual’s providing most or all of the capital. Barzel points out that this practice gives assurance to other factor providers that when business condi­tions appear to take a turn for the worse (whether or not they really have), the entrepreneur will not immediately terminate their contracts and hire or rehire factors at a lower rate on the spot market (another version of the holdup problem): “Employed factors will feel more secure if their contracts are structured so that an employer who fails to pay for the use of other assets stands to lose, which will happen... if a commensurate amount of the employer’s own capital is idled when he lays off other factors” (1987, 114). By pledging to idle substantial resources of his own, the firm’s owner assures these other factor providers that they will not be exploited in this way. Relatedly, if the central contracting agent has his own assets tied up in the project, it is much easier to get redress through the courts. An entre­preneur with no assets of his own at stake is what lawyers call “judgment­proof.” Of course, this arrangement does not preclude bluffing or simply stonewalling, which, if successful, would permit the entrepreneur to appro­priate some of the quasi-rents of others (i.e., to exploit them). Bluffing is not a stable strategy over the long haul, however, and reputation effects would likely mitigate this species of exploitation over time.

The Monitor-Central ContractingAgent-Residual Claimant-Product Director-Primary Provider of Capital as Ultimate Decision-making Authority

There is one final role to take account of—ultimate decision maker. The person who fulfills the role of ultimate decision maker has final say over the disposition of the firm’s assets. It is this individual in whose interests the firm is operated. There are clear advantages to concentrating all of the hard-to-monitor roles (i.e., monitor, central contracting agent, product director) in this individual. Complete, costlessly enforceable contracts can­not be written, which means that separating any of these roles from ulti­mate decision-making authority makes it possible for those who occupy these roles to get paid more or less than the true value of their services and, when the one party has nowhere else to go, to exploit or be exploited. Making the ultimate decision maker the primary supplier of capital gives him something to lose if he exercises his authority poorly, and making him residual claimant gives him something to gain if he exercises his authority wisely.

The Right OfAlienability and Liability to Execution of Debt

The final distinctive feature of the classical capitalist firm is that this indi­vidual who occupies all of these other roles can alienate any and all of the rights associated with these roles by selling them (usually as a package, for the reasons indicated in the preceding subsections). In addition, as owner of the firm, that individual is fully liable for execution of the firm’s debts. The most important implication of this multiple-role occupier’s also being able to alienate (i.e., sell) these roles and being liable for the firm’s debts is that—externalities to one side—the economic consequences of all the deci­sions made in these roles accrue to, or are capitalized into, the value of the firm and thus fall squarely on the shoulders of the individual who makes them. If some or all of these rights are inalienable and held only temporari­ly, then the decision maker does not have to take account of the conse­quences that his decision will have on the long-term value of the rights in question. Suppose, for example, that the boss could not sell his right of residual claimancy but would, instead, have to give it up without compen­sation when he leaves the firm. Under these circumstances, he could enter into a venture that would generate short-term gains but net losses over the long term—losses that could be incurred after he leaves the firm and which would be borne by the owners of specialized assets (including labor) that are locked into the firm for the long term. The right of alienability and the liability to execution of debt together encourage responsible use of assets by the boss, since this right and the corresponding liability result in his bearing the full economic consequences of his actions.