The most important result to be developed in this chapter can easily be stated: In a world in which competition is ideal, in a way to be described later, there is natural monopoly in a particular market if and only if a single firm can produce the desired output at lower cost than any combination of two or more firms. Natural monopoly is defined in terms of a single firm's efficiency relative to the efficiency of other combinations of firms in the industry. It is possible to test the proposition that a given firm is a natural monopoly.

Not all monopolies are natural. For example, monopoly may result from the control by a single firm of essential inputs into production, through trademarks or patents, or from the exclusive right to sell in a certain market. The essential characteristic of this form is that monopoly power is based on the inability of other firms to compete on an equal basis. Monopolies of this sort are generally transitory in nature and may also serve a useful social function, for example, when the granting of a patent gives an incentive for technological innovation. This form of monopoly, however, is not a natural monopoly.

Monopoly power may also result from unfair practices, such as predatory pricing, by one firm against its competitors, or from the formation of a cartel of several firms in a market.

Chapters 4 through 8 have developed a theory of natural monopoly in highly abstract terms. In this concluding chapter some of that theory will be applied in an examination of both the institutional and technological structure of the telecommunications industry. The function of this chapter is not to determine whether or not the telecommunications industry is a natural monopoly. Instead, the chapter will satisfy three more modest objectives: (1) The major elements of the theory will be reviewed in the more concrete and structured setting of the telecommunications industry. (2) Existing empirical studies of the industry will be evaluated. (3) Issues relevant to the future theoretical and applied research will be discussed.

Section 9.1 will consider the nature of demand in telecommunications markets. There are a number of important characteristics, some of which are shared by other industries and some that are unique to telecommunications. The demand for communication is inherently a two-party interdependent process. Furthermore, the demand is time dependent but varies over time, and the output of the industry is nonstorable. Finally, demand in telecommunications is a demand for access to the network and potential use, as well as a demand for actual transmission of messages. Each of these characteristics will be shown to have a specific effect on telecommunications costs and in particular on the subadditivity of costs.

In Section 9.2 telecommunications technology for transmission, switching, network planning, and network operation will be discussed.

This study was begun in the fall of 1978 following the publication in several major economics journals of results that cast the “theory of natural monopoly” in a new light. The traditional viewpoint had generally been that certain industries, primarily the regulated public utilities, might by nature be monopolies but that assuredly no theory was required to account for the natural monopoly status of any given industry.

Recently this viewpoint has been questioned by a number of sources. From a policy-making perspective the last decade has witnessed the transformation of many industries from a regime of pervasive regulation to one of substantial competition. From a theoretical perspective it has been observed that the proper definition of natural monopoly depends on the more elusive property known as subadditivity of costs, rather than on the simple and easily measured condition of economies of scale. Therefore it is no longer obvious, or self-evident, whether or not a given industry satisfies the conditions of natural monopoly.

The purpose of this book is to provide a comprehensive theory that will help those who wish to investigate natural monopoly from either a theoretical or a policy-making perspective. The mathematical level of presentation varies greatly from chapter to chapter, depending on the subject matter. However, at no point is the technical exposition more advanced than is necessary to convey the essential aspects of the theory.

This chapter continues the analysis of a contestable market that was begun in Chapter 7 by developing a model that uses the theory of noncooperative games. The analysis of such a model will provide new insights into three separate but related areas: (1) The determination of conditions under which a contestable market equilibrium exists will be useful in assessing the proper role of regulation in natural monopoly and natural oligopoly markets. (2) The noncooperative game theoretic framework is one method by which the theory of sustainability, as described in Chapter 5, can be made more rigorous and complete. (3) The noncooperative model will provide an alternative approach to the question of market stability, which was analyzed in a cooperative game theoretic framework in Chapter 6.

Section 8.1 contains a description of the economic structure of both the model and the basic concept of a noncooperative equilibrium. The market that will be studied consists of two or more firms that produce a single output and have identical U-shaped average cost functions. Firms are assumed to compete by simultaneously choosing prices. Outputs and profits to each firm are then determined as a function of the vector of prices chosen by each firm. A price vector is an equilibrium price if no firm can increase its profit by a unilateral change in its own price. One simple model of equilibrium is considered in Section 8.1 and rejected as unrealistic.

The purpose of this study is to define in substantial detail the conditions under which monopoly — production by a single firm — is a desirable form of market organization. In this chapter I will briefly describe many of the results that will be discussed at greater length later in the book.

The “theory of natural monopoly,” which will be developed in the following chapters, should be considered as a part of a larger theory of “market organization” or “industrial organization.” In the theory of industrial organization the concept of a “market” and of a “firm” in a market are fundamental. Although it will be assumed that the reader is familiar with the way in which these terms are used in economic theory, a brief discussion of each will help to lay the groundwork for the results that follow. Throughout this book the term “market” will refer to any collection of buyers and sellers and to the outputs that are produced and sold. A market is a competitive market if there are a large number of sellers and no seller is able to influence the market price by a unilateral change in output. Consequently, in a competitive market there is no strategic interaction among sellers. None of them can increase their profits by taking account of the behavior of other sellers.

A market is an “oligopoly” if the number of sellers is small, but greater than one, and if strategic interactions among sellers are important.

Many aspects of relational contract law can be illuminated by examination of a well-known case from outside the realm of contract law: Boomer v. Atlantic Cement. The facts of Boomer are simple. Atlantic Cement built a large cement plant which produced some pollutants creating a nuisance for some neighboring residents and businesses. The victims sought an injunction to prevent the nuisance. However, the court took into account that the magnitude of damages to the victims was only about $183,000 while if the victims received an injunction, they could conceivably force the cement plant, valued at over $40 million, to shut down.

Shutting down the plant would entail a great social cost and this could be avoided by denying the injunction. Of course, it is highly unlikely that the plant would be closed down even if the injunction were granted. For, if the plant were to close, the victims would receive only a modest benefit: the clean air that removing the nuisance would provide. The clean air might well be worth more to Boomer and friends than the $183,000 price tag assigned it by the court. Nevertheless, the victims could almost certainly do much better if they bargained with Atlantic, selling it the right to continue in business with the same level of pollution. If the plant has no other uses and little scrap value, the victims could conceivably obtain something close to the $40 million that the plant was valued at. The magnitude of the payment would de end upon the bargaining skill and the amount of resources devoted to the bargaining by the two sides.

… The typical case in which impossibility or some related doctrine is invoked is one where, by reason of an unforeseen or at least unprovided-for event, performance by one of the parties of his obligations under the contract has become so much more costly than he foresaw at the time the contract was made as to be uneconomical (that is, the costs of performance would be greater than the benefits). The performance promised may have been delivery of a particular cargo by a specified delivery date – but the ship is trapped in the Suez Canal because of a war between Israel and Egypt. Or it may have been a piano recital by Gina Bachauer – and she dies between the signing of the contract and the date of the recital. The law could in each case treat the failure to perform as a breach of contract, thereby in effect assigning to the promisor the risk that war, or death, would prevent performance (or render it uneconomical). Alternatively, invoking impossibility or some related notion, the law could treat the failure to perform as excusable and discharge the contract, thereby in effect assigning the risk to the promisee.

From the standpoint of economics – and disregarding, but only momentarily, administrative costs – discharge should be allowed where the promisee is the superior risk bearer; if the promisor is the superior risk bearer, nonperformance should be treated as a breach of contract. “Superior risk bearer” is to be understood here as the party that is the more efficient bearer of the particular risk in question, in the particular circumstances of the transaction.

Suppose that none of the work performed for a large firm required firmor job-specific skills. Further, assume that all the paperwork costs associated with labor turnover were nil. Even in these extreme circumstances there would still be good reason for the large firm to establish an elaborate governance structure for employees and for the employees to achieve considerable de facto job security.

To direct workers to perform certain tasks and to discourage behavior that impairs performance, the firm requires devices which impose costs on the worker for noncompliance. The ability to impose costs is enhanced by making quitting expensive for the worker. If the worker could simply walk away without cost, any particular punishment (say a suspension or fine) could be ignored; if, however, quitting imposed a substantial loss on the worker, he would be vulnerable to the threat of punishment and thus the deterrents become credible. Further, a high exit cost can be a powerful deterrent in its own right. The firm can use the threat of termination to influence the worker's behavior.

The firm has available a number of devices with which to penalize exit, or what amounts to the same thing, reward continuation. One device is to pay a premium wage (like Ford's five-dollar day), the sacrifice of that premium being a cost of leaving borne by the worker. Note that the premium is not paid “up front”; it is deferred so that the payment is contingent upon the worker's continued satisfactory (from the employer's viewpoint) performance. Deferral enables the firm to enforce the agreement without recourse to the expensive judicial system; if legal enforcement were free, then up-front payment would suffice.

In a discrete transaction, the parties need not rely upon performance by a particular trading partner. For many exchange relationships, however, that is not the case. Having entered into a contract with a particular supplier, a buyer will find that the costs of leaving this contract and dealing with an alternative supplier are high. The buyer's dependence on continued dealing with this supplier gives the supplier power over the buyer in the sense that the seller can threaten to impose costs on the buyer unless it acted in a certain way.

The first two selections are concerned with some issues regarding power within the contractual relationship. While Klein [6.1] focuses primarily on franchise contracts and I [6.2] focus on the employment relationship, much of the analysis is relevant to a broader class of contracts. The crucial point is that the power is not necessarily bad; rational parties often want to set up their relationship so that one party will be able to exercise power over the other. At the same time, they often want to utilize some devices for governing the exercise of that power. This might entail reliance on reputation with the trade, public enforcement, explicit contract terms, establishment of a private dispute resolution apparatus, or some combination of these.

Klein's central point is that it will often be the case that the most efficient arrangement will entail apparently unfair contracts. One party will be able to impose a substantial penalty on the other.

The lost-volume seller problem is a very confusing one that has been incorrectly analyzed by numerous commentators and has been a constant source of confusion to the courts. The issue is this: If a buyer breaches a purchase contract for a manufactured item and the seller subsequently resells the product at the same price, has the seller suffered any damage, and if so, should he be compensated for it? In cases in which the seller is a retailer, the conclusion is (a) yes, the seller does suffer damages, (b) the damages are the market price of the service of selling the goods, (c) the market price of selling is approximately the gross margin, (d) even though the damages are incurred, full compensation would probably be inefficient, and (e) the law ought to encourage the parties to use nonrefundable deposits as liquidated damages. This argument is developed in Selection [4.1], one of the few papers in this book that considers contracts between businessmen and consumers.

While that argument is of interest in its own right, it serves as a useful introduction to the lost-volume problem that arises in contracts between business firms. The line of argument is somewhat different, but the basic conclusion remains the same: Although the damages are real, the law should deny recovery and facilitate customized remedies via liquidated damages clauses. The treatment of the lost-volume problem in the nonretail case is explored in the notes and questions at the end of Part IV.

1. On May 15 shares of Acme Industries are selling at $16 per share. Smith agrees to deliver 1,000 shares to Jones on June 1 at $20. On June 1 Acme is selling at $30 per share and Smith reneges. By June 5 Acme is selling at $12. Jones sues for breach of contract asking for $10,000 – the difference between the contract price and the market price at the time of the breach. Smith argues that Jones might have held the shares rather than selling them immediately and that the market price of $12 should be used in determining damages; consequently, Smith argues, he should not have to pay anything. What is the appropriate measure of damage?

2. In the summer of 1973 X chartered Y's ship for seven years beginning in 1974 at a rental of $1 million per year. After the Yom Kippur War, the demand for shipping soared and Y refused to deliver the ship, renting the ship to someone else for a six-month charter at $3 million. X sued for breach of contract. Justice moves slowly, and by the time the matter reached trial, shipping was in the doldrums. Multiyear charters that were going for $4 million per year in 1974 were available for $200,000 in 1977 when the case finally came to trial. The defendant argued that it did the plaintiff a favor by breaching; since the plaintiff was better off following the breach, Y argued, there were no damages.

The first paper in this book appeared over two decades ago in a sociology journal. Nevertheless, Stewart Macaulay's paper on the use (and nonuse) of contracts by businessmen has had a considerable influence on economic scholarship. The paper provides a good picture of how businessmen view contracts and why contract language is often of little relevance in describing the behavior of contracting parties and influencing the resolution of disputes. It provides a commonsense backdrop for much of what comes later.

A distinguishing feature of this collection of readings is the focus on “relational exchange” as opposed to “discrete transactions.” These are analytical constructs, not categories for classifying existing contractual arrangements. The former concerns arrangements in which contracting parties are isolated, to some degree, from alternative trading partners and the outcomes depend in part upon the behavior of the parties during the life of the contract. The latter concerns exchange of commodities in thick markets (a lot of buyers and sellers); the fact that a buyer enters into a contract with a particular seller today does not give that seller any advantage or disadvantage vis à vis other sellers in subsequent dealings with that buyer. As we shall see, most of the interesting, and difficult, questions of contract law disappear in a world of discrete transactions. In Selection [1.2], I provide a brief introduction to the concept of relational exchange. This is followed by a discussion of “transactions costs,” a term I am somewhat uncomfortable with. The reasons for that discomfort are spelled out in [1.3].

1. Goebel v. Linn is discussed in three of the selections. Which do you find most persuasive?

1. 1.1 Suppose that Goebel accepted the modified contract until the end of the season. Instead of defaulting on the note, Goebel then sued the ice company for damages for breach of contract and argued that the second contract was simply the “cover” contract. Should it be able to recover the difference between the original contract price ($2.00) and the modified contract price ($3.50)? See Endiss v. Belle Isle Ice Co.

2. 1.2 The fact that the size of the ice crop would depend upon the weather is obvious and one would expect that ice contracts would reflect this. The twenty-five cent premium in the event of a small ice crop in Goebel's contract is one way to deal with this contingency. A second device would be to include a prorationing arrangement. Such a clause was used by at least one other seller of ice in that era. See Kemp v. The Knickerbocker Ice Co.

3. 1.3 Suppose that Goebel's contract did call for prorationing and that some buyers used their quotas to resell to those who were desperate for ice at prices in the $10- to $15-per-ton range.

4. […]

… Static efficiency considerations will generally require that contract modifications be enforced on the grounds that the immediate contracting parties perceive mutual gains from recontracting that cannot, at the time modification is proposed, be realized as fully by any alternative strategy. On the other hand, dynamic efficiency considerations focus on the longrun incentives for contracting parties at large imparted by a set of legal rules. In the modification context, these dynamic efficiency considerations adopt an ex ante perspective, rather than the ex post perspective implicit in the static efficiency considerations. Adopting the former perspective, rules that impose no constraints on recontracting may increase the overall costs of contracting by creating incentives for opportunistic behavior in cases where “holdup” possibilities arise during contract performance. As well, even where a genuine change has occurred in the economic environment of the contract between the time of formation and the time of modification such that, in the absence of modification, one party faces an increase in the costs of performance relative to expectations at the time of contract performance, allowing recontracting may facilitate the reallocation of initially efficiently assigned risks. This leads to moral hazard problems that may attenuate incentives for efficient risk minimization or risk insurance strategies by the party who subsequently seeks the modification. Thus, what is in the best interests of two particular contracting parties ex post contract formation when a modification is proposed and what is in the interests ex ante of contracting parties generally in terms of legally ordained incentives and constraints that minimize the overall costs of contracting may lead to divergent policy perspectives.

1. As Bishop notes, the foreseeability doctrine in contracts has little to do with the possible occurrence of low-probability events. It concerns the ability of the parties to control the magnitude of the damages that actually occur. Bishop emphasizes one device for doing so: providing notice of the special circumstances. An alternative, which he does not consider, is for the parties to arrange their affairs so that the magnitude of the damages would not be so great if the other party failed to perform. That is, in Hadley v. Baxendale the shipper could have reduced the expected cost of a breach by informing the carrier of the consequences of a failure to deliver or by carrying a greater inventory of shafts so that the plant would not have to remain idle in the event of a delay (breach) by the carrier.

The possibility that Hadley could have held a larger inventory of shafts (an input) has been widely recognized. Less attention has been given to other ways in which the costs of the failure could have been limited by timely effort on the part of Hadley. Hadley could have held a larger inventory of flour (the output). After the shaft was delivered and the mill again operating, Hadley could have made up for some of the lost output by running at a higher level of output than he otherwise would have. (In effect, that entails having a larger inventory of productive capacity – another input – than Hadley might otherwise carry.)