Introduction

In this appendix we review various theoretical models that have been proposed in order to reproduce some of the empirically observed properties of real networks. We consider only the models that focus on the local topological properties, in particular (in the language of Appendix A) on the first- and second-order properties. As a result, the higher-order properties of the networks generated by the models considered here are the result of local rules alone. Nonetheless, suitable local rules are often enough in order to reproduce most of the observed complexity of real networks. Moreover, it is believed that most real networks are indeed shaped by local rules alone, as higher-order mechanisms requiring the knowledge of the entire network are in most cases unfeasible.

The models presented here share a common aspect: the deviation of real networks from regular graphs is modelled through the introduction of some ‘disorder’ according to suitable stochastic rules. All the models described below (and largely most models in the literature) are therefore stochastic models. As a consequence they are also ensemble models, since they define a whole set of possible realizations of a network, rather than a single graph. Ensemble averages give the expected value of any topological property. They will be denoted by angular brackets 〈…〉 to avoid confusion with averages over the vertices of a single graph, which are instead denoted by a bar as in Appendix A.

Introduction

Biological systems react to changes in the surrounding environment by adjustment of their properties and functioning. The simplest cases include the capacity of prokaryotes to change the expression levels of specific proteins, as well as their distance from the source of chemical substances. In eukaryotes and multi-cellular organisms, the property of monitoring the environmental conditions and responding to their transformations has attained levels of particular complexity, through the development and evolution of means of supporting the communication among separate districts within the same organism, and among different organisms as well. Two general types of communication are classically described in biological systems. Neuronal communication is the first of them, comprising the networks of fibres connecting the different parts of organisms. Another kind of communication in living systems takes place by chemical signals that are produced and released from some cell sources, diffuse in the environment surrounding the emitting system, being it a liquid or air, and eventually reach the target cells. Two major features distinguish neuronal and chemical signalling: the means supporting the signals and the distance between the source and the target of signals. In neuronal communication the signal is mechanically supported by individual nerve fibres and travels distances related to the size of the organism about, being up to 10 m. The distance between the source of the chemical signal and its target, in contrast, is not limited by the existence of a physical wire connecting the emitting source and its target.

Economic life is critically dependent on a wide variety of privately owned networks: for communications (the Internet and broadband data, telephone, broadcast television and radio, cable television), for transportation (airlines, railroads, buses, trucks, shipping, electric power transmission and distribution, natural gas and petroleum pipelines), and for distribution of products (wholesale trade and postal services). Although different types of networks vary significantly with respect to a wide variety of technical details, they share certain technical and economic features that are critical to understanding network access.

This chapter introduces some basic aspects of network structure and design. The focus of our discussion is on physical networks, involving facilities such as telecommunications switching equipment and transmission lines. The purpose is not only to provide a technical background to our study but also to obtain some important insights into how networks are organized. Evaluating the effects of regulation on the organization or formation of networks is central to our public policy analysis.

This chapter also introduces a classification scheme that is useful for understanding the different types of access to networks. These types of access have generated a variety of public policy actions. Classifying the types of access provides insights into inconsistencies in public policies that can create economic inefficiencies. Our classification scheme provides the basis for the development of a new Coasian theory of communications networks.

Network designers choose the capacity necessary to provide the desired level of network services. The quality of service also determines the capacity required to deliver the specified quantity of network services. Because the architecture of the network is optimized by the network designers, the level of network capacity determines the underlying configuration. Different configurations are likely to accompany different capacity levels.

Choosing the aggregate level of services determines how much capacity to build into each node and link and the architecture of the network. In a telecommunications network, the capacity of switches and lines and the design of the network create constraints on the volume and type of traffic that the network can handle. Capacity usage might take the form of the number of calls and the time spent on each call in a circuit-switched network. In a packet-switched data network, capacity usage is measured by more complex definitions of bandwidth.

Pricing, that is, the selection of prices by decision makers, serves two general functions. First, prices ration consumer demand, because consumers reduce their capacity usage in response to higher prices. Second, prices stimulate capacity supply, because suppliers increase capacity offered for sale in response to higher prices. Prices implicitly communicate the scarcity of network capacity to consumers and also signal the strength of demand to suppliers. At a market equilibrium, prices tend to reflect consumer demand for network services and the supply of network capacity by firms.

Telecommunications has moved from traditional analog voice service to high-speed digital transmission, referred to as broadband. The explosion of new forms of transmission and new types of content (data, audio, video), has changed the debate over access to networks. Under the general umbrella of “network neutrality,” advocates of mandatory access argue for extensive regulation of prices and products.

The Supreme Court's recent decision in National Cable & Telecommunications Association v. Brand X Internet Services cleared the way for the FCC to resolve how to fit the leading broadband technologies, such as cable modems and digital subscriber line (DSL) services, into the existing regulatory regime (545 U.S. 967 (2005)). The fact that Congress largely failed to take the Internet into consideration when enacting the Telecommunications Act of 1996 has left policymakers with little definitive guidance. Regulatory decisionmaking is further complicated by the dynamism of the technological environment. The demands that end users are placing on the Internet are changing rapidly, as evidenced by the increasing popularity of bandwidth-intensive applications, such as streaming media, peer-to-peer downloads, and virtual worlds. In the meantime, a host of new communications platforms are waiting in the wings, such as third-generation mobile communications devices (3G) and wireless hotspots employing WiFi technology.

The emergence of platform competition and the shift from rate regulation to access regulation compel a different constitutional analysis of regulation. Because rate regulation simply adjusts the terms under which parties can contract, it represents the type of nonpossessory regulation traditionally subjected to a rather permissive standard of review under the Takings Clause. This standard requires only that the rate fall within a zone of reasonableness. Compelling access to a physical network, in contrast, invariably requires the network owner to permit third parties to locate equipment on its property. As such, access regulations are subject to the more restrictive standards associated with the Court's physical takings jurisprudence. Unlike nonpossessory regulations, in which reductions in the value of property are not necessarily compensable, physical takings necessarily command market-value compensation. Principles of constitutional law thus reinforce the basic economic conclusion that network access should be priced at market levels.

The Distinction between Physical and Nonpossessory Takings

The Emergence of the Nonpossessory Takings Doctrine

Initially, the Takings Clause was believed to protect only against direct government appropriations of private property or invasions that effectively divested the owner of possession as though the government formally condemned the property. Government actions that merely reduced property value did not qualify as takings. The Court subsequently recognized two types of takings that can arise without a physical occupation. First, the Court acknowledged that a rate regulation may effect a taking if the rate is set so low as to be confiscatory.

Antitrust courts have played a pivotal role in shaping the telecommunications industry, most significantly through the private antitrust suits brought against AT&T by MCI and the predecessor to Sprint in the 1970s and the case brought by the federal government against AT&T that culminated with the breakup of AT&T in 1984. All three of these cases relied upon the essential facilities doctrine, which requires owners of bottleneck elements unavailable elsewhere in the marketplace to make those elements available to competitors on reasonable terms.

The essential facilities doctrine has proven quite controversial, inspiring a welter of largely critical commentary challenging its conceptual validity, its administrability, and the extent to which it actually benefits consumers. After years of signaling ambivalence about the doctrine, the Supreme Court once again returned to the subject in Verizon Communications Inc. v. Law Offices of Curtis V. Trinko, LLP. Although the Court found “no need either to recognize…or to repudiate” the doctrine, it endorsed many of the criticisms advanced in the commentary (540 U.S. 398, 411 (2004)).

Most striking is the language in Trinko questioning the antitrust courts' institutional competence to implement access remedies. Since the opinion was issued, courts and commentators have struggled to determine how broadly to read this language. Some have suggested that the presence of a regulatory regime leaves no role for antitrust courts. Others have read Trinko more narrowly, arguing that the scope of judicial authority depends on a more nuanced assessment of the nature of the regulatory regime.

On the afternoon of August 14, 2003, a major power blackout struck the northeastern and midwestern United States and parts of Ontario, Canada. The power outage cascaded across cities and states, forcing businesses to close, cutting power to households, and shutting down communications networks, affecting over 50 million people. The outage stranded commuters with traffic jams, closed bridges and tunnels, and stalled public transportation systems. The problem began in Ohio when three high-voltage transmission lines of First Energy Corporation made contact with trees. Because of malfunctioning control-room equipment at First Energy, the utility failed to detect the problem and to warn other utilities. The downed lines caused power to overload the grid, which in turn led to further breakdowns in the transmission network. The Midwest Independent System Operator and the reliability control area for Pennsylvania, New Jersey, and Maryland (the PJM Interconnection) lacked the information and procedures to address the transmission problems.

The dramatic story of the Northeast power outage illustrates the complexities of networks – both in balancing the supply and demand for network capacity and in coordinating the interconnection of networks. Monitoring the performance of a single network is hard enough, and the transaction costs of coordinating network interactions are even higher.

Although communications networks face different types of problems than electric power networks, they also involve highly complex problems of network management and interconnection. Regulatory efforts to mandate access to communications networks can substantially increase this complexity.

The performance of network industries is determined not just by their underlying economics, but also by the manner in which they are regulated. The chapters in this part analyze the impact of regulation on network performance. This chapter begins the analysis by laying out the sources of market failure traditionally thought to justify widespread regulation of the telecommunications industry. It continues by reviewing the regulatory instruments employed to address those source of market failure. It then critiques the validity of those justifications, finding that they are generally based on economic theories and factual premises that are no longer regarded as viable and fail to take into account the full complexity of network behavior.

Chapter 5 offers a more complete description of the five-part system for classifying types of access mentioned briefly in the Introduction and analyzes the impact of each type of access on network capacity, configuration, and transaction costs. Chapter 6 examines the insights that graph theory can provide into the pricing of network access and uses that framework to critique existing approaches to establishing access prices. Chapter 7 considers the constitutional limitations on access imposed by the Fifth Amendment to the U.S. Constitution, which prohibits the taking of property for public use without the payment of just compensation.

Networks in communications, transportation, and distribution are fundamental features of the modern economy. Network industries are a major source of economic growth and a key component of economic development. The expansion of telecommunications, the Internet, and mobile communications has accompanied innovations in information technology. Although electronic commerce was once viewed as a specific category of business transactions, practically all business activity has come to depend on digital communications networks.

Public policymakers are keenly aware of the critical role of telecommunications in the economy. A great shift in regulatory focus has occurred, from more traditional utility regulation to access mandates. Examining changes in public policy in telecommunications requires a more fundamental understanding of the structure and function of networks. Expansion of the economics and law literatures on networks has accompanied economic and technological developments. The purpose of this book is to evaluate the implications of this public policy shift and to achieve a better understanding of public policy toward networks.

To achieve these objectives, the book develops a comprehensive framework for the study of telecommunications networks. We draw upon important developments in the graph theory that is used to represent networks. We examine economic models of networks and apply these developments to study the legal aspects of network industries. The result is an outline of a theory of telecommunications networks that generates insights into important public policy debates, including mandatory access and network neutrality.

The book is organized as follows: The first part introduces the economics of networks.

Most Internet users communicate through a suite of nonproprietary protocols known as the transmission control protocol/Internet protocol (“TCP/IP”). Widespread adoption of TCP/IP has given the Internet a nearly universal interoperability that allows all end users to access Internet applications and content on a nondiscriminatory basis. Commentators, led by Lawrence Lessig (2002), have long been concerned that cable modem and DSL systems will use their control of the “last mile” of the network to block or slow access to content and applications that threaten their proprietary operations. The concern is that the resulting reduction in interoperability will produce a less favorable environment for competition and innovation in the market for Internet content and applications.

Some regulatory proposals attempt to preserve the transparency of the Internet by regulating last-mile providers' relationships with end users. Other proposals seek to regulate last-mile providers' relationships with network and content providers. Some call for mandating interconnection of broadband networks along standardized interfaces such as TCP/IP. Others argue in favor of a presumption that any discriminatory access agreements are anticompetitive, leaving the precise regulatory requirements to be developed over time through case-by-case adjudication. Although these proposals vary considerably in both their terminology and details, they can comfortably be aggregated within the broad rubric of “network neutrality.”

The term “network neutrality” is something of a misnomer. Adoption of any standardized interface has the inevitable effect of favoring certain applications and disfavoring others. For example, TCP/IP routes packets anonymously on a “first come, first served” and “best efforts” basis.

In this part, we apply the framework developed in the earlier chapters to a series of current issues in telecommunications policy. The first issue, discussed in this chapter, is the regulation of local telephone networks. Chapter 9 examines how antitrust law has been applied to the telecommunications industry. Chapter 10 uses the framework we have developed to analyze the regulation of broadband networks. Chapters 11 and 12 examine one of the most high-profile recent issues in telecommunications policy: “network neutrality.”

The History of the Regulation of Local Telephony

Early State and Federal Regulation and the Communications Act of 1934

Under the system of federalism enshrined in the U.S. Constitution, the authority of the federal government is limited to interstate commercial activities. The regulation of intrastate telephone rates fell under the jurisdiction of the states. Although early legislation in five states had authorized some degree of regulation over local telephone companies, state regulation of local telephone service did not begin in earnest until 1907, when states began authorizing their public utility commissions to oversee the reasonableness of local telephone rates. By 1921, all but three states had instituted some form of regulation of local telephone rates (U.S. House 1921).

Federal regulation of interstate telephone service began in 1910 with the enactment of the Mann–Elkins Act, which gave the Interstate Commerce Commission (ICC) the power to review the reasonableness of interstate and international telephone rates (Statutes at Large 1910, pp. 544–46).

The consensus economic position is that so long as competition is sufficiently robust, market prices represent the best reflection of value. The market price is the outcome of the forces of supply and demand. The supply side of the market reflects the costs to sellers of providing a good; the demand side reflects the benefits to buyers from consuming the good. At market equilibrium, prices are thus determined by the marginal cost to sellers of providing a good and the marginal benefit to buyers of consuming it. Prices are adjusted through the process of exchange to balance supply and demand and to clear the market, so that prices are further reflections of scarcity, the meeting of consumer wants, and supplier capacities.

Because the services of a network are comparable to the output of other types of production facilities, market processes can allocate them. Markets refer to the interaction of buyers and sellers, with market prices mediating between what buyers are willing to pay and what sellers are willing to accept. Market prices are determined through the activities of suppliers, customers, and intermediaries such as retailers and wholesalers. In the short run, firms raise prices when demand exceeds supply and lower prices when supply exceeds demand. In the long run, suppliers make production decisions by comparing the prices of goods to their costs and to the prices of alternative goods the supplier might provide.

During the Internet's initial, narrowband phase of development, Internet service providers (ISPs) pursued a variety of architectural approaches. Some ISPs, such as CompuServe, Prodigy, and America Online, initially adopted relatively restrictive policies, which only provided end users with access to proprietary applications and content and charged them for network usage on a per-minute basis. Others followed a more permissive approach, opening up their networks to all content and applications providers on a nondiscriminatory basis and allowing end users to download any content, run any application, and attach any device for a flat monthly fee. The latter approach ultimately proved more attractive to consumers, and end users became accustomed to a world in which they faced few restrictions either on the ways they could use their network connections or on the amount of bandwidth they consumed.

This same debate has resurfaced as the Internet has begun to migrate from a narrowband to a broadband architecture. Once again, network owners have begun to experiment with more restrictive approaches. With respect to end users, some network owners have begun to offer bandwidth tiers to end users, in which the amount that customers pay varies with the amount of bandwidth with which they are provided. Others have placed restrictions on end users' latitude to run certain applications or attach certain devices. Still others have considered alternative pricing relationships with respect to content and applications providers, under which transmission speed would depend on the tier of service purchased.