Just how open must an open network be for an open network to be labeled "open"?
First Monday

Just how open must an open network be for an open network to be labeled "open"? by Jonathan Sallet

Just how open must an open network be for an open network to be labeled "open"? by Jonathan Sallet
In 2003, the U.S. Federal Communications Commission (FCC) will decide in multiple contexts the extent to which governmental action should be used to maintain the "openness" of telecommunications and Internet networks. At the same time, the European Union will put into effect its new, comprehensive Access Directive. 2003 may, therefore, be a critical year for the future of governmental policy towards the "openness" of next-generation networks.

This paper argues that the debate between "open" and "closed" networks has been insufficiently precise and, therefore, has failed to bring to policy makers' attention critical factors of decision. That is because the choice between "open" and "closed" networks is not binary; rather it consists of different policy bases operating from different perspectives on the network. Arguments for or against governmental opening of a network can be premised on a variety of disciplinary regimes that include, for example, engineering principles, economic theory, social philosophy and legal analysis. Often ignored is the plain fact that these disciplines do not always line up with each other. This will be critical to understand if in the future policy makers are asked to weigh claims of economic theory — say the need to encourage investment — against claims of social philosophy — say the value of free speech and experimentation.

Nor do contentions necessarily operate at the same perspective. From a user's perspective, the network can include the activities of an end user, competitive network provider, an independent content/software provider, or the network owner itself. Thus a claim of an end user's "right" to access content through a network may shed little light on the claim of a competitive network provider to use that same network.

This paper demonstrates the interplay of the policy bases and network perspectives with four examples: Access to Regional Bell Operating Company (RBOC) networks; access to U.S. cable networks; the European perspective as demonstrated in the Access Directive; and, the architecture of the Internet itself. Along the way, the paper also notes, as an aspect of future analysis, the extent to which the Internet as an "idea" influenced public policy in a manner that departed from normal interest-group politics.

The paper posits, as an example, a decisional template that could be employed, for each perspective on the network, to distinguish between policy disciplines. The paper concludes by noting those circumstances that reinforce the continuing importance of the "open/closed" network question. The goal of this paper is not to advocate for any particular policy outcome. It is, rather, to demonstrate that current policy analysis would benefit from applying greater analytical precision to the question of whether — and why — governments should act to open next-generation networks.

Finally, the paper includes an addendum reviewing regulatory activity in February 2003.


I. Introduction
II. Policy background
III. The multiple faces of open v. closed networks
IV. Perspectives on openness
V. Application of the bases of decision-making to perspectives on openness
VI. The interdependence of openness
VII. Conclusion




I. Introduction

The year 2003 will highlight — for technology communities on both sides of the Atlantic — the question: What if anything should a government do to "open" a next-generation network?

In the United States, the Federal Communications Commission (FCC) has just met a February deadline to consider the extent to which local telephone facilities — including facilities used for the provision of DSL service — will be available to competitors (A review of this FCC decision, announced after the completion of this paper, is included at the end of this paper entitled "Addendum — Recent Developments.") In the succeeding months of 2003, the FCC will review the appropriate regulatory treatment of broadband networks in a series of proceedings that concern both telephone and cable networks (Powell, 2003). These proceedings are likely to decide — for the foreseeable future — the extent to which competitors have access to the facilities of incumbent U.S. telephone and cable networks for the purpose of supplying alternative high-speed data services to residential and business end users [1].

At the same time, European attention is focused on the very same question — whether telecommunications and Internet networks should be opened by governmental action. On the heels of the promulgation of its comprehensive Access Directive that is slated to become effective in May of 2003, and which is applicable to all forms of electronic communications services [2], the EU sponsored consultations in February of 2003 to consider the manner in which both Digital Television and wireless 3G networks would be operated as "open platforms" (European Commission, 2002).

Thus, 2003 may prove to be the year in which the next generation of public policy for next-generation networks in Europe and the United States is set in stone — decisions that may have great significance for the future of broadband deployment and innovation, the recovery of the economically-damaged technology sector, the architecture of next-generation networks and, of course, the nature of the content and technological experiences that are available for end users throughout these societies.

This current attention — and the importance of the current set of policy decisions — is not surprising. In the recent period that this paper will label the "second generation" of public policy directed towards the creation and deployment of next-generation networks, considerable attention has focused on the circumstances in which government should require that a private network be opened in order to spur the development and deployment of next-generation network capacity and creation. But are governments asking the right questions and are applying the correct analysis?

This paper suggests that the current public policy dialogue is insufficient. Therefore, it analyzes the issue of "openness" in next-generation networks for the purpose of (i) demonstrating that the current debate — fought in different places over different methods of digital transmission — cannot be properly understood as posing a purely binary choice between "open" and "closed" networks; (ii) suggesting the importance, in considering historical examples of "openness" policy, of analyzing the role of the Internet as an "idea" in competition with more traditional components of interest-group and bureaucratic politics; and, (iii) offering for purposes of discussion an example of a prescriptive framework that could guide policy makers in determining whether next-generation networks should be opened through governmental intervention.

To that end, this paper proceeds as follows: Part II offers a short background on the economic and policy context in which the "openness" question has arisen in recent years, the claims that have been made on behalf of both "open" and "closed" networks, and the historical example of the Bell System, which demonstrates the complexity inherent in the issue.

Part III, after noting the circumstances that may obviate the need for any governmental intervention at all, reviews multiple disciplinary bases on which the policy of "open" networks has been presented — engineering principles, economic theory, social philosophy and legal precepts. The variety of forms of reasoning proceed from different fields of expertise and rely on different forms of argumentation — a recipe for a debate in which competing sides fail to confront the underlying premises of each other's contentions.

Part IV then divides the "network" into four perspectives in order to illustrate different network perspectives from which the "open/closed" dichotomy can be presented to policy makers: From the differing perspectives of the end user, a competitive network provider, a content/software provider and then from the perspective of a network provider itself.

Part V considers four examples of the openness issue in order to illustrate briefly how the various disciplines differ in their applicability to the diverse circumstances in which the "open/closed" network issue appears, and to illuminate issues concerning the manner in which a public policy maker would, in any event, actually apply these disciplines. Noted in this section are the questions of the access by horizontal competitors to the "fast-Internet" resources of incumbent telephone companies, the role of "free speech" values in the "vertical" world of cable modems, the future of the Internet itself, and the distinct perspective on these issues of the European Union. An observation offered in this section is that an analysis of the way public policy affecting next-generation networks was actually made during recent years could helpfully consider whether, to what extent and for what period of time, the Internet as an "idea" was able to disrupt the normal functions of interest group and bureaucratic politics.

Part VI focuses specifically on the prescriptive: How should policy makers advance their analyses of assertions that governmental action need be taken to "open" networks? The paper asserts that policy makers should take additional care to understand both the precise perspectives on the network (e.g., end users) that are implicated and the precise disciplinary principles that actually apply to any situation. As an illustration of how policy makers might make some progress, the paper offers, for purposes of discussion only, a template of presumptions that could be applied. This section emphasizes the particular need for policy makers to evaluate whether there is independent substantive weight to be given to arguments invoking freedom of speech and the freedom to innovate.

In its conclusion, the paper returns to the proceedings pending now at the FCC that may, in the short term, provide comprehensive conclusions to a number of the important issues noted in this paper, and then identifies issues that may tend to keep the question of "open" networks alive in the coming years regardless of the FCC's coming decisions. These include consolidation in the economic sector that provides next-generation networks, continuing impact of past market power in the traditional telecommunications sector, the potential collapse — or culmination — of alternative technologies that, at the present, may be seen as supporting additional next-generation networks, and control of content in ways that are seen to limit the legitimate expectations of end users.

Finally, in a short addendum, this paper notes recent developments in February 2003 that bear upon this question of "open" networks — including the FCC's announcement of its decision (without the issuance of a formal order and analysis) concerning competitive access to the networks of incumbent telecommunications companies.



II. Policy background

In April of 1995, the United States government took a step as momentous as any previous action affecting the development of information networks. Without great fanfare, the National Science Foundation shut off NSFNET — and for the first time the Internet backbone passed into the private domain. The network's day as a government-funded research experiment was over; its day as a multi-use network of networks had come.

The succeeding years have brought both boom and bust. Initial frenzy about the unlimited potential of the Internet focused on both personal empowerment and economic potential. The Internet was seen by many — especially from among its early colony of inventors and users — as the final culmination of a grand utopian vision of a data-driven world that would see "sales of computers overtaking those of phones and televisions; growth stocks soaring, venture capitalists creating prodigious wealth, youth triumphing over age, the West rising, bandwidth spreading, entrepreneurs growing rich, the power of ideas exceeding the status quo" [3].

Equally utopian — if not as intentionally so — were the predictions of the purely economic gains that would be garnered from use of the Internet. Touting its pre-millennium selection of Amazon.com's founder Jeff Bezos as "Person of the Year" Time (1999) gushed that

"[t]his year it's as if "ideagetthemoneyhireaCEOlaunchpartyIPO" has become one big, fast millennial screech. Companies that barely existed a year ago are publicly traded, their founders ungodly wealthy. Some argue the world has entered a long boom, a kind of economic speed loop, where the centrifugal force spins off nothing but wealth and happiness."

And the "some" who would make that argument included not just fast lane e-commerce magnates, but revolutionaries from the new world of competitive telecommunications. In 1999, WorldCom achieved a market capitalization of $160 billion — approximately $160 billion above its current level — and new "data" providers were hot on the trail of broadband business — companies like Covad Communications Group, Northpoint Communications and Rhythms NetConnections were "leading the pack" in 1999 (Grice, 1999), but in bankruptcy by the end of the last quarter of 2001 (Chidi, 2001; Wagner, 2001).

During the rise — and apparent fall — of the network economy, public policy makers were hard at work ensuring that the promise of the Internet would be fulfilled. A critical aspect of that work focused on policies that would lead to the build-out of next-generation networks [4]. These would be the networks that would, after all, carry the Internet to consumers and businesses. Without them — and existing networks were widely viewed as inadequate in the mid-1990s [5] — it would be impossible to fulfill both the promise of Internet prophets and Internet profits.

And so, in the United States and around the world, policy makers went to work to spur the deployment of next-generation networks. In a real sense, this was the second generation of public policy directed to next-generation networks. The first was characterized by action during the preceding decades best symbolized by the steps taken by the U.S. government to develop the Internet and formulate the competitive structure in which data transmissions were sent over telecommunications facilities. The funding by the Department of Defense of ARPANET and the structural separation between "basic" telecommunications services and "enhanced" data services both exemplify this first period (Hafner and Lyon, 1996).

The second generation of public policy had larger aspirations. Spurred by the grand predictions of the impact of the Internet, policy makers set out to ensure that its benefits were widely reaped by society. It is now possible to begin to take stock — against the backdrop of economic boom and bust — of the policy choices to which leading nations of the world committed themselves and to review the main engines of next generation policy initiatives during this second era of policy. They include [6]:

  • Global commitment to opening previously closed telecom markets, including the WTO agreement on basic telecommunications services [7], the passage in the United States of the Telecommunications Act of 1996 [8], and the adoption by the European Union of a broad set of market-opening measures [9].
  • Strong governmental emphasis on the ubiquitous provision of broadband as a means of achieving both economic and social goals, including tax incentives and similar financial incentives for the deployment of broadband services [10], as well as specialized efforts to ensure the availability of broadband services to certain user populations, such as schools and libraries [11].
  • Continued search for actions, like spectrum reallocation, that can spur new means of delivering next generation services, including the 3G auctions in Europe and the authorization of unlicensed WiFi and UltraWideBand services in the United States [12]; and
  • Careful scrutiny of the market structure of evolving network markets, including the recent rejection of the proposed Echostar-Direct TV merger [13], the European Union's rejection of the proposed WorldCom/Sprint merger on the ground that it might lead to domination of the Internet backbone [14], and the U.S. government's placement of conditions on the AOL-Time Warner merger [15].

The focus of this paper — to what extent, if any, should government take action to ensure that next-generation networks are "open"? — is implicated by at least two of the basic policy goals noted above. The requirement of open networks, as in the United States Telecommunications Act of 1996, may arise from regulatory actions in pursuit of open telecommunications markets. Thus, Section 251 of the Act includes the requirement that in specified circumstances the incumbent — and only incumbent — local exchange companies offer "unbundled" elements for lease to competitors at wholesale prices — a form of open networks [16].

Requirements of "openness" may also arise from case-by-case adjudications of the market structure of evolving network markets, as in the requirements imposed by the United States Federal Trade Commission (FTC) and the FCC requiring that the merged entity AOL-Time Warner permit unaffiliated ISPs to use their cable-television facilities in order to provide competitive "broadband" Internet access — a different form of open access (Federal Communications Commission, 2001).

The debate between "open" and "closed" networks — or, to use different language between "forced" and "voluntary" access — is neither dispassionate nor new. "Open" networks are, to their adherents, desirable because "[o]pen platform services will spur diversity in the electronic media," (Kapor, 1993), they will prevent "[n]onprofit groups, public interest coalitions, and independent artists (to name just a few) [from being] relegated to second-class status, leaving the Net in the hands of corporate interests" (Center for Digital Democracy, n.d.), and they will ensure that "the free flow of content [is not subject] to artificial constraints at the hands of either government regulators or would-be monopolists" (Gore, 1994). They will, in other words, enhance the deployment of next-generation networks because an open network "complements the deployment of new networks, because it generates customers for new entrants, as well as cash flow for financing network construction" [17].

The advantages of "closed" networks spark equal passion. The use of "open access" actually "limits Internet access by encouraging ... under-allocation of broadband spectrum [in cable systems], and by introducing coordination problems slowing technology deployment" (Hazlett and Bittlingmayer, 2001); in the telephony arena it "is simply a subsidy for companies like WorldCom and AT&T that is totally dependent on [incumbent telephone] companies ... being able and willing to make all the capital investment and absorb all the profit margin pressure" (SBC, 2002a), and, therefore, reduces the incentive to invest capital in — and slows the deployment of — next-generation networks [18]. Bad regulation, in the form of the pricing of network components has, in this view, led directly to weak earnings by the incumbent telephone companies — a direct financial blow at a sector already on the ropes (SBC, 2002a).

That arguments clash in a public-policy forum is axiomatic. In this case, however, the underlying concept of "open" networks is more complicated than the rhetoric suggests. Consider the question: Was the Bell telephone network at the height of its power, circa 1950, an open or a closed network? And the answer, like Schrodinger's Cat, is, of course, that it was simultaneously both open and closed.

At the level of content, and its philosophical partner free expression, the telephone network was "open" in every material way. For all intents and purposes, users were entirely free to say anything to anyone and to hear what anyone else had to say to them. Other than operator assistance and similar ancillary services, the Bell Systems placed no content on the network. The total sum of communication came from imagination — or at least the mouths — of end users.

At the same time, the Bell System was almost entirely closed at the level of terminal equipment and engineering. No one could attach any equipment to the network without the permission of the Bell System; the engineering of the network was entirely closed to users in order to preserve the integrity of the system — as protected by Bell technicians. If you wanted a telephone, you had one choice — to rent it from the Bell System [19].

And in the middle was access to the network by other telephone companies. Those non-Bell entities that operated in geographic regions not directly served by the Bell system had the ability to exchange traffic with the Bell network — a critical form of openness that provided ubiquitous service to U.S. telephone users. But no company trying to compete directly against the Bells, say for the provision of long-distance service, had the ability to connect its network to the Bell system [20].



III. The multiple faces of open v. closed networks

A. Circumstances that may obviate debate over governmental action

At the outset it is important to reiterate that the question posed by this paper concerns the circumstances in which it may be advisable to require some form of government action to "open" a next-generation network. Virtually all commentators would agree that this question follows, rather than precedes, those forces that may work to open previously closed networks in the absence of governmental action.

Technological innovation may threaten existing network usage in a manner that causes the network provider to re-think an earlier decision to operate in a closed manner. For example, imagine a wireline network that sees significant amounts of traffic migrating to a wireless competitor. One rational business strategy would be for the wireline network to "wholesale" its services or network elements to competitors in the hope that more competition on its network will offset the impact of competition between networks.

Competition from other networks that had not previously provided competitive services might have the same effect. The classic example of such competition in recent American experience is, of course, the competition in some parts of the United States between cable modem service and copper-based DSL service. Cable and telephone networks had existed side-by-side for a long time before they competed against each other for the provision of fast-Internet services to residential customers. To the extent that the network provider is focused on the total amount of traffic on its network (as opposed to upstream profits it may separately reap as a content or Internet service provider or the potential of horizontal competition between video-based Internet service and its more traditional cable television offerings), it could decide that it would be advantaged by inviting other ISP's onto its network. There is now some evidence, as a result of government-ordered actions, that such a strategy of openness does, in fact, work to the benefit of the network provider. For example, the CEO of Time Warner Cable was quoted in 2002 as saying that the presence of other "ISPs [have] boosted Time Warner Cable's broadband Internet subscriptions additions 20 to 25 percent in initial launch markets without cannibalizing subscriptions to its own Road Runner service" (Cable Datacom News, 2002).

Also, the government may engage in non-adjudicatory or non-regulatory actions that boost technology or help to seed markets in a way that creates these kinds of pressures. The U.S. government's seminal support for research that became the Internet is the leading example.

B. Proffered bases for and against governmental action

1. Engineering the network

a. The end-to-end principle of Internet architecture

The Internet is often extolled as an "open" architecture — a network of networks in which data can be transmitted between users without the need for homogeneity of application or device. The engineering principle on which this "open" architecture is built has been identified as the "end-to-end principle," famously articulated as such in 1981 by J.H. Saltzer, D.P. Reed and D.D. Clark (Saltzer et al., 1984).

The end-to-end (or "e2e") principle states a preference — not a requirement — for intelligence to be embedded in the devices at the end of the network rather than "in" the network itself. The engineering preference recognizes that placing additional functions in the network may raise the cost of performance by, for example, embedding functions in the network that only some higher-level applications require. An additional concern is that the lower-level network may not have all of the information needed to perform a specific function that is possessed by applications at the "edge" of the network. To take one example offered by the original proponents of the e2e principle, real-time voice transmission evokes "an unusually strong version of the end-to-end arguments" because "if low levels of the communication system try to accomplish bit-perfect communications [of the kind one might desire for voice recognition] they will probably introduce uncontrolled delays in packet delivery" [21]. The better option is to allow the end users to deal with the slight errors in transmission that occur by, for example, requiring the end-user to ask, for example, "Would you please say that again" [22]. This paper will label this form of strict engineering inquiry the "narrow" version of e2e.

From an engineering origin, however, e2e has established a "broader" version as a policy preference of potentially profound meaning. So, for example, the choice between e2e, on the one hand, or a network in which intelligence has been embedded, on the other, has been characterized as posing the tradeoff between "Freedom" versus "Control"; "Innovation" versus "Manageability" and "Popular Democracy" versus "Elitist Democracy". Thus the potential introduction of quality-of-service differentiations on the network, say between the treatment of real-time video and traditional e-mail, can be seen as a zero/sum choice, not just between technical specifications, but between a world in which some must be left behind if others are to be advantaged, on the one hand, and a world in which the entire network is improved to benefit all users, on the other [23].

In a similar vein, the broader e2e principle has been cited as a critical step in fostering user innovation. Under this view, the fact that Internet end users have been free traditionally to experiment, without restraint in terms of application or device, has provided a source of tremendous power to the technological evolution of the network (Bar et al., 2001). Perhaps the prime example is the invention of the World Wide Web by Tim Berners-Lee who created an easier way to share information among scientific colleagues at CERN (Berners-Lee, 2000).

b. The closed paradigm of the Bell System

The traditional view of "the" network — the Bell network that came to maturity in the Twentieth Century — was that it needed to be absolutely closed to outsiders on all technical levels.

In the middle of the Twentieth Century, the Bell Corporation had filed tariffs that expressly forbade attachment to the telephone instrument of any device "not furnished by the telephone company" [24]. Faced with a challenge to the propriety of this rule, the Bell System contended that the ban on "foreign attachments" was needed to ensure the smooth operation of its entire network; otherwise, it was said, responsibility for ensuring the operation of the network would be fatally divided and independent operators, committed to the use of their specific equipment, would resist improvements to the network [25].

The Bell concern for the sanctity of its network — and the need to keep it closed to outsiders — sprung from well-deserved pride in the creation, in the words of John deButts, an AT&T Chairman in the mid-1970s, of "the nationwide switched network, a technological achievement unmatched in scale and complexity and the precision of its operations" (Von Auw, 1983, Appendix B).

In two famous cases, Hush-a-Phone and Carterphone, federal authorities rejected the Bell System's contention that a closed network at the engineering level was required in order prevent harm to that network. Instead, through means that included the creation of certification standards for non-Bell equipment, the FCC mandated that "foreign attachments" be barred only if they were harmful to the network. This limited conclusion, applicable initially to a plastic cup attached to the end of a telephone receiver, was later extended dramatically to reach inside the workings of the network itself. Under the Telecommunications Act of 1996, incumbent local exchange companies are governed by a legal standard that, in the appropriate circumstances, provides for "nondiscriminatory access to network elements" and for "physical collocation" at their premises [26].

Those early rulings did not resolve the controversy. A few years after the Carterphone decision, AT&T complained that intercity private lines equipped with customer-provided equipment generated trouble reports at least fifty percent higher than those with only Bell-supplied terminals (Von Auw, 1983, Appendix B). But, as the subsequent provisions of the Telecommunications Act demonstrate, the tide had clearly turned.

Concerns that had been long shelved resurfaced in the wake of the September 11 terrorist attacks. By 2001, the successors to the Bell networks had been opened on an engineering level to such an extent that in Manhattan competitive communications companies were able to lease portions of the incumbent network and to install their own equipment — connected to the leased equipment — in co-location cages located on the premises of the incumbent but under the control of the competitive provider. In the years since the passage of the Telecommunications Act of 1996 co-location had been an issue of continuing discussion, but one not seriously discussed with regard to the engineering integrity of the network [27].

The terrorist attacks on the World Trade Center badly damaged a Verizon switching facility located nearby, temporarily disrupting service to 3.6 million data circuits and a large number of voice lines (Dawson, 2001). Although the attacks were exogenous to operation of communications facilities themselves, concern was soon raised about the nature of access afforded by federal regulation. In a speech delivered in early October, 2001, the president of the United States Telecom Association warned that: "Interconnection, unbundled access, and collocation requirements present security challenges" (McCormick, 2001). In the same period of time, the co-CEO of Verizon, while citing, "some new security concerns related to safeguarding our key network assets," emphasized that "true security lies in having a diversity of technologies that give customers redundant capabilities and provide alternative ways for Americans to communicate" (Seidenberg, 2001) — a perspective that could similarly justify lessened emphasis on the use of shared network facilities [28]. In October, 2002, the same thought was repeated by FCC Chairman Powell, who said that "[o]nly through facilities-based competition," which does not require the opening of any entity's network, "can our Nation attain greater network redundancies for security purposes and national emergencies" (Powell, 2002).

2. Economics

a. Network Economics

The rise of electronic networks, and particularly the Internet, brought to the fore economic concepts that, while not wholly new, are said to have particular saliency in the new world of next generation services. Of particular importance is the concept of "network effects" — the value of a network increases as more people join it (Katz and Shapiro, 1994). The telephone network is a classic example — a single telephone standing alone may find its highest value as a paperweight but a mature network creates great value for all of those on it and thus a great attraction for anyone thinking of joining it. When faced with a choice between two networks — one of which has millions of end users and the other of which has only a few — the typical end user would find the choice easy — choose the network that provides the greatest number of connections. And that would, of course, complicate the ability of a rival network to build a customer base [29].

In the world of technology, network effects may arise from "first mover advantages" that, through slight differences in market share at first, cascade into market leadership as network effects strengthen the initial lead and turn it into market dominance. Similarly, network effects that reinforce the initial superior qualities of a network can lead to a circumstance in which the first network is essentially unassailable by competitors attempting to introduce a rival network employing the same basic technology. The battle between VHS and beta videotape formats is often cited as an example of a circumstance in which small initial differences between competing technologies create positive feedback for the leader that eventually leads to market domination [30].

The danger, according to economic theorists, is that networks are particularly suited to gain an initial advantage and then, once advantaged, can become strong enough to maintain that advantage by refusing any sort of open connection to other, competitive networks, thus maintaining market dominance, limiting consumer choice, and diminishing opportunities for innovation [31]. So, for example, in the early years of the Twentieth Century, local telephone networks did not exchange traffic, which required customers to subscribe to multiple networks in order to obtain ubiquitous coverage — a circumstance that was thought to demonstrate the necessity of a single monopoly telephone network [32].

An alternative approach could have relied on principles of "openness". Had regulators required that each local telephone network exchange traffic with each other network, then all subscribers would have been able to call each other and multiple networks could have competed on performance characteristics other than market share — and the concomitant network effects — alone. Thus this school of economics can be used to posit a world in which governmental action — assuring competition and innovation through various requirements of openness — can be properly understood as a natural corollary of the nature of network markets.

b. The economics of investment

Someone must, of course, acquire the capital to build the networks in the first place and circumstances that hinder capital acquisition may have the effect of preventing the creation or enhancement of next-generation networks in the first instance. This investment-based view of the economics of open platforms does not deny that private parties may, as a matter of business choice, open their networks, nor that such open access is desirable, nor even that, in some limited circumstances, the presence of monopoly power may support "openness" requirements. What it does deny is that open platform requirements are costless and what, as a generalized matter, this point of view does contend is that the costs of governmental intervention outweigh the purported benefits.

In focusing on the costs, three major contentions are made. First, that openness requirements, by allowing competitive firms to reap profits from use of the network, diminish the value of the network to the original network provider and, therefore, diminish the inventive to invest in network upgrades and new network construction. Because old-line copper networks may require significant technological upgrade before they are capable of carrying DSL, and because the provision of DSL itself requires the installation of equipment not used for standard voice traffic, the Bell companies assert that allowing others to piggyback on their investment directly lowers their incentive to build out DSL-capable networks [33]. Lower incentives to build networks are not, of course, commensurate with the stated desire of stimulating the creation of next-generation networks.

One particular mechanism of profit-maximization can be found in price discrimination. Use and usage restrictions can be utilized to promote price discrimination that, in turn, could be defended in certain circumstances as a legitimate means for increasing the incentives to build out networks. For example, one might argue that restricting residential Internet connections to only non-commercial use could increase the incentives for the build out of advanced networks to commercial customers [34].

Second, it is contended that openness requirements that promote shared usage of network stimulates competition that is less beneficial than full facilities-based competition. In this view, which has been explained recently by FCC Chairman Powell, only through facilities-based competition can an entity offer "true product and pricing differentiation for consumers [and] bypass the incumbent completely and force the incumbent to innovate to offset lost wholesale revenues" (Powell, 2002). In other words, the most robust and socially beneficial form of competition arises when two or more distinct networks compete to provide the same service as, for example, when wired and wireless networks provide competing voice services or cable and telephone networks compete to provide fast-Internet service to residential customers. Such competition, it is said, provides society with a choice in technologies and the technical characteristics of different networks. By contrast, it is argued that competition that rides on "someone else's network" cannot provide these distinct product features and therefore does not serve the public interest in promoting competition (especially if the presence of "sharing" competitors has the effect of discouraging full facilities-based competition) [35].

Third, the argument is made that government intervention itself imposes dead weight costs on burdened network providers. Regulatory processes can certainly burden economic activity — especially when they are lengthy, when the meaning of controlling authority is uncertain (often resulting in litigation) and when they require the marshalling of political and legal, as opposed to technological and product marketing, resources. The economic argument described here asserts that these costs are non-productive (a contention often connected with the view that, left to their own devices, network providers may choose openness).

In the fall of 2002, SBC Communications presented a particularly stark example of this economic contention. Announcing substantial layoffs and a decrease in planned capital spending, SBC put the blame squarely on the existence and implementation of the Telecommunications Act requirements of open access — in this case the specific requirement that incumbent telephone networks allow competitors to lease "unbundled" elements of their network. The theory behind the provision was that the availability of leased elements would permit new competitors to enter a market without having to incur the costly investment of laying down wires to each and every customer, thus obviating the monopoly power of the incumbent telephone companies. But in SBC's view the impact of that openness requirement was simply to "subsidize competitors" who have no requirement of investing in network construction and no responsibility to maintain the networks they use. The pronounced result: SBC's decision to reduce its investment in its network (SBC, 2002b).

SBC returned to this contention when, in October 2002, it released its quarterly earnings. As noted above, SBC asserted that network unbundling "is simply a subsidy for companies like WorldCom and AT&T that is totally dependent on companies like SBC being able and willing to make all the capital investment and absorb all the profit margin pressure," and that the effect of unbundling is to reduce incentives to invest capital in and speed the build out of next-generation networks (SBC, 2002a).

3. Social philosophy

a. Jeffersonian democracy — towards an open society

That the First Amendment to the United States Constitution states principles of great importance would seem to be axiomatic. That these principles — cloaked in the penumbra of the nation's first great philosopher of liberty and extended to non-governmental actors — would inform the operation of next-generation networks might be less obvious at first glance.

But the spirit of Thomas Jefferson finds easy companionship in any group hospitable to the notion of a decentralized, non-hierarchical society in which individual freedom, above all, is prized. And the Internet, in its initial pre-commercial manifestation, had many attributes of such a society. Supported by governmental research grants, operated by a reasonably homogenous group of experts, shielded from short-term commercial aspirations, lying atop a heavily regulated telecommunications structure, the Internet could easily be seen as a prototype for a new, open society.

Such an argument was made in an influential essay published in 1993 by Mitch Kapor (1993), who made the case "for a Jeffersonian Information Policy" that would be "founded on the primacy of individual liberty and a commitment to pluralism, diversity and community." In such a world, Kapor explained, "openness, as exemplified by the personal computer and the Internet, can be a driver of prosperity and diversity in any new market." And diversity, the "support of non-mass ideas — is a good thing: more people thinking their own thoughts, reaching their own conclusions. Today's fringe idea is often tomorrow's mainstream."

In a similar vein, John Perry Barlow's "A Declaration of Independence of Cyberspace" asserted that the world of cyberspace is "creating a world where anyone, anywhere may express his or her beliefs, no matter how singular, without fear of being coerced into silence or conformity"; a world that will constitute "a civilization of the Mind in Cyberspace" (Barlow, 1996). Nor surprisingly, Barlow himself has been labeled "the Thomas Jefferson of Cyberspace" [36].

Much of Jefferson's enduring power is, of course, linked to his appeal both to those who seek the use of governmental power to protect individual liberty and those who seek his ideal of a government exercising very limited power. The Internet, in its pre-commercial composition, is often said to embody both — considerable individual liberty with little or no governmental oversight. The "Internet" as a model thus poses a metaphorical attraction similar to that of Jefferson himself — it can be invoked as a model for no governmental regulation or for the proposition that individual rights of self-expression must be protected — even at the cost of governmental intervention.

Recent comments by members of the high technology community demonstrate the gingerly embrace of possible governmental action even by those who favor a "deregulatory" approach when it comes to governmental intervention into the workings of technology markets. In a letter lodged at the FCC in November, 2002, a number of high-technology companies, including Amazon.com, eBay, and Yahoo, urged the FCC to "assure that consumers and other Internet users continue to enjoy the unfettered ability to reach lawful content and services, and to communicate and interact with each other and reach desired Internet destinations without impediments imposed by transmission network providers" (Coalition of Broadband Users and Innovators, 2002). Despite its preference for free market solutions, the High Tech Broadband Coalition endorsed the potential for governmental action in order to ensure that customers of cable modem services "have unrestricted access to their choice of content" and have the general right "to run applications of their choice" (High Tech Broadband Coalition, 2002). Why? According to the Coalition, to protect the "innovation and consumers' freedom which have been the engines of the information age" (High Tech Broadband Coalition, 2002).

Also useful in this context is Cass Sunstein's (2001) warning of the danger of group polarization as a result of the Internet promoting only like-to-like speech. He says, "Group polarisation is highly likely to occur on the Internet. Indeed, it is clear that the Internet is serving, for many, as a breeding ground for extremism, precisely because like-minded people are deliberating with one another, without hearing contrary views." Thus, the "most reasonable conclusion is that it is extremely important to ensure that people are exposed to views other than those with which they currently agree, in order to protect against the harmful effects of group polarisation on individual thinking and on social cohesion." In this context, open access that promotes a diversity of views could be viewed as vindicating not only the individual right to self-expression but also a societal interest in promoting diverse information flows.

b. Freedom From governmental dictates

There is no monopoly on the cause of freedom, of course, and so it is not surprising that some Internet philosophers have concluded that the nature of the Internet has rendered obsolete the kind of traditional notions of governmental action that might support the mandatory opening of networks.

In 1994, four leading theorists — Esther Dyson, George Gilder, George Keyworth and Alvin Toffler — published a self-styled "Magna Carta for the Knowledge Age." They proclaimed the advent of the "Third Wave" — the Knowledge Age — that would require redefinition of "the meaning of freedom," and "structures of self-government" among other concepts. A critical attribute of the Third Wave is "the overthrow of matter" and the notions of scarcity that accompany it. They said that "[t]he world of Cyberspace would be characterized by Unlimited Knowledge" because "new information technologies are driving the financial costs of diversity ... down toward zero ... " (Dyson et al., 1994). With the "death of the central institutional paradigm of modern life, the bureaucratic organization," the government should turn its attention to "massively deregulating the fast-growing telecommunications and computing industries." Taking direct aim on Kapor's Jeffersonian Vision, the statement concluded that the "only effective near-term path to Universal Access" requires "reducing barriers to entry and innovation." In other words, enhancing the freedom of private network providers to build and configure their own networks is the best governmental mechanism to achieving the goals of an "open" network. (Similarly, it can be argued that a network provider's "freedom" to chose and "edit" the content that flows over its network is a substantial form of freedom in and of itself.)

George Gilder has subsequently focused directly on the openness question as applied to both telephony and cable systems. As to the regulation of incumbent telephone systems, Gilder (2001) explained, "in dynamic technology markets such as Internet broadband, monopolies are inevitable, virtuous, and fleeting" (emphasis added). Virtuous, because "[e]very innovation creates a monopoly at the outset, and monopoly rents pay for financial risks and costs entailed in bringing innovation to market."

Similarly, Gilder critiqued so-called "open access" as applied to cable systems. Returning to the notion of bandwidth abundance, Gilder explained that "[t]rue open access — allowing any ISP to send its content over any cable conduit — is desirable," but that "when bandwidth is abundant, consumers will overwhelmingly prefer conduits open to everyone's content" — again obviating the requirement, and the justification — for government action (Gilder, 2000).

4. Competition policy

For the purposes of this paper, a critical means of legal analysis includes the application of competition policy to the development of next generation markets. That is not a new conclusion. Competition policy, in the form of anti-trust litigation, played a critical role in the opening of the U.S. telecommunications market, beginning with anti-trust suits brought in the late 1940s and culminating in the break-up of AT&T. More recent telecom mergers have added to the repository of anti-trust analysis in both Europe and the United States.

Moreover, anti-trust law offers three additional attributes that make it a suitable candidate as the starting point for legal analysis. First, it proceeds on a case-by-case basis, generating precedent but not detailed regulatory rules. Second, it is transnational, with both Europe and the United States actively involved in considering the dimensions of its application. Third, its tradition has, at various times and in various places, considered aspects of competition beyond only narrowly-defined antitrust doctrine (although the legitimacy of non-economic considerations itself has sparked a spirited debate, which is outside the scope of this paper.)

As noted above, the United States enjoys a long tradition of using competition policy to provide for "open" attributes of communications networks. The original antitrust action filed by the United States Government resulted in a settlement agreement including a provision for the interconnection of Bell with non-Bell networks [37]. The subsequent anti-trust action that led to the break-up of AT&T on 1 January 1984 also included provisions ensuring that all long-distance companies would be able to exchange traffic with the local networks of the Bell companies [38].

Three recent decisions illustrate the potential for the application of competition policy principles to the issue of "open" networks. The proposed merger between WorldCom and Sprint was to have combined two of the world's largest providers of top-level Internet connectivity or, to use the vernacular, two of the largest Internet backbones. The European Union concluded that the market for the provision of Internet services is hierarchical and that access to top-tier backbone providers is, therefore, absolutely necessary if other Internet backbone or service providers are to provide global connectivity to their customers. In the Commission's (2000a) view, the proposed merger would (if it had been allowed to be consummated):

"[r]esult in an entity of such absolute (more than [35-45] percent of the market) and relative ([several] times larger than its closest competitor) size, that this would enable the merged entity to behave independently of competitor and customers. For instance, it will be able to increase prices to customers or to impose its own standards upon the industry. Its ability to diminish its rivals' quality of service at any time through selective degradation will make it possible for the combined entity to discipline the market. It can therefore be concluded that [this] transaction will lead either to the creation or the strengthening of a dominant position for the provision of top-level or universal Internet connectivity."

Accordingly, the European Commission prohibited the merger, even after the merging parties offered to divest the Sprint network (European Commission, 2000b).

In a similar vein, the U.S. FCC intervened when faced with the matter of instant messaging on the Internet. The issue arose in the context of the AOL-Time Warner merger when, drawing on principles of network economics discussed above, the FCC scrutinized AOL's refusal to make its instant messaging service interoperable with the instant messaging services of any of its competitors. In the FCC's view, instant messaging is a market with strong network effects and AOL was "by far" the leading provider of instant-messaging services [39]. That market position could be strengthened by the addition of Time Warner's customer base and valuable content in a manner that, the FCC concluded, could give the merged company a dominant position in the provision of new advanced messaging services, thus making it easier for AOL-Time Warner to "leverage the network effects" of current instant messaging into new, more advanced services. To counter that possibility, the FCC imposed a "prophylactic" condition that requires interoperability whenever, in the future, such advanced instant messaging services are deployed [40].

In both of these cases of potential anti-competitive impact on competitors, the effect of government intervention — although it did not mandate the exchange of Internet traffic through peering or private arrangements and did not mandate interoperability in the current marketplace for instant messaging — was to promote openness.

A third example concerns the nature of vertical restraints that are often described as creating the power to control a "bottleneck". The ability of network providers to leverage market domination of a downstream network into advantages in upstream markets is not a new concept. The antitrust action brought by the U.S. Justice Department against the Bell system in the late 1940s, for example, asserted that the Bell system was using its monopoly to harm competition in the equipment-manufacturing sector [41].

In recent years, the paradigmatic example has been waged in the context of cable television systems [42]. In 1992, Congress enacted provisions to limit the ability of cable networks to harm against upstream content providers by "address[ing] the fear that cable system acquisition of program suppliers (or vice versa) may lead the system owners unfairly to favor the programs of firms in which they have an ownership stake over the programs of suppliers they do not own" [43]. The issue of bottleneck cable control arose in the context of the Internet when objections were raised to the decision of cable systems to require residential customers to choose the single Internet Service Provider (ISP) that was affiliated with the cable network. Again, fears were raised that "bottleneck" control of the transmission plant would be used to discriminate against competitive ISPs, to raise prices for ISP services artificially and to limit the kinds of content that would be available to residential customers [44].

Arguments over so-called "open access" quickly prompted recitation of policy bases noted above. Some advocates for open access argued that the e2e engineering principles of the Internet — and the innovative, diverse environment it spawned — justified governmental intervention to assure customer choice of ISPs (Bar et al., 2001; Lemley and Lessig, 2001). Others argued more narrowly from anti-trust principles that open access should be imposed only in merger cases involving cable networks that are vertically integrated with content providers (Hausman et al., 2001). And, of course, others argued, on the basic of economic principles, that mandated openness would adversely impact investment incentives.

Governmental authorities in the United States declined to state any rule of general applicability. But, in the context again of the AOL-Time Warner merger, the FTC and FCC imposed conditions that required the merged AOL-Time Warner cable systems to permit at least three unaffiliated ISPs to access its cable modem customers, to not interfere with the content provided by such unaffiliated ISPs, and to not attempt to bias its customers choice of an ISP [45]. That rule was not applied to the subsequent acquisition by Comcast of AT&T's cable systems [46].

5. The multiple faces of open v. closed networks

In sum, multiple disciplines, including engineering, economics, social philosophy and the law, offer competing bases on which a policy maker can decide whether the government should intervene to order a next-generation network to be opened. Even if there were an easy tool for a public policy maker setting a legislative or regulatory rule to choose alternatives within a discipline (say, e2e v. the traditional Bell preference for closed networks), there is no public policy syntax that permits easy resolution of arguments that proceed from different disciplines (say the interest in free and diverse speech v. the need to encourage investment).

On this basis alone, mapping an appropriate public policy approach to the question of openness would likely generate considerable confusion. But the question is not, in any event, that easy.



IV. Perspectives on openness

To say that one network is "open" and another is "closed" is — outside of the context of truly closed governmental or corporate networks — to say very little indeed. That is because different functions of the network can be open or closed and very few networks serving public subscribers are all one or the other.

This paper proceeds to analyze the network from the perspective of specific users, including the network provider itself. This should not be viewed as inconsistent in any way with that form of analysis that distinguishes between the physical, logical and content layers of a network [47]. Indeed, the concerns of various users may track specific layers — residential end users for example may be concerned exclusively with their access to the content layer. The slightly different formulation used in this paper — which characterizes an actor's perspective on network access — is employed merely to focus, as a policy debate would be likely to do, on the motivation and contentions of those who seek to use the network.

A. From the perspective of the end user

From the perspective of an individual or corporate end user, a series of issues affect the extent to which the network appears to be open. First are criteria for network access. Traditional telephone networks were required, because of their status as common carriers, to provide their service to all potential customers who could pay the approved tariffs. Consumer-oriented commercial networks follow the same approach. There are, however, social questions involved in this question: Is a network "open" if socially or racially insulated portions of society cannot afford access? Is it open if its operation makes it usage difficult for people with disabilities? These kinds of questions, which usually proceed under the rubric of "universal service" or "universal access" and involve a series of important social and distributional questions, are a form of the openness inquiry that lies outside the scope of this paper.

Once on the network, is the end user restricted in the manner in which she may participate? For example, does a network restrict her usage only to personal, and not professional, content? Are there limitations on the applications she may run or on the content — like streaming video — that she may access? Is the network biased in some non-transparent fashion to certain content (perhaps to content that bears a commercial relationship to the network provider)? And can she choose what sort of device to use to access the network and, in general, to connect hardware to the network not supplied by the network provider itself? In each of these ways, an end user may find herself on a network that does not permit her to use it as she wishes [48].

B. From the perspective of a competitive network provider

A competing network provider may be thought, at first glance, to have little legitimate interest in the openness of a competing network — except as a basis for potential competitive advantage. In fact, governmental policy has long found circumstances in which, for the perceived benefit of competition, public policy takes cognizance of the ability of competitive network providers to have another's network open to them.

For example, does a competing network have the legal right to exchange traffic with another network? [49] Does it have access to any of the technical attributes of the first network — including the ability to transmit information over the other network's facilities? [50] Does it have the right to purchase the services of the competing network and resell them under its own name and associated with its own services? [51] Does it have the right to lease portions of a competing network and combine them with its own facilities in order to provide service to a wider geographic region? [52]

C. From the perspective of a content/software provider

A content and/or software provider stands in a slightly different position than a competing network provider because, depending on the extent of vertical integration in the network at issue, the content and/or software provider may or may not be the direct competitor of the network provider.

Issues that arise in this context include the following: From the perspective of a commercial provider of content/software does it have the ability to place its content on the network or, as in cable television systems, must it seek the "editorial" consent of the network? Can its applications run freely on the network? If it provides hardware, can the hardware be connected to the network or supplied to the network owner?

D. From the perspective of the network owner

Limitations, or qualifications, on the identity of the network owner itself may be viewed as implementing openness policies as well. For example, industry conditions may be of relevance as where, for example, incumbent telephone companies were guaranteed cellular-telephone spectrum in the United States in the 1980s, broadcast television licensees were guaranteed the ownership of digital television spectrum or, to the opposite effect, where cable television and telephone companies were barred from owning both sets of networks in the same geographic area [53].

Issues may also arise with regard to foreign ownership of next-generation networks, in particular where the foreign ownership involves a foreign governmental entity. At a congressional hearing held in the fall of 2000, the Federal Bureau of Investigation (FBI) warned that "there may be no practical way to conduct lawful surveillance effectively and securely if the facilities [to] process U.S. communications are located outside of the United States," that "[o]wnership and control of U.S. communications networks give a foreign government the capacity to gain relatively easy access to confidential information about the targets of U.S. national security and law enforcement investigations," and that "[o]wnership and control of U.S. communications networks could also provide a foreign government with the ability to direct key employees to utilize routine monitoring capability to access confidential private communications and data of U.S. corporations and citizens communicating over the network" (Parkinson, 2000). Thus, despite the general presumption that ownership by foreign firms of WTO member countries is to be permitted, the case of foreign acquisition of a key U.S. telecommunications or Internet network would clearly require careful scrutiny by the FCC, which retains the authority to block such an acquisition that would not be in the public interest (Kennard, 2000). A recent illustration of this policy in action was of course the FCC's probe of the acquisition of U.S. wireless provider VoiceStream by Deutsche Telekom, given the German government's significant stake in Deutsche Telekom (Ames, 2001).



V. Application of the bases of decision-making to perspectives on openness

This listing of perspectives on openness — and the identification of issues that can arise at from each perspective — is designed merely to set the stage for further analysis. One preliminary observation can, however, be noted from the recitation of these categories alone. The bases for decision on openness that were identified earlier are not all equally applicable to each perspective. For example, the narrow version of the e2e principle may say very little about the advisability of a set-top box located at the edge of the network that provides preferential treatment to content originating from specific content providers. Accordingly, it is worth noting issues that arise in four contexts as a means of illuminating some of the complexities inherent in the open/closed distinction.

It is important to understand that the means of analyses that should be applied by the creators of public policy making may be different from a factual description of how policy is actually made. The disciplinary approaches discussed earlier suggest what factors should be considered by policy makers. But consideration of the following four examples also suggests potential methods of analyzing the manner in which the policy making system actually operates.

1. The incumbent telephone network circa 2002 — economic contentions between horizontal competitors

Are the incumbent telephone networks in the United States "open"? In very significant ways, they are. End users are free to place any content on the network and free to receive any content (that is not otherwise illegal); as a dial-up platform, it is open to any Internet Service Provider; traffic must be exchanged with any other telephone provider — eliminating the difficulty that plagued telephony in its early years when a small network could not offer its customers comprehensive connectivity. In addition, a competing network can buy and then resell the incumbent's services.

The current controversy — and it is a very big one — focuses on the "openness" by which networks that are horizontal competitors can lease elements of the Bell networks in order to provide, for example, competing DSL services. (The same is true for voice traffic). In the parlance used earlier, this issue is presented as a question of openness from the perspective of a competing network provider. The Telecommunications Act mandated such "openness" in certain circumstances in order to counter the residual monopoly power of the Bells. This is not surprising; the European Union has adopted similar analysis [54].

But what may be surprising is that now, almost seven years after the fact, a fierce controversy has arisen about whether this "openness" is counter-productive to the goal of broadband deployment. As noted above, SBC, echoing the classic economic argument about investment incentives, not only says that it does; it blames this policy for its poor earnings report at the beginning of October, 2002 and for recent layoffs [55]. That contention has not been universally accepted. Competitive firms — like the Florida competitive company Z-Tel — reply that wholesale provisioning is profitable for the Bells and necessary to the goal of bringing competition in the local markets (Z-Tel Technologies, 2002).

A more discrete application of this general dispute concerns the question whether "openness" principles should continue to be applied to those elements of the Bell network that used to provide DSL or other forms of broadband access. To date, DSL has, of course, competed against incumbent cable companies to provide "fast Internet" service to residential customers.

It is the contention of Bell companies that, because DSL equipment is not part of the traditional voice network and because the Bell companies do not have the largest market share in the "fast Internet" network, there is no compelling public policy rationale to permit such equipment to be "open" for the use of Bell competitors. To the contrary, they contend that this is an instance in which such openness creates disincentives for investment in the deployment of next-generation networks. Similarly, focus has been placed on the installation by the Bells of new "fiber" facilities to carry broadband communications — fiber that was not part of the traditional, copper-based infrastructure and that may be seen, therefore, as different than the facilities to which competitors have been given access.

The view to the contrary favors the continued ability of competitors to gain access to incumbent facilities for inclusion in competitors' broadband offerings on the basis that Congress made a broad public policy judgment in the Telecommunications Act of 1996 that competitors should be able to access portions of the Bell network. This congressional judgment is said to be premised upon powerful network effects stemming from the continued value of the incumbent telephony network, including those portions that are used to provide broadband services.

A full discussion of the intricacies of this issue is beyond the scope of this paper. Some preliminary observations, however, can be made. First, issues of free speech or freedom in any non-economic sense have little pertinence to the resolution of this question. The Bell networks are "open" to speech of all kinds and the Bell networks have never classified themselves as "editors" of information that would justify, even in their own eyes, any limitation on the speech of others. (Moreover, there are substantial alternative methods of conveying speech on the Internet, including the open dial-up platform on the Internet itself, which may be relevant in judging the impact of even a closed network on free speech values.)

Second, non-economic engineering issues would seem to have little relevance. The e2e principle in its narrow version focuses on functionality of network components not, it would seem, on their ownership or the extent to which usage of them is shared between competitors. Although post-September 11 concerns have been raised about the access of competitors to networks, they do not, at the moment, appear to offer any justification for distinguishing between DSL and traditional voice portions of the Bell networks.

Third, there are abundant economic rationales on which to decide this issue. For example, the European Union requires incumbent telephony companies to offer access to their DSL facilities on the ground that they have market power in the product market of fast-Internet services [56]. In the United States, however, the Bell companies lag behind the cable companies in the provision of fast-Internet [57] and, therefore, the European reasoning could be seen as inapplicable. A contrary view would argue that, whatever the market share in the DSL market specifically, the incumbents enjoy continuing advantages in the provision of DSL because of their past position as monopolies.

A related issue concerns pricing. Even if open access is justified, the very important question arises of what competitors should pay to gain access to the elements of a network. This issue, in the context of the application of the Telecommunications Act generally, has already generated years of controversy and been reviewed by the United States Supreme Court [58]. Put simply, if competitors pay too little, then they may have disincentives to invest. On the other hand, if competitors pay as much as the wholesale costs that the incumbent would bear if it were to self-supply DSL access, then the incumbent would seem not to lose any revenue or incentives to invest.

One might conclude, therefore, that here a policy maker could confine her consideration of disciplines to the realm of economics. That would, of course, lead to the question: To what extent do policy makers considering the evolution of next-generation networks have and/or use effective tools for evaluating competing economic claims?

That may be the manner in which this issue should be decided, but it is worth noting two related points about the manner in which it is, in fact, decided.

A legal or political system may purport to answer the question — but perhaps not by supplying a reasoned justification. That is, it may be that the text of the Telecommunications Act, as a matter of statutory definition, resolves the issue without invoking any rationale except for principles of statutory interpretation. Viewed from this angle, the legal system thus demonstrates an important limitation on the attempt to build a set of principles on which the question of open/closed networks should be decided. Answers derived from legal principles may have force but lack an articulated justification that extends beyond the manner in which the law itself allocates authority between branches of government.

And that, in turns, implicates the manner in which the political process — the progenitor of the text to which principles of statutory construction are applied — actually applies any discernible discipline or discipline(s) to the resolution of next-generation network issues [59]. Thus, issues of legal/political outcomes, such as the study of the influence of interest groups, may be a fruitful avenue of study to the extent that one wishes an accurate descriptive of what forces have actually influenced the construction of policy. There is nothing surprising about that notion — it is a standard approach of political science. The next two sections suggest, however, that the concept of the "Internet" as a basis for decision may have brought something new to the game.

2. Cable networks — the substantive use of concepts of free expression

As noted above, in the special case of the AOL-Time Warner merger but not more generally, federal regulatory authorities required that the merged AOL-Time Warner cable systems permit at least three unaffiliated ISPs to access its cable modem customers. This issue can be viewed a presenting the classic "bottleneck" circumstance in which a downstream chokehold interferes with the ability of upstream entities to reach customers. As such it can be analyzed as a matter of economic theory and decided, either way, on that basis. In terms of the perspectives on network openness, this captures the perspective of a content and/or software provider.

A different perspective, however, obtains from the perspective of the end-user who wishes to participate in legitimate activities by, for example, accessing content, running specific applications, or using the network for a variety of purposes. As noted above, high-technology groups, which in general oppose government intervention, have nonetheless endorsed at least the potential for government action in order to ensure that customers of cable modem services "have unrestricted access to their choice of content" and have the general right "to run applications of their choice."

The disciplinary basis for those end user "rights" can be found in both the notion of Jeffersonian democracy (although not the precise scope of the First and Fourteenth Amendments, which apply only to governmental action) and the broader version of the end-to-end principle. Jeffersonian principles would seem to dictate a preference for systems that permit free expression — for the vindication of both social and political goals. The broader version of the end-to-end principle extols decentralized freedom (intelligence at the edges of the network) over centralized control (intelligence embedded in the center of the network).

What is a policy maker to do with these substantive policy goals if she wishes to grapple seriously with a circumstance in which, it can be assumed, the network provider asserts independent economic reasons to curtail these social and political "rights" [60]. Or, to put it another way, what tools are available to decide which is controlling in a particular fact situation — the "right" of freedom at the edge of the network as embodied in the decisions of an end user or the economic "right" of a owner of a network to control its use?

Current policy making tends to default to economic analysis. So, for example, the FCC decision in the AOL-Time Warner case focused its treatment of instant messaging on the application of principles of network economics, not on any inherent "right" of users to communicate freely with each other.

That kind of analysis could easily become a barrier to full understanding of the implications of the Internet. If non-economic rationales lack any decisional power, it would be better for regulators to simply say so and clear the field of non-economic contentions based on the "freedom" of users. If, on the other hand, there is normative power behind the concepts of Jeffersonian democracy and the application of the end-to-end principle that exceeds the calculation of economic benefits, then it is necessary that further research be conducted to examine how a policy maker could, in actuality, justify a decision in terms of those concepts [61].

Moreover, such research could reasonably begin to explore the power of the Internet not as a technology, but as an idea of independent political significance in the world in which public policy governing next-generation networks is to be made.

3. The Internet itself — architecture as an idea

The Internet itself is, of course, the classic open architecture network. But it may or may not remain so. To some in the e2e community, commercialization efforts, such as the introduction of quality of service standards for certain kinds of data (say video) from certain kinds of users (say corporations), threatens the existence of the architecture that it credits for fostering innovation at the edges of the network.

Consider two limitations on "openness" even in this paradigmatic system. There is no governmental requirement that networks exchange traffic through peering arrangements — unless one concludes that the basis for the European Union's rejection of the WorldCom/Sprint merger was an implicit requirement of peering or, at least, a rejection of those market structures in which peering would not be to the "voluntary" benefit of all large backbone providers.

Nor is there any requirement that applications run on the network be "open" to end users. Witness AOL's successful contention that it should be able to maintain its Instant Messaging as a closed system if it wishes to do so — a position marred only by a slightly perplexing decision of the FCC requiring that certain future versions of IM be open if they ever come into existence.

The irony here is that it is the idea of the Internet, and not its actual governance, that largely fuels the policy bases for governmental intervention into the development and operation of open networks. The Internet is open mainly as a manner of consensus, not legal rule. But the idea of the Internet — what it came to embody to second-generation policy makers as a means of reaching both social and economic outcomes — appears to have attained significance as a factor impacting policy in its own right. Another way to consider this would be to say that, at various times and to various degrees, the idea of the Internet, as expressed in the end-to-end principle and the competing visions of freedom, captured the attention of policy makers in a manner that competed for influence against the normal forces of interest group and bureaucratic politics.

If second-generation policy was shaped by the idea of the Internet as a transforming force independent of normal political interests, then it would be worth exploring the circumstances in which the power of that idea attained independent political significance. How did this idea act as an organizing principle for forces that had not been previously engaged in next-generation network policy making? Did traditional forces learn to live with, succumb to, successfully oppose or ultimately co-opt the force of that idea?

Understanding the manner in which the idea of the Internet may have upset the usual politics of policy making over the past decade therefore would be important in offering a prescription for third generation policy. Just to take one example, if it were to be concluded that the independent significance of the Internet as an idea, not just a technology, contributed to policies designed to open closed telecommunications networks (and that those policies have met or could meet with success), then it might also be concluded that there are beneficial spill-over effects that could arise from the way that future issues, like the use of WiFi, are decided. Further consideration should be given, in other words, to the possibility that policies that further the power of the Internet as an "idea" further an independent policy goal [62].

4. The European perspective

Even a preliminary examination of the application of "openness" policy would be incomplete without consideration of an international perspective. That is particularly important because the European Union has created an Access Regime that can be applied across technologies, that has a disciplinary starting point, and that encompasses multiple disciplinary perspectives.

In 2002, the European Union issued its Access Directive, scheduled to take effect in May 2003. The purpose is to replace existing regulatory requirements promulgated on a technology-specific basis, including requirements for unbundling the local telecom loop, with an overarching approach that applies on a technologically neutral basis to all electronic communications networks. In so doing, the Access Directive takes a specific disciplinary approach. It embraces the notion that government should "regulate only where absolutely necessary" and pegs the justification for such action expressly to the requirements of competition policy (Liikanen, 2002b).

At the same time, the new policy expressly notes the existence of other policy bases. For example, it states that "[c]ompetition rules alone may not be sufficient to ensure cultural diversity and media pluralism in the area of digital television," a concern sufficient to justify the "obligation" for Digital TV providers to provide "access on fair, reasonable, and non-discriminatory terms, in order to make sure that a wide variety of programming and services is available" [63]. The Access Directive is careful to note that Digital TV is an example, but not the exclusive province, of the principle of "non-discrimination," which can be applied generally to electronic networks in order to ensure that entities "with market power do not distort competition, in particular where they are vertically integrated [companies] that supply services to [companies] with whom they compete on downstream markets" [64]. Importantly, however, the non-discrimination rule does not appear to be limited only to firms with market power. Rather, the Access Directive is careful to reiterate the obligation of "openness" "to the extent that it is necessary to ensure accessibility for end users to digital radio and television broadcasting services" [65]. At the same time, the EU reiterates that pre-existing requirements of "openness" would continue, subject to immediate market review. Those requirements include access to incumbent telephone networks [66].

With this foundation, the EU has begun to move forward in the application of its Access Directive. Currently, the Commission has launched a consultation to consider the existence of "open platforms" in both digital TV and 3G mobile communications (European Commission, 2002). That process is based on a staff working paper that focuses on the use of proprietary technology in set-top boxes or mobile devices that could limit end-user access to content. Proposing that the Commission monitor the creation and use of open standards at these network edges, the staff notes that "[t]he new regulatory framework requires an examination of the extent to which interoperability and freedom of choice have been achieved in Member States by no later than July 2004" (European Commission, n.d.).

The EU Access Directive thus provides an important early example of a governmental policy that attempts to work its way carefully through the disciplinary thicket. The initial resort to competition policy provides a doctrinal framework through which to view both engineering and economic theories. But the careful reservation of the application of non-discrimination rules in order to vindicate end-user "freedom of choice" of content suggests a regime that has taken seriously, as well, the non-economic goal of diversity of speech and may seek to apply it in a considered fashion at the perspective on network use that is directly implicated by the policy goal.



VI. The interdependence of openness

The essential goal of this paper is to demonstrate that the question of openness in the operation of next-generation networks is more complicated than might be thought. Openness is more complicated because (i) the rationales both for and against openness stem from a widely-dispersed set of disciplines that do not necessarily share common reference points guiding a policy maker towards inter-disciplinary agreement; and, (ii) a next-generation network can be both "open" and "closed" simultaneously.

Thus, it would seem that, before a policy maker could decide that a next-generation network would be open, it would be necessary to know (i) to what extent and at what level a network should be treated as open or closed; (ii) which disciplines are properly invoked to analyze the claim that the government should take action; (iii) whether, in fact, governmental action is required to maintain or require openness; and, (iv) what should be the precise characteristics of governmental action, if any action is taken.

There may be a short cut through this analysis. An understanding of the diverse argumentation and the different perspectives on openness may allow a policy maker to construct a set of policy presumptions that would ease the resolution of many issues and offer a construct — if not a solution — to the resolution of problems that remain.

Consider the following set of rules, proffered merely as an example of how such a template could be constructed and for purposes of stimulating discussion:

  • Engineering principles are best understood by engineers and therefore, to the maximum extent possible, governments should find ways to refer engineering issues to groups or formal bodies of engineers for resolution by consensus. Given the history, however, of entities employing engineering arguments as an initial, and unjustified, tactic in order to slow competitors, governments should first employ engineering expertise to assess the facial validity of private sector engineering claims and should be vigilant to ensure that government-sanctioned engineering processes are not manipulated for the commercial gain of one party over another. This approach asks engineers to choose in the first instance between competing engineering issues (like the question of whether the Hush-a-Phone plastic device would harm the telephone network) and thus leaves to engineers the core, but not ultimate, task of applying engineering principles.
  • Private sector actors themselves are best positioned to understand the economic incentives that they face and, therefore, economic decisions should be left in the first instance to network owners. Because economic incentives are supposed to advance the self-interest of entities, there is no need for governments to establish panels of economists akin to the groups of engineers that might, for example, focus on open network standards. The presumption that private actors should decide the treatment of the networks that they own is rebuttable where the network owner has procured improper market power that is damaging to competition, including the innovation that competition stimulates. (The effect of this rule might be to treat economic considerations as a subset — and perhaps sole substantive driver — of competition policy).
  • Neither governments nor the private sector have any advantage in determining the best means of pursuing social freedom. Government does not, because the core values of free speech assume (although they do not require) that diversity in debate will include anti-governmental sentiment. Private sector actors have no advantage because in the realm of diversity of thought they have no greater standing than individuals (although this formulation does not suggest that they have any less.) In the area of social freedom, therefore, the presumption should be in favor of the "speaker" — for diversity of thought is peculiarly the product of multiple individual decisions. Similarly, the e2e principle, as it is applied in its broader version to promote innovation at the edge of networks rather than more narrowly focused engineering principles, should be applied to promote innovation in general and thus the greatest freedom of individuals to innovate. Acceptance of this policy preference would still require a policy maker to decide what, if any, weight she should attach to this goal.
  • Separate from the policy bases discussed above is the question: At what level of government is the decision appropriately made? One example comes from the American doctrine of federalism (i.e. when is a decision appropriate for federal and when is it appropriate for state and local authorities?) The Telecommunications Act of 1996 (and subsequent FCC decisions) led to extensive debates about the appropriate role of state authorities in administering a federal statute. The perspective of industry players may shift dramatically over time — witness the evolution of views of competitive entrants into local telephone markets who originally called from extensive federal regulation but who have, much more recently, been defending the preservation of local regulatory authority. Although this question of allocation of power is ostensibly unrelated to the substance of policy analysis, the choice of forum can, in fact, have enormous implications for the outcome of policy decisions. Similarly, in Europe decisions have been — and continue to be — made against the background of discussion about the appropriate authority of the Commission vis-à-vis the member states.
  • Finally, there may be policy areas where the government, as a result of its own governmental functions, is an expert. Consider the question of the circumstances under which foreign governments (or foreign companies in which foreign governments have substantial holdings) are permitted to acquire U.S. networks — and the subsidiary issues of the effect on national security and the pursuit of U.S. trade policy. Those are issues on which it may be thought that the government itself has special knowledge and in which, therefore, any policy regime must include a means by which government expertise can be considered and evaluated.

Application of this template might simplify but assuredly would not end, the inquiry. Consider the quite common circumstances in which end users (who we will assume to be non-commercial consumers) sit on one side of a network and frustrated content providers (who we will assume to be motivated solely by commercial considerations) sit on the other. Assume also a network provider that refuses to allow its end users to access all of the online applications that they wish to reach and, in the particular instance, does not permit end users to engage in online gambling even though a willing content provider wants access to that customer to offer an online gambling site.

The interest of the content provider is economic and can be dealt with as a matter of competition policy which, in classic terms, would look to see if the network were an essential facility or, in a less stringent sense, a bottleneck that would allow its owner to unfairly leverage its control into upward vertical integration. The interest of the end-user is not principally economic — rather she asserts a freedom to swap content with whomever she likes.

If the competition policy argument fails, because the network provider lacks market power, does that mean that the end user's free speech argument necessarily fails? If it does, then it means that issues of freedom of speech and to innovate are nothing more than subsets of economic theory (which themselves have been sublimated to competition policy in the above template). If, however, principles of social philosophy actually have independent power, then it will be necessary for a policy maker to establish a means of mediating between social and economic arguments.

A less abstract version of this statement might be as follows: for all of the talk of freedom to speak and innovate, there is little to demonstrate that such considerations have been of decisional significance to date in the resolution of "open" versus "closed" network issues, at least outside of the world of cable television in the United States, or television and radio services in the EU (and, of course, upstream programming interests may provide an economic rationale in both instances). Such an approach may be quite correct when the issues are primarily economic, such as the request of a competing network to exchange traffic with another network provider. But where the interests of end users as a definable group are implicated (and not just treated as a rhetorical flourish to argumentation by various upstream commercial entities), that approach does not actually inform the analysis because it does not treat the issues of social philosophy with the care they would deserve if they were to stand as an independent decisional principle. In this respect, the recent EU Access Directive could be seen as an early step towards analysis that directly confronts this question.

In addition, the debate over openness sometimes fails to identify with precision the particular perspective from which, and disciplinary point at which, the issue resides and thus fails to weed out issues that do not, in fact, involve particular disciplines. So, for example, consider whether a wireless data network is open or closed. Suppose that the end users can have access to any content on the Web that they seek but that competitive networks cannot gain access to the wireless transmission stream to offer competing proprietary content. Whether such a network is open or closed may depend on whether one wishes to examine the interests of end users to reach what is generally available or competing networks' ability to add to the mix of content. This paper has called those different interests different "perspectives" on network use.

And, even within a single network perspective, the precise disciplinary interest that is being advanced must be correctly identified. To add to the same example, suppose end users of a 3G device can access any content but can do so only with a wireless device supplied by the 3G network provider. Unless a strong contention can be made that mandated use of devices substantially impairs the social freedoms of speech and innovation, the network can be seen as "closed" to alternative devices, but not in a fashion that implicates the same social values that would be implicated were end users barred from reaching certain content. (Of course, one could vary the hypothetical to implicate social values if the hardware device included technology that preferred certain content over others.)



VII. Conclusion

The Federal Communications Commission now has pending, and is expected to determine in the coming months, a series of proceedings that directly address the question of open next-generation networks. Among the questions are the terms under which the unbundled elements of incumbents telephone networks will continue to be available, the appropriate characterization of the Bells' market position in the provision of broadband services, and the appropriate regulatory classification of DSL and cable modem services [67]. It is widely expected that the FCC will take a less expansive view of the openness principle than either it, or similarly situated regulators, have taken in the past.

Would such an outcome obviate the need to consider the questions discussed in this paper? There are four reasons to think not.

First, the market power in incumbent networks has not entirely disappeared. The copper network that forms the last mile to most American users may look old-fashioned sometimes but it retains great value (Krim, 2003). That means that competitors will continue to press for access to it — even if, in the short term, current policies in the United States are reversed or substantially modified.

Second, consolidation in the telecommunications sector is likely to lead to greater, not less, market power in leading firms. If anyone could afford consolidation it would be here today. But, with stock prices lower in the United States and the debt overhang on European telecommunications firms, they can't — yet. When consolidation does come, issues like those that arose in the WorldCom/Sprint merger — dominance of the Internet backbone — could become important again.

Third, technological alternatives to existing networks could prove to be less robust than many hope. What if, for example, 3G fails as a commercial application, or if other wireless alternatives, such as ultra-wide band (UWB) or WiFi, fail to offer an alternative to the local loop? Would that, combined with other possible trend lines noted above, lead to markedly fewer network providers and would that kind of market concentration give rise to "openness" concerns?

Fourth, network providers may not voluntarily open networks to content, applications and user devices. Such behavior could lead to the kinds of circumstances of which high-tech entities are currently complaining, including restrictions on or bias towards content and the ability to run applications. Those types of restrictions, as next-generation networks become more important, could spark increased emphasis on the applicability of principles of free speech and innovation, including the potential free speech interests of those who assert that they are performing editorial services by choosing which content to place on their networks.

In sum, however this year's proceedings in front of the FCC turn out, public policy will surely be advantaged if the issue of "open/closed" networks is understood to involve several issues that implicate distinct rationales for governmental decision and different perspectives on network access. The core goal of this paper is to advance an initial taxonomy of such policy bases and network perspectives in order to permit policy discussion to operate in a more discrete and, one hopes, more insightful manner.



Addendum — recent developments

The FCC Unbundling Order

On 20 February 2003 the U.S. Federal Communications Commission issued its decision concerning the obligation of incumbent telecommunications companies to make available their networks to competitive providers. That decision, which was soon awash in political controversy far beyond the scope of this paper, was not accompanied by the formal Order or Commission analysis, which will be forthcoming. Indeed, Commissioner Adelstein noted his discomfort in "voting on this item before the [Commissioners'] offices have seen the draft order," a result of the fact that those offices "reached some agreements on major issues at the eleventh hour — and I mean that literally, around 11:00" pm the previous night [68].

Each of the commissioners issued a separate explanatory statement and the final result was far from unanimous — four of the five Commissioners (all but Commissioner Martin) dissented in part. Thorough analysis must, of course, await review of the final order and its accompanying textual explanation but the statements issued on the day of the decision do permit the following brief comments:

1. Broadband investment incentives

The Commission concluded that incumbent telecommunications companies will not have to provide competitive access to "new build/greenfield [Fiber-to-the-Home] loops for both broadband and narrowband services," nor is there any "unbundling requirement for overbuild/brownfield [Fiber-to-the-Home] loops for broadband services" nor for "the packet-switching features, functions and capabilities of incumbent ... loops" [69]. That means that new fiber facilities to the home are entirely exempt from open access requirements and that access to hybrid fiber/copper loops (composed of partly new and partly pre-existing facilities) will be limited "to preserve existing access rights but refrain from imposing new unbundling obligations on upgraded hybrid loops" [70]. In taking this position, the Commission decided to "endorse and adopt in total the High Tech Broadband Coalition's proposals for the deregulation of fiber to the home and any fiber used with new packet technology" [71].

These conclusions appear to have been based heavily — if not solely — on the contention that freedom from access regulations would encourage investments by the incumbents in broadband facilities. Chairman Powell explained that "[t]oday's decision makes significant strides to promote investment in advanced architecture and fiber" [72], and Commissioner Abernathy said that "[w]e have taken bold action to restore incentives for carriers to build next-generation fiber-based facilities," based on a "record [that] suggests that the uncertainty regarding possible broadband unbundling obligations has chilled investment substantially" [73].

The Commission's de-regulatory approach to broadband facilities was not, however, unanimous. Commissioner Kopps attacked the decision on straightforward economic grounds. Charactering local loops (fiber and copper) as "the ultimate bottleneck facility," Commission Kopps argued that the Commission's decision would operate "on a nationwide basis without adequate analysis of the impact on consumers, without analyzing different geographic or customer markets" and without consideration that "in some markets such as the small and medium business market, there may not be any competitive alternatives if competitors cannot get access to loop facilities" [74].

Complete understanding of the Commission's reasoning must, of course, await issuance of its Order and analysis. At first blush, however, the ascendancy of the economic theory proposing the importance of creating incentives for investment in next-generation networks seems obvious.

2. The end of line sharing for the provision of data services

The issue of line sharing arises when, for example, a residential customer wishes to obtain traditional voice service from an incumbent telecommunications company (say Verizon) but wishes to purchase DSL service from a competitive data provider (say Covad). It only takes one copper loop to provide both services (DSL services being provided on the high-frequency portion of the loop) and so, under the FCC rules previously in existence, Covad would have the right to require Verizon to share the loop so that it could provide DSL services to the requesting residential customer.

The FCC decided to abolish that open-access right, over the course of a three-year transition period. Which policy was used for this decision? That remains uncertain because none of the Commissioners who voted to adopt this measure explained in their statements why that was the correct outcome.

Also notable was the fervent criticism of the abolition of the line sharing by two commissioners who fervently favored the de-regulation of broadband facilities. Their analysis can be seen as attempting to define a limit to the "investment incentive" economic reasoning — even if not the limit apparently adopted by a majority of their colleagues. The critical distinction: "line sharing rides on the old copper infrastructure not on the new advanced networks that we are attempting to push to deployment" [75]. Because the copper infrastructure already exists, "there is simply no copper upgrade that the incumbents are deterred from making" [76]. Thus, in the view of Chairman Powell and Commissioner Abernathy, there is no logical application of the "incentive" analysis to the copper infrastructure in question. Moreover, both found significant benefits from this form of open access. Chairman Powell said that "[l]ine sharing has given birth to facilities-based broadband telecommunications carriers" that have "in turn, provided a valuable source of broadband transmission services to independent internet service providers ... " [77] and that has also brought consumers "lower prices for new broadband services" [78].

In addition, Commissioner Abernathy said that "the record suggests that line sharing spurs [incumbent] investment in DSL, rather than retarding it." With an estimated "40% of DSL providers us[ing] line shared inputs," Chairman Powell explained why, in his view, the continuation of line sharing would boost, not impede, investment: "[T]he continued availability of line sharing and the competition that flowed from it likely would have pressured incumbents to deploy more advanced networks in order to move from the negative regulatory pole [of open access] to the positive regulatory pole [of de-regulation], by deploying more fiber infrastructure" [79]. In other words, Chairman Powell argued that the line sharing decision was inconsistent with the goal of broadband de-regulation because it would decrease the incentive for incumbents to move from regulated copper to de-regulated fiber.

One awaits with great interest the Commission's explanation of its decision to abolish line sharing. Chairman Powell's and Commissioner Abernathy's critiques contend that the line sharing decision cannot be squared with the economic theory of increasing incentives for investment. Is that really true (keeping mind that Commissioner Abernathy was careful to note arguments against line sharing)? If it is, does that mean that the majority has adopted a flawed economic analysis (and the seeming tension between line sharing and broadband deployment is sure to become fodder for appellate lawyers)? Or does it mean that the majority, in the circumstances of high-speed Internet services at least, has adopted a different economic rationale, applied the same rationale with a different understanding of the facts or, perhaps, embraced a broader, non-economic rationale for ending access requirements for the delivery of next generation services?

3. The structure of government — what's best decided by whom?

The regulation of traditional voice telephony — and the open access requirements applied to it in the United States and Europe — lies largely beyond the scope of this paper. It is worth noting briefly, however, the importance of the question of governmental structure in the world of "open" access as illustrated by the recent FCC decision. Without delving deeply into the details, the basic question before the FCC regarding residential voice service was whether competitive operators would continue to have access to the complete network of the incumbents in order to provide their own voice services. As one would expect, the battle over this issue invoked a series of economic contentions of the type discussed earlier in this paper. But the FCC's ultimate decision did not firmly decide the question — instead the bulk of the analysis and the ultimate outcomes were left to state regulators to complete. And this set off a veritable firestorm of commentary. Chairman Powell criticized it extensively, arguing the majority had "abdicate[d] its responsibility," that a series of state-by-state proceedings would "prove too chaotic for an already fragile telecom market" [80] and that the Commission has given "the states a subjective and unrestricted role in determining the fate" of this form of competitive access [81]. Commissioner Abernathy added that this approach "[i]n addition to jettisoning the principle of regulatory certainty ... tramples on the goal of promoting facilities-based competition" [82]. But Commissioner Martin argued in response that the state-by-state approach was the more careful one — "putting in place a granular analysis that recognizes that competitors face different operational and economic barriers in different markets" [83], and Commissioner Adelstein embraced state regulators "as our full partners in [the Telecom Act's] implementation" [84].

Much more could be said about this disagreement — and about its implications for competition in the market for voice telephony. The briefest conclusion is this: the application of the policy bases articulated in this paper would be incomplete without an understanding of the separate logic — and the politics — that surround the decisions appropriately taken on the level of a governmental body.

A final word on content

On 21 February 2003 the National Cable & Telecommunications Association responded to the contention of some in the high-tech community, as discussed above, that the ability of end-users to unfettered Internet content must be protected, including through the use of governmental regulation [85]. Without analyzing its legal contentions about the scope of FCC authority, the letter is notable for its statements that:

  • "cable modem customers have — and have always had — access to all lawful content on the Internet"
  • "regulation purporting simply to guarantee unfettered access to the Internet could be used to restrict all sorts of commercial arrangements ...", and
  • "[t]hese efforts would deter investment and innovation in the provision of high-speed Internet services."

One suspects that this correspondence comes near the beginning of a debate that will continue for a long time over end-user access to content, applications and devices. Indeed, it is possible that the continuation of economic de-regulation — as characterized by the FCC's February decision — may increase, not decrease, the pressure on regulators to consider the power of de-regulated and un-regulated networks to "close" access to end users (and their upstream suppliers). That is because the presence of a small number of de-regulated next-generation networks may re-invigorate the old-time discussion of "common carrier"-type obligations that, as the earlier discussion of the "classic" Bell network illustrated, imposed the requirement of open content without any requirements of open network access to competitors. Or that kind of concentrated market structure could lead to new forms of regulatory action being proposed. Depending on the structure of future market economics — and even noting the unequivocal endorsement of open content by the cable operators — it is possible that the focus on open access in the future may, in order words, increasingly become focused on content-related claims and rebuttals. End of article


About the Author

Jonathan Sallet is Co-Director of the National Security Agency (NSA) Project in the Center for International and Security Studies at the University of Maryland (CISSM), School of Public Affairs and Visiting Scholar, Berkeley Roundtable on International Economics. Previously, he served as Chief Policy Counsel of MCI WorldCom, in the years immediately after the passage of the Telecommunications Act of 1996, and as Director of the Office of Planning & Strategic Policy at the U.S. Department of Commerce (1993-1996) where he worked in the Clinton Administration's creation of telecommunications and technology policy. He served as a law clerk to Associate Justice Lewis F. Powell, Jr. of the United States Supreme Court and is a graduate of the University of Virginia School of Law, where he served as Editor-In-Chief of the Virginia Law Review, and Brown University.
E-mail: jonathan@sallet.com



Mr. Sallet gratefully acknowledges the research assistance of David Bach and Hasan Zafar Khan. Mr. Sallet also appreciates the counsel of Mary Brown, Michael Pelcovits, Robert Latham, Viktor Mayer-Schoenberger, Steven Weber and John Zysman, none of who bear any responsibility for errors in the paper. An initial version of this paper was presented as part of the National Security Agency (NSA) Project in the Center for International and Security Studies at Maryland (CISSM), School of Public Affairs, the support of which Mr. Sallet also gratefully acknowledges.



1. The proceedings are (i) the Triennial Review, (ii) the "Wireline Broadband Item" addressing, in part, "the appropriate statutory classification of broadband Internet access services provided over the traditional or future wireline telephone network", (iii) the "Second Cable Modem Service Order", addressing, inter alia, "whether [the FCC] should require cable operators to offer ISPs access to their facilities", and (iv) the "Dom/Non-Dom Proceeding" which asks whether the Commission should "make changes, based on marketplace developments, in its traditional regulatory requirements of ILECs' broadband transmission services." See Powell, 2003.

2. European Communities, Directive 2002/19/EC of the European Parliament and of the Council of 7 March 2002 on access to and interconnection of, electronic communications networks and associated facilities (Access Directive), at http://europa.eu.int/information_society/topics/telecoms/regulatory/new_rf/documents/l_10820020424en00070020.pdf, accessed 6 February 2003.

3. Hundt, 2000, p. 226.

4. By "next-generation networks" this paper means to include facilities open to public subscription that, in the eyes of policy makers, mark a technological advance over the data abilities of fixed, wire-line telephone networks either in their ability to carry additional amounts of data ("bandwidth") or their ability to operate more flexibly for the end-user's convenience (such as wireless data networks). Under this definition, "next-generation networks" would therefore include, but not be limited to, DSL deployment over traditional wire networks, the substitution of fiber for copper in traditional telephone networks, the use of cable networks through cable modems and the use of any form of wireless network for Internet communication, including UltraWideBand, WiFi, 2.5 G and 3G networks.

5. Hundt, 2000, pp. 8-11.

6. As an example of a broader agenda, in 1993, the Clinton Administration endorsed nine principles for action that included (i) promoting private investment, (ii) extending the concept of universal service, (iii) government action as a catalyst to promote innovation and applications, (iv) promoting "seamless, interactive, user-driven operation" of a National Information Infrastructure, (v) ensuring information security and network reliability, (vi) improved spectrum management, (vii) protection of intellectual property rights, (viii) international harmonization and (ix) improved access to government information and better government procurement. See Association of Research Libraries, 1995.

7. See World Trade Organization, n.d., at http://www.wto.org/english/tratop_e/serv_e/telecom_e/telecom_results_e.htm, accessed 6 February 2003.

8. See the Telecommunications Act of 1996, at http://www.fcc.gov/telecom.html, accessed 6 February 2003.

9. See European Communities, 2002b, at http://europa.eu.int/information_society/topics/telecoms/regulatory/new_rf/documents/l_10820020424en00070020.pdf, accessed 6 February 2003.

10. See Umino, 2002, discussing, among others, Sweden, Korea and Japan.

11. See U.S. Department of Commerce — National Telecommunications and Information Administration, 2000, at http://www.ntia.doc.gov/reports/ruralbb42600.pdf, accessed 6 February 2003.

12. See U.S. Department of Commerce — National Telecommunications and Information Administration, 2001, at http://www.ntia.doc.gov/ntiahome/threeg/33001/3g33001.pdf, accessed 6 February 2003.

13. See Federal Communications Commission, 2002, http://epicentertech.net/vrmad3d/pdf_docs/FCC_Broadband.pdf, accessed 6 February 2003.

14. See European Commission, 2000a, at http://europa.eu.int/comm/competition/mergers/cases/decisions/m1741_en.pdf, accessed 6 February 2003.

15. See Kennard, 2001.

16. See the Telecommunications Act of 1996. Some would distinguish "unbundling" of incumbent networks from "openness" requirements on the ground that "unbundling" stems from monopoly characteristics of the incumbent network while other openness requirements, although based on economic analysis, do not. For the purposes of this paper, "unbundling" requirements are treated as a subset of "openness", a characterization that does not require any specific justifications of "unbundling" to be ignored.

17. George Ford (2002), as quoted in "Z-Tel Releases UNE-P White Paper," The Digest (25 July), at http://www.thedigest.com/more/142/142-042.html, accessed 6 February 2003.

18. See USTA v. FCC, United States Court of Appeals for the District of Columbia, at http://pacer.cadc.uscourts.gov/common/opinions/200205/00-1012a.txt, accessed 6 February 2003.

19. Vietor, 1994, pp. 190-191.

20. Ibid., pp. 194-196.

21. Saltzer et al., 1984, p. 284.

22. Op. cit.

23. See Yochai Benkler's e2e map at http://cyberlaw.stanford.edu/e2e/e2e_map.html, accessed 6 February 2003..

24. See Hush-A-Phone Corporation v. United States, 238 F.2d 266 (D.C. Cir. 1956).

25. See Re Use of the Carter phone Device in Message Toll telephone Service, 13 FCC2d 420 (1968).

26. See Section 251(c)(3)&(5), Telecommunications Act of 1996.

27. See Federal Communication Commission, 2000, at http://www.fcc.gov/Bureaus/Common_Carrier/Orders/2000/fcc00297.txt, accessed 6 February 2003.

28. In this vein, Mr. Seidenberg said that: "Riding on someone else's network and investment is not a growth business. It will not help us — or the interexchange companies and CLECs, for that matter — compete against the cable, satellite, wireless, software and content companies for the information-age customer." See Seidenberg, 2001.

29. The now classic formulation of these economic principles and their application to the information age, described with much greater sophistication, can be found in Shapiro and Varian, 1999.

30. Shapiro and Varian, 1999, p. 177.

31. Ibid., pp. 196-203.

32. Federal Communications Commission, 2001, pp. 68-69.

33. Note the Ivan Seidenberg comment cited earlier: "Riding on someone else's network and investment is not a growth business. It will not help us — or the interexchange companies and CLECs, for that matter — compete against the cable, satellite, wireless, software and content companies for the information-age customer."

34. Of course, other policy disciplines, including the application of the anti-trust laws, might offer a different view on the appropriateness of price discrimination as an appropriate business tactic.

35. Even more starkly, it has been argued that FCC "openness" rules "encouraged many entrants and their investors to believe that the government would support even uneconomic entry in the marketplace, especially the local service segment, merely for the sake of toting up new 'competitors' on a scorecard." See May, n.d.

36. See http://www.eff.org/~barlow/, accessed 6 February 2003.

37. Vietor, 1994, p. 172.

38. Ibid., p. 211.

39. Federal Communications Commission, 2001, pp. 158-160.

40. See Federal Communications Commission, 2001, para 188. The FCC also noted that AOL-Time Warner would be able to seek a waiver of the condition, should its imposition ever come to pass, by showing that market conditions had changed and that market conditions had changed. See Ibid., at para 195. In a significant dissent, then Commissioner (and now Chairman) Michael Powell criticized the Commission's analysis on the ground that the majority's analysis had not demonstrated that (using network effects analysis) the market for instant messaging had "tipped" to the permanent advantage of AOL. He noted, for example, that AOL's competitors were growing at a much faster rate than AOL itself whereas "if a market has tipped one would normally expect to see exponential growth by the leader and a precipitous fall off by its competitors." See Powell, 2001, p. 6.

41. Vietor, 1994, p. 184.

42. This is not to say that all vertical combinations are anti-competitive or even successful. AT&T's acquisition of cable systems in the 1990s and its Excite/@Home strategy for linking Internet access to local network ownership was not successful. Excite/@Home failed and AT&T divested its cable holdings in 2002. See Schaff, 2001.

43. Krattenmaker, 1998, p. 535.

44. It is worth repeating that these issues did not, and do not, exist in the world of dial-up Internet access because the local telephone network, for these purposes, is an "open" platform. Just as any customer may call any other customer, so may any ISP provide its customers with the unfettered ability to reach it without regard to the market power of the incumbent telephone provider. The initial unconcentrated structure of the market for Internet Service Providers was no doubt influenced by the existence of this "open access" requirement.

45. Federal Communications Commission, 2001, para 47, 126.

46. See the FCC's online dossier on this case at http://www.fcc.gov/transaction/attcomcast.html, accessed 6 February 2003.

47. For an analysis of "openness" along these lines, see Bar, 2002.

48. See the discussion on cable modem limitations at U.S. Robotics, n.d., at http://www.usr.com/education/bb1.asp, accessed 6 February 2003. .

49. A telephone network does have that right vis-à-vis a competitive telephone network. See Section 251, Telecommunications Act of 1996. An Internet backbone network does not have the right to exchange traffic with another backbone — which formed part of the market structure at issue in the WorldCom/Sprint merger.

50. See 77 FCC2d, p. 474, showing that facilities-based providers of information services must sell transmission to unaffiliated ISPs on the same terms and conditions that they sell that transmission to themselves or affiliated companies.

51. Telecommunications Act of 1996 — Section 251(b)(1): "Each local exchange carrier has the following duties: (1) RESALE — The duty not to prohibit, and not to impose unreasonable or discriminatory conditions or limitations on, the resale of its telecommunications services."

52. Telecommunications Act of 1996, Section 251(c)(3)(Incumbent telephone companies only): "UNBUNDLED ACCESS — The duty to provide, to any requesting telecommunications carrier for the provision of a telecommunications service, nondiscriminatory access to network elements on an unbundled basis at any technically feasible point on rates, terms, and conditions that are just, reasonable, and nondiscriminatory in accordance with the terms and conditions of the agreement and the requirements of this section and section 252. An incumbent local exchange carrier shall provide such unbundled network elements in a manner that allows requesting carriers to combine such elements in order to provide such telecommunications service."

53. See the discussion of cellular-telephone spectrum allocation in Roessner et al., 1998, at http://www.sri.com/policy/stp/techin2/chp4.html, accessed 6 February 2003; the discussion of digital television spectrum in Federal Communications Commission, 1997, at http://www.fcc.gov/Bureaus/Mass_Media/Orders/1997/fcc97116.pdf, accessed 6 February 2003; and Section 652, Telecommunications Act of 1996, on cable-telephone cross-ownership.

54. See European Communities, 2002b, http://europa.eu.int/information_society/topics/telecoms/regulatory/new_rf/documents/l_10820020424en00070020.pdf, accessed 6 February 2003.

55. SBC is reportedly urging the FCC to distinguish between business and residential customers for purposes of permitting competitor access to unbundled elements. See Levin et al., 2002.

56. According to Erkki Liikanen, Europe's Commissioner for the Information Society, "In Europe, the broadband market is currently dominated by ADSL." See Liikanen, 2002a.

57. In the U.S., in mid-2002, cable modems served 9.2 million lines; DSL service served 5.1 million lines. See Federal Communication Commission, 2002, http://epicentertech.net/vrmad3d/pdf_docs/FCC_Broadband.pdf, accessed 6 February 2003.

58. See Verizon Communications v. FCC, at http://a257.g.akamaitech.net/7/257/2422/13may20021500/www.supremecourtus.gov/opinions/01pdf/00-511.pdf, accessed 6 February 2003.

59. Of course, legal principles of statutory construction themselves may permit a reviewing court to delve into the "intentions" of the framers of the legislation. But this is an analysis that does not, strictly speaking, attempt to find an independent disciplinary basis for decision.

60. In real life, the network provider would likely also assert engineering concerns and the opposing right to be free to edit content that travels over its network — as a cable provider is free to decide which television channels its networks will carry.

61. Separate from this question, which is posed in a hypothetical manner, is the question of how a regulator can persuade a reviewing court that it has applied only those considerations of the governing law that have been authorized for its consideration.

62. Such research inquiries should consider the extent to which epistemic communities provided the basis for the consideration of the Internet as an "idea" and the extent to which their influence fluctuated during the second-generation period of policy making. As Professor Ernst Haas has explained, epistemic communities "are carriers of scientific knowledge into politics ... associations of professional experts in a particular field who, because of the knowledge they have, have an unusual influence on politicians and bureaucrats, and are, therefore, able to penetrate government departments and make their ideas part of policy" (http://globetrotter.berkeley.edu/people/Haas/haas-con3.html, accessed 6 February 2003). See also Haas, 1992 and, Haas, 1990.

63. European Communities, 2002a, para 10.

64. Ibid., para 17.

65. Ibid., chapter II, article 5-1-b.

66. Ibid., para 13.

67. The five FCC Commissioners (Michael Powell, Kathleen Abernathy, Kevin Martin, Michael Kopps and Jonathan Adelstein) all delivered statements on 14 January 2003 to the Committee on Commerce, Science and Transportation of the United States Senate that discussed the importance of the upcoming proceedings and laid out their approach to the openness issues. These statements can be found on the FCC's Web site, http://www.fcc.gov.

68. Separate Statement of Commission Jonathan S. Adelstein Approving in Part, Concurring in Part, Dissenting in Part. All of the Commission statements noted in this discussion are available at the FCC Web site at http://www.fcc.gov. References to those statements delivered by commissioners to the Subcommittee on telecommunications and the Internet, Committee on Energy and Commerce, U.S. House of Representatives, 26 February 2003 are labeled as "Congressional Statement".

69. See Attachment to Triennial Review Press Release.

70. Press Statement of Commissioner Kathleen Q. Abernathy.

71. Commissioner Kevin J. Martin's Press Statement on the Triennial Review, emphasis in original.

72. Separate Statement of Chairman Michael K. Powell Dissenting in Part.

73. Abernathy Statement.

74. Press Statement of Commissioner Michael J. Copps, Approving in Part, Concurring in Part, Dissenting in Part.

75. Powell Statement.

76. Abernathy Statement.

77. Powell Congressional Statement, emphasis in original.

78. Powell Statement.

79. Op. cit.

80. Powell Statement.

81. Powell Congressional Statement.

82. Abernathy Statement.

83. Martin Statement.

84. Adelstein Statement.

85. The text of the letter is available at http://www.ncta.com.



Sam Ames, 2001. "Wireless merger gets OK despite debate," CNET News.com (25 April), at http://news.com.com/2100-1033-256505.html, accessed 6 February 2003.

François Bar, 2002. "Open networks: Theoretical and policy lessons from the Internet," Annenberg Colloquia, University of Southern California (2 April), at http://www.stanford.edu/~fbar/Presentations/index.html, accessed 6 February 2003.

François Bar, Stephen Cohen, Peter Cowhey, Brad DeLong, Michael Kleeman, and John Zysman, 2001. "The next-generation Internet: Promoting innovation and user-experimentation," In: Tracking a transformation: E-commerce and the terms of competition in industries. Washington, D.C.: Brookings Institution Press, pp. 435-473; and at http://www.stanford.edu/~fbar/Publications/Brookings-NextGenInternet.pdf,accessed 8 March 2003.

John Perry Barlow, 1996. "A declaration of the independence of cyberspace," (8 February), at http://www.eff.org/~barlow/Declaration-Final.html, accessed 6 February 2003.

Tim Berners-Lee, 2000. Weaving the Web. New York: HarperCollins.

Cable Datacom News, 2002. "Time Warner Cable touts multi-ISP upside," Cable Datacom News (1 May), at http://www.cabledatacomnews.com/may02/may02-3.html, accessed 6 February 2003.

Center for Digital Democracy, n.d. "Open access," at http://www.democraticmedia.org/issues/openaccess/, accessed 6 February 2003.

George A. Chidi, Jr., 2001. "WorldCom buys Rhythms' gear to keep service running," ITworld.com (26 September), at http://www.itworld.com/Net/2574/IDG010926worldcom/, accessed 6 February 2003.

Coalition of Broadband Users and Innovators, 2002. "Letter to Chairman Powell et al.," (18 November), at http://www.itaa.org/isec/docs/itaa111802cbui.pdf, accessed 6 February 2003.

Fred Dawson, 2001. "Emergency demands: Industry likely to address network reliability during times of crisis," XCHANGE (1 November), at http://www.xchangemag.com/articles/1B1front.html, accessed 6 February 2003.

Esther Dyson, George Gilder, George Keyworth, and Alvin Toffler, 1994. "Cyberspace and the American dream: A Magna Carta for the knowledge age," Progress & Freedom Foundation (22 August), at http://www.pff.org/position.html, accessed 6 February 2003.

European Commission, 2002. "Commission launches public consultation on open platforms in digital TV and 3G," press release (9 December), at http://europa.eu.int/rapid/start/cgi/guesten.ksh?p_action.gettxt=gt&doc=IP/02/1818|0|RAPID&lg=EN, accessed 6 February 2003.

European Commission, 2000a. "Case No. COMP/M.1741-MCI WorldCom/Sprint," (28 June), at http://europa.eu.int/comm/competition/mergers/cases/decisions/m1741_en.pdf, accessed 6 February 2003.

European Commission, 2000b. "Commission prohibits merger between MCI WorldCom and Sprint," press release (28 June), at http://www.europa.eu.int/rapid/start/cgi/guesten.ksh?p_action.gettxt=gt&doc=IP/00/668|0|RAPID&lg=EN, accessed 6 February 2003.

European Commission, n.d. "Commission staff working document on barriers to widespread access to new services and applications of the information society through open platforms in digital television and third generation mobile communications," at http://europa.eu.int/information_society/topics/telecoms/regulatory/publiconsult/documents/211_29_en.pdf, accessed 6 February 2003.

European Communities, 2002a. "Directive 2002/19/EC of the European Parliament and of the Council of 7 March 2002 on access to and interconnection of, electronic communications networks and associated facilities (Access Directive)," Official Journal of the European Communities L 108/7 (24 April), at http://europa.eu.int/information_society/topics/telecoms/regulatory/new_rf/documents/l_10820020424en00070020.pdf, accessed 6 February 2003.

European Communities, 2002b. "Regulation (EC) No 2887/2000 of the European Parliament and the European Council of 18 December 2000 on unbundled access to the local loop," Official Journal of the European Communities L 336/4 (30 December), at http://europa.eu.int/information_society/topics/telecoms/regulatory/new_rf/documents/oj-regull-en.pdf, accessed 6 February 2003.

Federal Communications Commission, 2002. "FCC Declines to approve Echo-Star-DirecTV merger," (10 October), at http://www.fcc.gov/Daily_Releases/Daily_Business/2002/db1010/DOC-227263A1.doc, accessed 6 February 2003.

Federal Communications Commission, 2001. "Memorandum, Opinion and Order: AOL-Time Warner," (22 January), at http://www.fcc.gov/Bureaus/Cable/Orders/2001/fcc01012.txt, accessed 6 February 2003.

Federal Communications Commission, 1997. "In the matter of advanced television systems," (21 April), at http://www.fcc.gov/Bureaus/Mass_Media/Orders/1997/fcc97116.pdf, accessed 6 February 2003.

George Gilder, 2001. "Tumbling into the telechasm," Wall Street Journal (6 August), and at http://gilder.com/Gilder.comNews/GGWSJ08.06.01.htm, accessed 8 March 2003.

George Gilder, 2000. "Open access now! Wait, never mind," Wall Street Journal (18 February), p. A14.

Al Gore, 1994. "Speech delivered at the Information Superhighway Summit at UCLA," (11 January), at http://www.uibk.ac.at/sci-org/voeb/texte/vor9401.html, accessed 6 February 2003.

Corey Grice, 1999. "Making waves in the data world," CNET News.com (22 February), at http://news.com.com/2009-1033-221422.html?legacy=cnet, accessed 6 February 2003.

Ernst B. Haas, 1990. When knowledge is power: Three models of change in international organizations. Berkeley: University of California Press.

Peter M. Haas, 1992. "Introduction: Epistemic Communities and International Policy Coordination," International Organization, volume 46, number 1 pp. 1-35. http://dx.doi.org/10.1017/S0020818300001442

Katie Hafner and Matthew Lyon, 1996. Where wizards stay up late: The origins of the Internet. New York: Simon & Schuster.

Jerry A. Hausman, Hal J. Singer, and J. Gregory Sidak, 2001. "Cable modems and DSL: Broadband Internet access for residential customers," American Economic Review, volume 91, number 2 (May), pp. 302-307. http://dx.doi.org/10.1257/aer.91.2.302

Thomas W. Hazlett and George Bittlingmayer, 2001. "The political economy of cable 'open access'," at http://aei.brookings.org/admin/pdffiles/working_01_06.pdf, accessed 6 February 2003.

High Tech Broadband Coalition, 2002. "Comments in the matter of appropriate regulatory treatment for broadband access to the Internet over cable facilities (CC Docket No. 96-45), " FCC Filing (17 June), at http://www.unibex.com/ibx/users/GetCom1/CIAJ1/website/index.cfm, accessed 6 February 2003.

Reed E. Hundt, 2000. You say you want a revolution: A story of information age politics. New Haven: Yale University Press.

Mitch Kapor, 1993. "Where is the digital highway really heading? The case for a Jeffersonian information policy," Wired, volume 1, number 3 (July-August), pp. 53-59, 94, and at http://www.wired.com/wired/archive/1.03/kapor.on.nii.html?person=mitch_kapor&topic_set=wiredpeople, accessed 6 February 2003.

Michael L. Katz and Carl Shapiro, 1994. "Systems competition and network effects," Journal of Economic Perspectives, volume 8, number 2 pp. 93-115. http://dx.doi.org/10.1257/jep.8.2.93

William E. Kennard, 2001. "Statement of FCC chairman on conditioned approval of the AOL Time Warner merger," Federal Communications Commission (11 January), at http://www.fcc.gov/Speeches/Kennard/Statements/2001/stwek101.html, accessed 6 February 2003.

William E. Kennard, 2000. "Statement of William E. Kennard, chairman of the Federal Communications Commission, on foreign government ownership of American telecommunications companies," U.S. House of Representatives, Committee on Commerce, Subcommittee on Telecommunications, Trade & Consumer Protection (7 September), at http://com-notes.house.gov/cchear/hearings106.nsf/a317d879d32c08c2852567d300539946/77fe9d23603418b68525695100537dae?OpenDocument, accessed 6 February 2003.

Thomas G. Krattenmaker, 1998. Telecommunications law and policy. Durham: Carolina Academic Press.

Jonathan Krim, 2003. "Copper lines regaining luster: With the obstacles to fiber, phone companies are tapping the old infrastructure," Washington Post (7 February), p. E01, and at http://www.washingtonpost.com/ac2/wp-dyn?pagename=article&node=&contentId=A38106-2003Feb6¬Found=true, accessed 8 March 2003.

Mark Lemley and Lawrence Lessig, 2001. "The end of end-to-end: Preserving the architecture of the Internet in the broadband era," Berkeley Law & Economics Working Papers, volume 2000, number 2, article 8, at http://www.bepress.com/blewp/default/vol2000/iss2/art8, accessed 8 March 2003.

Blair Levin, et al., 2002. "SBC offers UNE plan as FCC and states eye jurisdiction," Washington Telecom & Media Insider (Legg Mason Wood Walker, Inc.), (22 November), at http://www.leggmason.com.

Erkki Liikanen, 2002a. "Ambitions for Europe's telecommunications sector," European Competitive Telecommunications Association (ECTA) (11 December), at http://europa.eu.int/rapid/start/cgi/guesten.ksh?p_action.gettxt=gt&doc=SPEECH/02/627|0|RAPID&lg=EN&display=, accessed 6 February 2003.

Erkki Liikanen, 2002b. "The EU communications regulatory framework," Forum for EU-US Legal-Economic Affairs (3 April), at http://europa.eu.int/rapid/start/cgi/guesten.ksh?p_action.gettxt=gt&doc=SPEECH/02/140|0|RAPID&lg=EN&display=, accessed 6 February 2003.

Randolph J. May, n.d. "They just want to be free: Telecom can end its crisis if only the FCC will let firms truly compete," Progress & Freedom Foundation, at http://www.pff.org/RandysPOVsinLegalTimes/MaysPOV111402.htm, accessed 6 February 2003.

Walter B. McCormick, Jr., 2001. "Telecom president calls for additional focus on network security," press release (8 October), at http://media.usta.org/pr/2001/pr100801.html, accessed 6 February 2003.

Larry R. Parkinson, 2000. "Statement of Larry R. Parkinson, general counsel for the Federal Bureau of Investigation, on foreign government ownership of American telecommunications companies," U.S. House of Representatives, Committee on Commerce, Subcommittee on Telecommunications, Trade & Consumer Protection (7 September), at http://www.fbi.gov/congress/congress00/parkinson.htm, accessed 6 February 2003.

Michael K. Powell, 2003. "Statement of Michael K. Powell, chairman of the Federal Communications Commission, on competition issues in the telecommunications industry," United States Senate, Committee on Commerce, Science and Transportation (14 January), at http://hraunfoss.fcc.gov/edocs_public/attachmatch/DOC-230241A1.pdf, accessed 6 February 2003.

Michael K. Powell, 2002. "Remarks at Goldman Sachs Communicopia XI Conference," (2 October), at http://hraunfoss.fcc.gov/edocs_public/attachmatch/DOC-226929A1.pdf, accessed 6 February 2003.

Michael K. Powell, 2001. "Statement of Commissioner Michael K. Powell, concurring in part and dissenting in part," Federal Communications Commission (22 January), at http://www.fcc.gov/Speeches/Powell/Statements/2001/stmkp104.html, accessed 6 February 2003.

David Roessner, Robert Carr, Irwin Feller, Michael McGeary, and Nils Newman, 1998. The Role of NSF's Support of Engineering in Enabling Technological Innovation. Final Report — Phase II. Arlington, Va.: SRI International, and at http://www.sri.com/policy/stp/techin2/, accessed 8 March 2003.

J.H. Saltzer, D.P. Reed, and D.D. Clark, 1984. "End-to-end arguments in systems design," ACM Transactions on Computer Systems, volume 2, number 4 (November), pp. 277-288. http://dx.doi.org/10.1145/357401.357402

SBC, 2002a. "SBC provides updates on long-distance entry and UNE-P line loss: Company comfortable with full-year earnings consensus of $2.26," press release (8 October), at http://www.sbc.com/press_room/1,5932,31,00.html?query=20263, accessed 6 February 2003.

SBC, 2002b. "SBC to cut 11,000 jobs and investment due to outmoded regulatory scheme and weak economy," press release (26 September), at http://www.sbc.com/press_room/news_search/1,5932,31,00.html?query=20257, accessed 6 February 2003.

William Schaff, 2001. "Taking stock: Ma Bell's apron tears as AT&T divests its broadband division," InformationWeek (24 December), at http://www.informationweek.com/story/IWK20011221S0019, accessed 6 February 2003.

Ivan Seidenberg, 2001. "Remarks at National Press Club, " (3 December), at http://www.npr.org/programs/npc/2001/011203.iseidenberg.html, accessed 6 February 2003.

Carl Shapiro and Hal R. Varian, 1999. Information rules: A strategic guide to the network economy. Boston: Harvard Business School Press.

Cass Sunstein, 2001. "The future of free speech," Little Magazine (March/April), at http://www.littlemag.com/mar-apr01/cass7.html, accessed 6 February 2003.

Time, 1999. "Introduction: Why the founder of Amazon.com is our choice for 1999," Time (27 December), at http://www.time.com/time/poy/intro.html, accessed 6 February 2003.

U.S. Department of Commerce, National Telecommunications and Information Administration, 2001. "The potential for accommodating third generation mobile systems in the 1710 to 1850 MhZ mand," at http://www.ntia.doc.gov/ntiahome/threeg/33001/3g33001.pdf, accessed 6 February 2003.

U.S. Department of Commerce, National Telecommunications and Information Administration, 2000. "The challenge of bringing broadband service to all Americans," at http://www.ntia.doc.gov/reports/ruralbb42600.pdf, accessed 6 February 2003.

U.S. Robotics, n.d. "What is cable Internet access?," at http://www.usr.com/education/bb1.asp, accessed 6 February 2003.

Atsushi Umino, 2002. "Broadband infrastructure deployment: The role of government assistance," Organization for Economic Cooperation and Development (OECD) (22 May), at http://www.olis.oecd.org/olis/2002doc.nsf/43bb6130e5e86e5fc12569fa005d004c/42158ef983225772c1256bc100560c01/$FILE/JT00126526.PDF, accessed 6 February 2003.

Richard H.K. Vietor, 1994. Contrived competition: Regulation and deregulation in America. Cambridge, Mass.: Belknap Press of Harvard University Press.

Alvin Von Auw, 1983. Heritage & destiny: Reflections on the Bell System in transition. New York: Praeger.

Jim Wagner, 2001. "Covad to file for bankruptcy," internetnews.com (7 August), at http://www.internetnews.com/xSP/article.php/861601, accessed 6 February 2003.

World Trade Organization, n.d. "Results of the basic telecommunication negotiations," at http://www.wto.org/english/tratop_e/serv_e/telecom_e/telecom_results_e.htm, accessed 6 February 2003.

Z-Tel Technologies, 2002. "Bell companies' claims of financial troubles "mere theatrics," Z-Tel tells FCC and state utility commissioners today," press release (23 September), at http://www.corporate-ir.net/ireye/ir_site.zhtml?ticker=ZTEL&script=410&layout=-6&item_id=336474, accessed 6 February 2003.


Editorial history

Paper received 11 February 2003; revised version received 5 March 2003; accepted 7 March 2003.

Contents Index

Copyright ©2003, First Monday

Copyright ©2003, Jonathan Sallet

Just how open must an open network be for an open network to be labeled "open"? by Jonathan Sallet
First Monday, volume 8, number 3 (March 2003),
URL: http://firstmonday.org/issues/issue8_3/sallet/index.html

A Great Cities Initiative of the University of Illinois at Chicago University Library.

© First Monday, 1995-2019. ISSN 1396-0466.