Avoiding "dog in the manger" regulation - a nuanced approach to net neutrality in Canada.
L'article soutient que les formes de neutralite des reseaux proposees exigent dans l'ensemble des garanties reglementaires dans un marche deja fort concurrentiel, mais que rien ne demontre que les regles de prescription enoncees s'averent vraiment plus avantageuses que les forces actuelles du marche, tant sur le plan de l'acces que du contenu et des applications Internet. L'argument est fait que la Loi sur les telecommunications canadienne, en particulier les articles 27 et 36 et leur application cas par cas privilegiee par le Conseil de la radiodiffusion et des telecommunications canadiennes (CRTC) offre un plan efficace de gestion des problemes de neutralite des reseaux et qu'il n'y a pas lieu de prevoir de nouvelles regles ex ante. Le Canada s'etant engage a intervenir le moins possible dans le secteur des telecommunications, en vertu des articles 27 et 36, il est allegue que la norme applicable pour evaluer le comportement des fournisseurs d'acces Internet (FAI) est l'entrave importante a la concurrence creee par ce comportement sur le marche des contenus et des applications. Il est suggere que la plupart des formes de neutralite des reseanx proposees ne sauraient repondre a ce critere de la concurrence. Afin de constituer un principe de politique publique valable dans un milieu deja concurrentiel, la << neutralite des reseaux >> devrait etre definie de facon plus etroite, en dehors du grand plan reglementaire des garanties, de facon a completer et non a jouer au detriment des forces du marche. Une telle complementarite serait possible la ou les consommatrices et les consommateurs se heurtent a un manque d'information ou a une asymetrie datas les pratiques de gestion des reseaux des FAI. Voir la neutralite des reseaux comme un moyen d'assurer une gestion transparente des reseaux ajoute a l'orientation politique ainsi qu'aux dispositions des articles 27 et 36 de la Loi sur les telecommunications et sert l'interet de la concurrence, car plus la pratique de gestion des reseaux est claire pour les consommatrices et les consommateurs qui sont en bout de ligne, moins grands sont les risques d'effets negatifs aux echelons superieurs du marche du contenu et des applications Internet.
Table of Contents I. INTRODUCTION II. NET NEUTRALITY AND CANADA'S TELECOMMUNICATIONS REGULATION POLICY A. The Government's Policy for Economic Regulation in the Telecommunications Sector B. Normative Goals of Net Neutrality Regulation C. Evaluating Net Neutrality Proposals Under the Policy Direction and Canada's Telecommunications Regulatory Framework 1. The Purpose of Net Neutrality Regulations (a) Blocking, Degrading, Interfering with Content and Applications (b) Discriminatory Network Management Practices (c) Access Tiering or Price Discrimination Among Content Applications Providers (d) Discriminatory Pricing of Consumer Access Service 2. Defining the Relevant Market for Competitive Outcome Analysis (a) The High Speed Internet Access Market (b) The Internet Content and Applications Market 3. The Benefits and Costs of Regulatory Guarantee: Innovation and Investments Within and at the Edges of the Network III. TOWARDS A LIGHTER, FOCUSED, LESS INTRUSIVE NET NEUTRALITY POLICY A. Regulatory Guarantee vs. Other Measures to Reduce Downstream Competitive Risks B. Redefining Net Neutrality as Transparency in Network Management Practices 1. Free Press v. Comcast--The FCC's Relapse into Regulatory Guarantee 2. CAIP v. Bell Canada--The CRTC's Competitive Outcome Approach 3. Anti-competitive Incentive Under the Competitive Outcome Approach 4. The Competitive Outcome Approach and Transparency of Network Management Practices IV. CONCLUSION
The discussion in this paper proceeds in two parts. In Part I, the paper reviews various forms of net neutrality regulation proposals against Canada's telecommunications policy and regulatory framework. (1) Like the US, there is currently no Canadian legislation or regulation that codifies a detailed ex ante "net neutrality" set of rules. (2) However, the objectives specified in Section 7 of the Telecommunications Act are broad enough to be interpreted as mandating a policy that advances competition, innovation and investments in the Internet. (3) The Canadian Radio-television and Telecommunications Commission (CRTC) is vested with the authority to regulate telecommunications service providers by the Telecommunications Act. The CRTC has determined that certain specific sections of the Telecommunications Act, in particular sections 27 and 36, can be used to address issues of net neutrality. (4) Policy direction was further clarified in 2007 when the Canadian Government issued binding guidelines to the CRTC on matters of economic regulation in the telecommunications sector. The guidelines, which were issued as a Policy Direction, are rooted in competition law principles and direct the CRTC, in assessing existing and new regulatory measures, to rely on market forces to the maximum extent possible, weigh the costs and benefits of regulation and opt for less intrusive measures than behaviour and price regulation. (5)
Most popular forms of net neutrality proposals have their intellectual pedigree in the works of Professors Lawrence Lessig and Timothy Wu, which are discussed in this paper. (6) This paper suggests that the fundamental argument in the works of Lessig, Wu and other similar scholarship is that, notwithstanding the existence of competition in the high speed Internet access market, net neutrality regulation is required as a guarantee that Internet Service Providers (ISPs) will not behave in the access market in manners that can impede competition, innovation and investment in the Internet content and applications market. The paper calls this the regulatory guarantee approach. This paper argues that the Policy Directions' call for commitment to minimal interference and proportionate regulation, the competitions principles applied by the Competition Bureau under the Competition Act, (7) as well as the CRTC jurisprudence on matters arising under sections 27 and 36 of the Telecommunications Act, require that net neutrality proposals be evaluated from a competitive outcome approach. In other words, what is the impact on competition of the conduct or behaviour that a particular form of net neutrality regulation seeks to constrain?
This paper argues that many forms of net neutrality proposals are justified on the basis of the need for regulatory guarantee but are unlikely to pass scrutiny of the competitive outcome test because they are aimed at regulating, ex ante, behaviour that has no significant negative impact on competition. The most prominent are proposals aimed at: (1) prohibiting or limiting ISPs' ability to charge differential pricing among consumers (such as usage caps); (2) prohibiting ISPs from contracting with Internet content and applications providers for prioritized packet delivery (so-called "access-tiering"); and (3) specifying ex ante what kinds of network management practices are permissible as net neutral. Moreover, these forms of net neutrality regulation are intrusive, unnecessary and disproportionate to their objectives. Because they aim to provide a guarantee against anti-competitive behaviour, they tend to generalize in their scope of application thereby impeding legitimate business and network decisions that are made to advance innovation, efficiency and investment in the network. In contrast, the competitive outcome approach requires that ISP behaviour in the access market be assessed ex post on a case-by-case basis with the overriding question being whether the impugned behaviour could substantially lessen or impede competition in the content and applications market. No blanket or detailed rules of prohibition or constriction should be imposed. Instead specific, real and significant threats to competition should be identified and rectified as the facts may arise.
Using this more nuanced approach, this paper argues that the vast majority of concerns behind proposals for net neutrality regulation are sufficiently addressed by competition. However, net neutrality can play a valuable role as a public policy ideal if it is narrowly defined to address known market failure in the downstream high speed Internet access market. One exceptional area of concern is where, in spite of competition in the high speed Internet access market, there is failure or asymmetry of information between service provider and consumer which prevents or limits the ability of the consumer to make fully informed choice to the extent the competitive market would have dictated. Part II of the paper suggests that it is this information failure or asymmetry that is behind the emerging debates over ISPs' network management practices. Part II examines two recent cases in Canada and the US involving network management practices. In August 2008, the US Federal Communications Commission (FCC) issued a decision in Free Press v. Comcast in which it determined that Comcast's practice of targeting and shaping peer-to-peer (P2P) file sharing traffic violates the FCC's net neutrality policy and cannot be defended as reasonable network management practice. (8) In CAIP v. Bell Canada, the CRTC denied an application by third-party ISPs (supported by content and applications providers and consumer groups) alleging that Bell Canada's shaping of P2P file sharing traffic violates, among others rules, general net neutrality principles as well as specifically sections 27 and 36 of the Telecommunications Act. (9) The CRTC found that Bell Canada's traffic shaping practice in relation to its wholesale ADSL service does not violate any rules. Noting that the decision is specific to wholesale service, the CRTC initiated a public hearing to review ISPs' traffic management practices for both wholesale and retail Internet services. (10) The paper argues that the FCC appears to have been swayed by the regulatory guarantee line of thinking and that it would likely have come to a different conclusion if it had applied a competitive outcome test. In contrast, although the CRTC cautioned that its decision was specific to the wholesale relationship between Bell and the ISPs, the paper suggests that the CRTC in fact endorsed and applied the competitive outcome standard of analysis, as it has done in the past and as it is enjoined to do by the Policy Direction.
Part II argues that the danger in applying regulatory guarantee reasoning to issues of network management practice is that it could lead unintentionally to laying down ex ante prescriptive rules of management in an industry that is going through leaps of technological changes at a fast pace. This in turn could result in regulators inadvertently second-guessing an ISP's engineering judgment as the FCC appeared to have done in Comcast. It is suggested that regulators should identify transparency of network management practices as the broad policy goal of net neutrality, and then leave the industry and each ISP to develop appropriate measures to achieve this goal. In this respect, transparency complements the competitive outcome standard of analyzing net neutrality concerns. Transparency is not a necessary condition for a network management practice to be reasonable under section 27 or permissible under section 36 of the Telecommunications Act if the specific practice at issue does not have substantial negative impact on competition. However, a network management practice that is fully transparent in a market that is competitive enhances consumers' ability to make a fully informed choice even if it treats consumers differently from one another. The competitive outcome of such treatment could be evaluated ex post if there is a complaint that it is anti-competitive.
II. NET NEUTRALITY AND CANADA'S TELECOMMUNICATIONS REGULATION POLICY
A. The Government's Policy for Economic Regulation in the Telecommunications Sector
Gregory Sidak observed that while "[a]n economist would attempt to define network neutrality by asking, 'what is such regulation supposed to maximize?,'" most proponents define net neutrality in vague and ambiguous terms, with often conflicting goals and reasoning. (11) While this kind of generality unfortunately obfuscates genuine concerns that many forms of net neutrality proposals aim to address, there is truly a tremendous amount of scholarly work on net neutrality to leave the practical policymaker befuddled. (12) Luckily, Canada has clear policy guidelines against which to evaluate proposals for net neutrality regulations. In 2006, guidelines on issues of economic regulation in the telecommunications sector were issued by the Government of Canada in a Policy Direction to the CRTC. (13) The backdrop to the Policy Direction was a comprehensive report and set of recommendations for a new telecommunications regulatory framework submitted to the government by the Telecommunications Policy Review Panel on March 22, 2006. (14) In addition to its review of current telecommunications regulation, the TPR Panel specifically examined net neutrality regulation and concluded that:
Canada's telecommunications policy and regulatory framework should include provisions that confirm and protect the right of Canadian consumers to access publicly available Internet applications and content of their choice by means of public telecommunications networks that provide access to the Internet. However, because consumer access issues are complex and rapidly evolving, the Panel also believes it is important to distinguish between various kinds of concerns that arise in relation to consumer access, and to address them through the most appropriate regulatory mechanisms. These concerns include:
* first, concerns arising as a result of anti-competitive conduct
* second, concerns arising as a result of business decisions taken in the context of normal commercial business practices
* third, concerns arising from decisions taken for non-commercial reasons. (15)
The Panel concluded that the first type of concern should be addressed through maximum reliance on market forces and competition law principles, with economic regulation imposed only where market forces are insufficient to discipline the players. The Panel believed that, in most cases, network operators and ISPs will have little or no incentive to interfere with customer access. Regarding the third type of concern, the Panel stated:
[N]on-commercial reasons for blocking access could include legitimate legal prohibitions, for example, national security, child pornography or other criminal concerns. Restrictions on access might also arise because of copyright. In such cases, the Panel believes that blocking access would be legitimate because the access provider would merely be implementing the law.... The Panel believes the most difficult regulatory issues related to consumer access are likely to be those arising as a result of normal, ordinary business decisions that effectively limit or deny access to applications or content, even though they do not involve anti-competitive conduct ... or illegitimate forms of censorship. (16)
The obvious example is where a network operator decides to implement network management practices such as those that will be discussed later in this paper, or implement usage-based billing or bandwidth caps to control bandwidth utilization and recover costs of service. The Panel described the key issues in the following terms:
In some cases, there may be sound business reasons for blocking access to applications and content or degrading service. In other cases, these business practices may exploit customers unreasonably. One simple example of the issues involved in distinguishing between acceptable and unacceptable limitations on consumer access arises in relation to applications or content that require large amounts of capacity on the access provider's network, for example, streaming video. If a sufficiently large number of customers access such an application simultaneously, the level of service for everyone may deteriorate to an unacceptable level. Providing additional capacity may not be feasible in the short run. Even in the longer run, extra capacity is costly, and the prices currently paid by access customers may not justify the necessary investments to expand network capacity. In such a case, an access provider might decide for business reasons to block access to the relevant application, or allow access only to those customers prepared to pay a premium for the service. More generally, some customers may be willing to accept reduced access in exchange for lower prices. Conversely, other customers may be willing to pay a premium for a higher grade of service. It is common business practice to offer a range of levels of service at prices that reflect the differences.... Even more difficult questions arise when an access provider enters into an arrangement with an application provider to give preferential access to that provider's applications. For example, an Internet access provider may give priority to instant messages carried by one system and degrade instant messages carried by other systems to the point where they are difficult to use, so as to encourage customers to switch instant messaging systems. Access providers thus leverage their market power in the Internet access market to try to extract more profit, either directly or in partnership with a preferred third party, in the applications market. (17)
With respect to this second category of concerns, the Panel did not endorse proposals for prescriptive and detailed net neutrality legislation or regulation. Instead, it recommended that "[t]he Telecommunications Act should be amended to confirm the right of Canadian consumers to access publicly available Internet applications and content of their choice by means of all public telecommunications networks providing access to the Internet." (18) The Panel recommended this declaration of right "take into account any reasonable technical constraints and efficiency considerations related to providing such access." (19)
In terms of the overall direction for Canadian telecommunications policy, the Panel recommended that the federal government issue a binding Policy Direction to the CRTC to guide the Commission's regulatory initiatives. The Policy Direction issued to the CRTC essentially directed that, in determining whether there is need for economic regulation, the CRTC must apply the disciplines of welfare economics (through the weighing of the relative benefits and costs of regulation) and competition law (through the identification of the conditions under which market forces are not likely to achieve the purposes to which a regulation is directed). (20)
The CRTC has outlined a framework for assessing whether or not a particular regulation or proposed regulation is consistent with the Policy Direction. (21) For purposes of evaluating forms of net neutrality regulations, the framework can be applied as follows:
What is the purpose of the form of net neutrality proposal and can the purpose be achieved through reliance on market forces and without economic regulation?
ii) Does the proposed form of net neutrality regulation minimally interfere with market forces, including market-driven efficient entry into the market?
iii) What are the benefits and costs of relying on this particular form of net neutrality regulation to the exclusion of market forces and do those benefits outweigh the costs?
iv) If market forces are not sufficient in themselves to attain its objectives, are there any alternative regulatory measures other than this particular form of net neutrality regulation that would be less intrusive?
The CRTC's analytical framework has elements similar to those that have been used by law and economics scholars on both sides of the net neutrality debate in the US. These scholarly works offer provocative welfare economics (22) and competition law (23) insights into net neutrality regulation and are referenced or discussed where helpful in the analysis in this paper. But first, we must examine the various objectives aimed to be achieved by forms of regulations that are at the center of the net neutrality debate.
B. Normative Goals of Net Neutrality Regulation
There is no consensus as to what exactly "net neutrality" means, and even though it is a fairly recent topic in intellectual and regulatory discourse, a lot of economic, academic and business scholarship has emerged on the topic. The scholarship is very diverse in the disciplines from which they approach the topic, even if quite bipolar in terms of their conclusion on the question of whether or not a net neutrality regulation is warranted. (24) In order to keep perspective on this question amidst the intellectual noise, this paper will focus only on a few scholars whose works represent the fabric of the thinking on both sides of the debate.
Network neutrality was originally an engineering architectural concept. (25) However, "net neutrality" emerged as a regulatory policy issue in the late 1990s and early 2000s out of concern over the control over access to the emerging Internet by cable and telephone companies. In his 2001 seminal work, The Future of Ideas, Lessig articulated the central concern that cable modem and ADSL (26) high speed Internet access providers would use their control of the "last mile" of the network to block or slow access to content and applications that threaten their proprietary platforms. This, it was feared, would stymie competition and innovation in the emerging market for Internet content and applications. (27) Then, in 2002, Wu conducted a landmark review of possible restrictive practices by both cable modem and ADSL providers. (28) Wu found that most of the surveyed providers practiced several restrictions, including prohibiting customers from operating a server, home networking, overusing of bandwidth and attachment of some equipment. Wu argued that these restrictive practices are commonplace between the two broadband platforms because the two platforms dominated the broadband access market as a duopoly.
As a solution, Lessig, Wu and other commentators began to propound net neutrality as an Internet policy ideal articulated in general as the right of broadband consumers to access and/or download any legal content and application with no discrimination among consumers or among content and applications (with notable exceptions such as virus and illegal content). (29) The basic argument was that restrictions on customer use and any type of actual or potential discrimination against content and applications providers would stymie innovation in this emerging Internet market. Net neutrality regulations were therefore required to preserve competition and innovation. (30)
While net neutrality proposals share this Lessig-Wu fundamental argument, they differ in terms of the actual positive rules that should be imposed on high speed Internet access providers and how those rules should be implemented. In some forms, the proposals declare an absolute right of consumers and users to access and use any legal content and applications of their choice. (31) In 2006, Canada's Telecommunications Policy Review Panel recommended that the Telecommunications Act be amended to declare "the right of Canadian consumers to access publicly available Internet applications and content of their choice by means of all public telecommunications networks providing access to the Internet." (32) Similarly, in the application filed early in 2008 by the Canadian Association of Internet Providers (CAIP) against Bell Canada for "P2P Internet traffic throttling" (discussed later), CAIP asserted that Bell Canada's conduct was in breach of "net neutrality," defined as "the right of individuals to gain unfettered and uninterrupted access to Internet content and applications of their choice." (33)
In other forms, the positive rules would vest rights in the content and applications providers themselves, proclaiming a right to transmit their services to the user without interference or discrimination by the access provider. For example, in 2006, Professor Michael Geist remarked that:
While the definition of net neutrality is open to some debate, at the core is the commitment to ensuring that Internet service providers treat all content and applications equally with no privileges, degrading of service or prioritization based on the content's source, ownership or destination. (34)
Net neutrality proposals also vary with respect to how and where the rules could be implemented. Some are targeted at consumers in the access market. Some would forbid any type of price discrimination among consumers and users, including usage-based billings. Some would permit price discrimination if it reflects variability in consumers' demand and consumption such as usage-based pricing. (35) Others would permit any price discrimination among end users as long as it is not based on the type of content or applications. (36) Some forms of net neutrality rules are directed at the content and applications market. The prime rule championed by Lessig is to prohibit access tiering by forbidding high speed Internet access providers from contracting with content and applications providers for packet delivery or enhanced Quality of Service ("QoS"), even if the provider desires it and is willing and able to pay. (37) Lessig had also advocated banning ISPs from vertically integrating into the content and applications market altogether. (38) Some forms of net neutrality proposals would forbid any blocking or differential treatment of unaffiliated content and applications altogether, while others (these are more recent formulations as discussed later in the paper) would permit such blocking or differential treatment if they are done as part of a "reasonable" network management practice that is transparent to the user and perhaps developed with the cooperation of the content or applications provider. (39) Many of these proposals also combine any number of these forms without rigorous examination for internal consistency and practicability.
The potential for some forms of net neutrality proposals to prohibit differential pricing in general (including purely consensual and mutually beneficial business decisions between ISP and content and applications providers and consumer usage-based billing and bandwidth caps) led Sidak to liken net neutrality to Aesop's fable of "The Dog in the Manger," in which a dog prevents an ox from eating its hay simply because the dog cannot eat the hay itself. (40) The point is that net neutrality proposals tend to impose blanket restrictions on behaviour that is beneficial to and desired by some, because it is not desirable to others.
As noted previously, the original argument in the earlier works of Lessig and Wu was that as a result of insufficient competition in the high speed Internet access market, cable modem and ADSL service providers were imposing restrictions in that market that impeded competition and innovation in the Internet content and applications market. However, by 2005, both the FCC in the US and the CRTC in Canada had deregulated the high speed Internet access market on the basis that competition was vibrant and sufficient. And there was never any serious allegation of collusion, tacit or direct, between the cable modem and ADSL provider serving any local market as to merit an investigation by Canada's Competition Bureau under the Competition Act or the US Department of Justice under its anti-trust laws. Notwithstanding, Lessig, Wu and other emerging scholars like Barbara van Schewick argue in response that even if the access market is sufficiently competitive, the level of concentration in that market requires the application of a non-discrimination form of net neutrality regulation in order to ensure continuing competition and innovation in the content and applications market.
Lessig testified to the Senate in 2006 that "given the increasing concentration in broadband provision, the question whether ISP competition could protect end-to-end neutrality is now effectively moot. Whether or not competition among ISPs is enough, America no longer has sufficient broadband ISP competition. In most markets, an effective duopoly controls access to high speed Internet." (41) In their earlier testimony before Congress, Wu and Lessig argued that it was not enough to look at the level of innovation and investment under the status quo; rather, one must also consider whether the structural conditions could, in the future, create the environment for sub-optimal results. According to Wu and Lessig, "[t]he question an innovator, or venture capitalist, asks when deciding whether to develop some new Internet application is not just whether discrimination is occurring today, but whether restrictions might be imposed when the innovation is deployed." (42) In 2007, van Schewick made a similar argument that in the absence of network neutrality regulation, there is a real threat of anti-competitive discriminatory behaviour by broadband network owners and that "the threat of discrimination reduces the amount of application-level innovation [that is content, applications and portals] by independent producers," a reduction that is not offset by the corresponding increases noted in the network owners' innovations within the core. (43)
The theory behind this argument is not so much that competition or innovation and investment are sub-optimal today in the absence of a net neutrality regulation, but rather that ex ante codification of detailed and precise net neutrality regulation provides certainty that cable modem and ADSL service providers would not act anti-competitively because of their market share in the access market. (44) In general, ex ante regulatory regimes tend to set detailed and precise rules which a regulated entity must comply with before it makes any business decisions in the regulated field. In contrast, ex post regulatory regimes generally tend to set broad requirements and then evaluate the conduct of the regulated entity only in the event of a complaint and primarily in light of the impact of the conduct on competition. The argument for net neutrality regulation appears to be that although the high speed Internet market may be competitive, the level of concentration creates a risk of anticompetitive behaviour; therefore ex ante net neutrality regulation is necessary to give the market certainty that cable modem and ADSL service providers would not act in manners that could impede competition and innovation in the content and applications market. This "regulatory guarantee" argument was articulated by Skype in its arguments in CAIP v. Bell Canada as follows:
Even in an environment where there is a competitive Internet access marketplace, network neutrality rules are needed to prevent unreasonable interference with Internet content or applications, or favouring some content or applications over others including for anti-competitive purposes. While competition among network operators may serve to reduce anti-competitive actions, operators may still discriminate against unaffiliated content and service providers that rely on the Internet to deliver their content and services. (45)
Similarly, Google submitted in the same proceeding that prohibition of P2P network management practices is necessary because:
The open neutral internet (throttling-free) is what has driven the extraordinary innovation experienced over the past decade. Throttling [P2P traffic shaping] has a chilling effect, as innovators will not be able to develop new tools secure in the knowledge that users on every network will be able to access and run them. In turn, they will be less able and motivated to create new technologies in the first place.... Canadian carriers degrading particular lawful applications chosen by consumers threatens to reduce incentives to invest in infrastructure capacity. It encourages carriers to build their business model around managing scarcity, rather than developing more abundant capacity. (46)
C. Evaluating Net Neutrality Proposals Under the Policy Direction and Canada's Telecommunications Regulatory Framework
1. The Purpose of Net Neutrality Regulations
As described in the foregoing, all forms of net neutrality regulation are, at varying levels of specificity, addressed at the concern that despite a level of competition, concentration in the high speed Internet access market poses a risk that ISP behaviour could impede the robust competition and speedy innovation occurring in the content and applications market. In order to bring clarity to the analysis under the Policy Direction, we can generalize, with documented incidents, that when looked at in totality, all forms of net neutrality proposals are concerned that concentration in the high speed Internet access market could impede competition and innovation upstream in the Internet content and applications market as a result of one or more of the behaviours described below.
(a) Blocking, Degrading, Interfering with Content and Applications
A high speed Internet access provider could block its end users from accessing certain Internet content and using an application, or severely degrade the quality of such content or application. The form of regulation proposed to address this risk is to prohibit ISPs from blocking, degrading or interfering with access to legal content and applications. The classic case in Canada involved TELUS and the Telecommunications Workers Union (TWU).
On July 24, 2005, the Canadian Broadcasting Corporation (CBC) reported that TELUS, the largest telephone company and high speed Internet access provider in western Canada, blocked consumers from visiting a website run by the TWU during a labour strike by the TWU because it posted pictures of TELUS employees crossing the union's picket lines. (47) TELUS justified the blocking due to "an overriding need to protect the safety and privacy of our employees who were being targeted and the subject of intimidation...." (48) The Court of Queen's Bench of Alberta agreed and granted TELUS an interim injunction enjoining the TWU, its members and anyone else having knowledge of the order from posting for public viewing on any website any photographs or identifying features with the intent of intimidating or threatening TELUS employees, contractors, customers, suppliers and others. TELUS re-enabled access to the website as soon as the injunction was in place and discontinued the court action once the photographs intended to intimidate employees were removed.
The incident itself reveals the extremes of the concerns of both sides of the net neutrality debate--the concern that an ISP could block access to content for political or economic reasons, and the concern that net neutrality regulation could prevent a network operator from taking steps to do things on its networks necessary for security and safety of consumers (or employees, in this case). This theme is analyzed more critically later in the section on network management practices.
(b) Discriminatory Network Management Practices
A high speed Internet access provider could engage in network management practices that discriminate in terms of quality and speed of packet delivery between two content providers or between two applications. It could also discriminate between its own and non-affiliated content and applications providers. In general, net neutrality proposals would prohibit such network management practices subject to narrow "reasonable" constraints. For example, the TPR Report's net neutrality declaration would "take into account any reasonable technical constraints and efficiency considerations related to providing such access," (49) and the proposed Net Neutrality legislation, Bill C-552, would exempt discriminatory network management practices that are carried out "in a reasonable manner in order to relieve congestion." (50) However, the debate over what constitutes "reasonable" network management practice has proved to be as bipolar as the debate on the net neutrality ideal itself. Two recent cases that illustrate these concerns at a very detailed level are the Comcast case in the US and the CAIP v. Bell Canada case in Canada, both of which are discussed in greater length later in the paper. (51) In CAIP v. Bell Canada, independent ISPs alleged that Bell Canada's use of Deep Packet Inspection (DPI) to "shape" P2P file sharing traffic violates the rule against discrimination both under section 27(2) of the Telecommunications Act and under net neutrality principles. Content and applications providers including Google and Skype also filed comments arguing that such network management practices must be scrutinized heavily so as to ensure that they discriminate neither against non-affiliated content nor applications providers.
(c) Access Tiering or Price Discrimination Among Content and Applications Providers
A high speed Internet access provider could treat two content and applications providers unequally by contracting directly with one to provide priority delivery of packets or enhanced quality of service, for a fee. This practice, called "access tiering," is the major concern at the heart of much of the works of Lessig, for example. In his view, access tiering should be prohibited because of the risk it poses to competition and innovation. (52) Lessig and Robert W. McChesney argued that access tiering would enable ISPs "to sell access to the express lane to deep-pocketed corporations and relegate everyone else to the digital equivalent of a winding dirt road. Worse still, these gatekeepers would determine who gets premium treatment and who doesn't." (53) There has been no documented incident of access tiering in Canada. However, a recent incident in which an analogous issue was raised involved Shaw Communications, a Canadian cable high speed Internet access provider.
In late 2005, Shaw announced that it had an enhanced QoS technology on its network that could enhance the performance of VoIP service by reducing jitters and latency. Shaw proposed to offer this to the customers of unaffiliated VoIP providers like Vonage Canada for $10 per month. Vonage Canada made a submission to the CRTC, where it challenged Shaw. Vonage took the position that charging $10 for a service that Shaw's own voice telephone customers do not have to pay for was intended to price-squeeze independent VoIP providers. Vonage submitted that such price-squeeze was made possible as a result of insufficient competition in the Canadian high speed Internet access market:
Shaw's actions are also part of a bigger issue of network neutrality and who controls how Canadians use their Internet service. Vonage Canada wants to ensure that the monopoly telephone and cable Internet service providers don't restrict what services, applications or content Canadians can access. Canadians demand and deserve freedom of choice. (54)
The CRTC made no ruling on this particular submission, and it is widely believed that Shaw made a voluntary undertaking to pull the service. (55)
(d) Discriminatory Pricing of Consumer Access Service
A high speed Internet access provider could influence consumer demand for and consumption of certain Internet content and applications by charging different Internet access fees to consumers based on the type of content and applications that they access; or it may charge usage-based fees or impose bandwidth caps on consumer usage, or both. These concerns were reflected in Wu's 2002 survey of potential restrictions placed on consumers by cable modem and ADSL service providers. (56) However, as cable and ADSL providers removed the more objectionable usage restrictions such as home networking and device attachment, and as flat-rate pricing became entrenched in the industry, differential prices among consumers became less of a concern for net neutrality advocates and were even encouraged by some like Lessig. (57)
More recently, concerns over the need to recover the huge capital investment associated with building or augmenting for next generation networks to accommodate ever-increasing bandwidth-hungry content and applications are causing ISPs to rethink the current flat-rate "all you can eat" consumer high speed Internet pricing model, and to experiment with new revenue models, notably usage-based billing and bandwidth caps. The exponential growth of Internet content and applications that have been witnessed over the last ten years is expected to continue into the future. This will produce what Bret Swanson recently called "the coming exaflood" of real time bandwidth-intensive content and applications in the areas of medical imaging, video sharing, online gaming and digital/video surveillance. (58) At the same time, other services and applications are emerging that may not necessarily need high bandwidth but need high quality transportation and QoS delivery assurances--for example VoIP, alarm signaling and medical applications such as remote surgery. These so-called "QoS-needy" (59) applications will require broadband network owners to invest billions in building advanced intelligence into various layers of their networks to ensure timely and uncompromised delivery of packets. (60)
Many commentators are of the view that widespread metered consumer pricing and caps are imminent and efficient measures under these emerging conditions, and several cable modem and ADSL service providers are now introducing such usage-based measures. (61) As these are the very types of measures that Wu investigated and condemned as impeding net neutrality in 2002, debates over these new practices are beginning to morph into the larger net neutrality debate. After its network management practices were roundly condemned as unreasonable, in August 2008 Comcast announced its usage cap plan with an unprecedented maximum residential usage cap of 250 gigabytes of data transfer (uploading plus downloading). Notwithstanding that very few consumers will ever get anywhere near that cap with current video, critics berated Comcast's plan as foreshadowing things to come from other ISPs. Widespread usage caps, some now say, "is the end of the Internet as we know it." (62) In Canada, a legislative proposal, Bill C-555, proposes a regime under which telecommunications service providers can be compelled to provide to consumers accurate costs for services, including any "caps" and "overage," before contracts are signed; and to provide clear and accurate information with respect to network speeds and limitations on those speeds, including minimum, maximum, and average speeds obtained by customers. (63) Although Bill C-555 does not propose prohibiting usage-based billing per se, it is addressed at the emerging practice of usage-based billing and bandwidth caps. It seeks to compel disclosure to individual consumers at various stages in the customer relationship process and various levels of consumption. The level and scope of disclosure insinuated in Bill C-555 seems designed to discourage usage caps and usage-based billing. If implemented, the level of granularity of disclosure that Bill C-555 demands could make service providers reluctant to move from, or stop adopting, usage-based billing and bandwidth caps.
The foregoing has laid out the objectives behind specific forms of net neutrality proposals and rules. The pivotal question in testing compliance with the Government's Policy Direction under the CRTC's analytical framework is whether, despite the apparent concentration in the high speed Internet access market, these concerns can be addressed through reliance on market forces. Moreover, even if market forces cannot be relied upon and certainty through some intervention is desired, the Policy Direction requires that one look for alternative measures that minimally interfere with market forces and impose lesser costs than the forms of net neutrality regulations being proposed.
2. Defining the Relevant Market for Competitive Outcome Analysis
In evaluating the need for regulatory intervention in competitive telecommunications markets, both the CRTC and the Competition Bureau have established that the starting point is to properly define the relevant market under consideration. In Telecom Decision CRTC 94-19, the CRTC used the Bureau's framework to describe the relevant market for this purpose as "the smallest group of products and geographic area in which a firm with market power can profitably impose a sustainable price increase." (64) Each relevant market will have both a product and a geographic component.
Defining the relevant product and geographical aspects of the market to which a net neutrality regulation will be applied is necessary for evaluating the market failure which regulation is intended to address. Because net neutrality is often generally defined as the rights of consumers to access any legal Internet content and application of their choice, there is a perception that the relevant product market under consideration is the market for Internet access service--the access service subscribed to by end-users. There is, however, a second product market that is equally relevant to analyzing net neutrality regulation under the Policy Direction: the market occupied by the content and applications providers.
In a recent paper, Christopher Yoo described why net neutrality analyses should view broadband network providers as "intermediaries" in a "two-sided" product market. FIGURE I below illustrates Yoo's model showing an upstream component (Internet content and applications) and downstream component (broadband "last mile" or consumer high speed Internet "access"). (65)
[FIGURE 1 OMITTED]
Yoo observed that:
The fact that most consumers confront at best a last-mile duopoly has led many analysts to jump to the conclusion that the relevant markets are overly concentrated. In so doing, they ignore the fundamental insight that last-mile providers operate in what amounts to a two-sided market. In one side of the market, last-mile providers bargain with end users. On the other side of the market, last-mile providers bargain with content and applications providers. (66)
Yoo argued that while the geographical market for broadband "last mile" is local (most consumers are today served by their local area cable or telephone companies), the geographical market for Internet content and applications is national and perhaps even international. Last mile providers located in other cities cannot serve as substitutes for the consumer's own local provider, whereas Internet content and applications offered by different providers can serve as substitutes for one another even if they are located in another part of the country. Yoo concluded that absent significant market power or concentration upstream in the content and applications market, the fact that any one downstream Internet access provider may be able to block, degrade or overcharge for transmission of content and applications to a particular locality is of little consequence. The content and applications offered by different providers can serve as viable competitive substitutes. (67) In its final report to the Government of Canada, the TPR Panel also observed that:
In theory, the marketplace should take care of customers' interests in such situations. If customers feel strongly about restrictions on their ability to access other applications or content, they will make their feelings felt by switching either to another access provider or to a substitute application. However, if all access providers in a market decide to enter into such preferential arrangements, customers may be deprived of a real choice. In such a case, regulatory intervention to ensure a form of equal access to the application in question may be desirable. (68)
In Free Press v. Comcast, discussed later in this paper, Comcast made a similar argument which was rejected without analysis by the FCC. The FCC simply noted Comcast's argument that terminating peer-to-peer connections did not equate to blocking access to content because Internet users may upload such content from other sources. The FCC determined that:
[W]hether or not blocking content was Comcast's intent, Comcast's actions certainly had that effect in some circumstances. In any event, the semantic dispute of "delaying vs. blocking" is not outcome determinative here. Regardless of what one calls it, the evidence reviewed above shows that Comcast selectively targeted and terminated the upload connections of its customers' peer-to-peer applications and that this conduct significantly impeded consumers' ability to access the content and use the applications of their choice. These facts are the relevant ones here, and we thus find Comcast's verbal gymnastics both unpersuasive and beside the point. (69)
It is submitted that the FCC either misunderstood Comcast's argument or was overly swayed by other aggravating factors, as discussed later in this paper. This particular aspect of Comcast's comments asked that the FCC weigh the competitive impact of its network management practice in the upstream content and applications market and see if the targeted P2P users continue to have a choice. As argued in Part II of this paper, a competitive impact analysis is at the core of the Policy Direction's guidelines and the CRTC's traditional approach to section 27(2) of the Competition Act. Such an analysis, which will be referred to as the "competitive outcome approach" for the remainder of this paper, would have required that the FCC seriously consider the availability of alternatives to those affected by Comcast's practice, irrespective of whether the practice was "blocking" or "delaying."
The competitive outcome approach is consistent with the approach used by Canada's Competition Bureau in evaluating abuse of dominance under the Competition Act. The Competition Bureau is responsible for the administration and enforcement of the Competition Act. The statutory purpose of the Competition Act is to maintain and encourage competition in Canada in order to promote efficiency and adaptability of the Canadian economy; to expand opportunities for Canadian participation in world markets while recognizing the role of foreign competition in Canada; to ensure that small and medium-sized enterprises have an equitable opportunity to participate in the Canadian economy; and to provide consumers with competitive prices and product choices. (70)
In advancing its mandate under the Competition Act, the Bureau has produced a document--Information Bulletin on the Abuse of Dominance Provisions as applied to the Telecommunications Industry (71)--which outlines the Bureau's approach to a denial of access to facilities, functions or services under the Competition Act. The Bureau filed evidence and arguments in the regulatory proceeding recently held by the CRTC with respect to determining what facilities of the incumbent telephone companies and cable companies should be made available to competitors on a mandatory basis. The Bureau recognized that the threshold for mandatory regulated access under the Telecommunications Act needs to be set lower than the threshold for finding culpability under the strict anti-trust principles of the Competition Act. Nonetheless, the Bureau proposed that consistent with the general approach of the Draft TAB, the test of whether access to facilities, services and functions should be provided by a carrier to competitors is whether denial of access would result in a substantial lessening of competition in the market in which the facilities, services and functions are used as input. (72) The Bureau further explained:
[A]ssessing the ability and willingness of downstream consumers to substitute is a necessary component of assessing market power upstream. Demand for an upstream facility is a derived demand (i. e., it depends on the ability of consumers to switch to downstream substitutes offered by competitors that do not rely on the facility). If downstream competition is effective and consumers can engage in sufficient substitution, any attempt by the owner of the facility to exercise market power upstream will have limited, if any, effects downstream. This point is missed if only upstream substitution by competitors is assessed. (73)
If this approach were applied to ISPs in the two-sided market model, the question for evaluating any net neutrality regulation of ISP behaviour in the downstream access market is whether the impugned behaviour will substantially lessen competition upstream in the content and applications market. At the high level of analysis conducted in this paper, this approach will be called the "competitive outcome approach." We will see later that the same approach has been used by the CRTC to implement the common carrier provisions of the Telecommunications Act that are relevant to net neutrality.
(a) The High Speed Internet Access Market
In its 2008 Communications Monitoring Report, the CRTC reported that there were 500 companies providing Internet access services in Canada: 9% are traditional telephone companies, 18% are cable companies; 19% are utility companies using fibre, satellite or fixed wireless; and the remaining 54% are companies reselling the services of others. (74) As in the US, however, combined cable and telephone companies hold almost 90% of the total high speed Internet market--split 50/50 respectively. It is as a result of this that the high speed Internet access market is considered "concentrated" or "duopolistic" and why a net neutrality regulation is considered necessary under regulatory guarantee arguments. Nonetheless, under the Policy Direction and the analytic framework adopted by the CRTC, the relevant consideration is not that regulation makes it certain that firms in a concentrated market will not behave anticompetitively; rather, the question is whether market forces can ensure that they will not act anti-competitively.
Notwithstanding the dominant shares of cable modem and ADSL service providers, the Canadian consumer high speed Internet access market does demonstrate characteristics of competition in terms of pricing, service innovation and marketing. The prices for a given speed of service for both cable and ADSL high speed Internet access have declined in real terms compared to five years ago. (75) Independent industry analysts have also documented competitive behaviour between cable and telephone companies. A December 2007 study by GMP Securities documented recent pricing competition and concluded that:
The DSL providers may continue to use pricing promotions to mitigate the cable operators' telephony offerings in their territories and the cable operators will likely continue to benefit from the pull-through impact of the roll-out of their telephony services. Telus continued to offer big ticket promotional items for long-term contracts in its territories to mitigate the impact of Shaw's telephony and high-speed offerings. (76)
As further evidence of competition, both cable and ADSL providers have introduced variable speed products that allow consumers to tailor bandwidth and price to their needs. In 2008, as many as 25% of cable and ADSL consumer broadband customers subscribed to "Lite" broadband (below 1.5 mbps). (77) Moreover, new modes of entry have emerged that are offering competitive alternatives to cable modem and ADSL. They include satellite, wireless and power utility systems. (78)
From a competition analysis perspective, the CRTC has determined that market forces are sufficiently strong to justify deregulating the retail high speed Internet services market in Canada. (79) The CRTC's deregulation was carried out pursuant to its powers to forbear from regulation under section 34 of the Telecommunications Act. The Telecommunications Act sets out requirements for comprehensive fact-finding and application of competition law analysis with respect to an application for forbearance. (80) In addition, the Competition Bureau has not initiated a single investigation into the conduct or behaviour of high speed Internet access providers since they were deregulated.
It is understandable that concerns would exist because of concentration in the market, but from competition analysis perspective there is no demonstrated market failure that requires a net neutrality regulation. In the absence of collusion between cable and telephone companies, (81) market forces (the ability for consumers of content and applications to switch from one high speed Internet access provider to the other) provides certainty that that neither is likely to behave downstream in a manner that impedes competition in the upstream content and applications market. As Yoo argued, consumers could still migrate to the alternative provider even in a duopoly environment if their access to a particular content or application service was being blocked or compromised.
Further, the complementary nature of the two markets ensures that behaviour which harms competition in the content and applications market could also impact demand for and use of access, and therefore access revenues, downstream. In a general study of competition between complements, M. Keith Chen and Barry Nalebuff also analyzed complemetats in the context of net neutrality. Specifically, they examined the concern that a broadband access provider would have an incentive to degrade the quality of Skype and Vonage VoIP services in order to increase the attractiveness of its own competing internet phone service. Chen and Nalebuff concluded that this would not be profit maximizing since the access provider benefits from valuable complements such as VoIP services. They concluded that "... profit-maximizing internet service providers have no incentive to disrupt net neutrality." (82)
In her recent work, van Schewick provides additional arguments as to why a net neutrality regulation could still be required despite the existence of sufficient market forces that would otherwise have ensured that consumers could switch even between only two suppliers. According to van Schewick:
[S]witching costs may prevent consumers from changing providers to get access to the excluded application. This is the case, if the increased value from being able to use the excluded application is smaller than the costs of switching to another network provider. Thus, the higher switching costs, the lower the number of customers lost to other network providers. (83)
Switching costs could be a potential factor for inter-modal switching among cable, ADSL and WiFi, due to the need to purchase a new modem and the time-cost of reconfiguring the gear. However, it is doubtful if these costs will be higher than the value of switching in most cases. For the most part, modems are usually provided free or at discounts in promotions. A December 2007 study of broadband pricing in Canada shows that both cable network and ADSL network operators generally provide basic installation for free and full installation for free when combined with a term contract of at least 12 months. They also have modem lease and instalment payment plans that enable consumers to pay as little as $3/month for a modem. Accordingly, switching costs do not appear to be a compelling market failure that requires a net neutrality regulation. (84)
In line with van Schewick's arguments, it should be noted that it is increasingly the norm in Canadian telecommunications markets to bind customers to one or two-year contracts. The penalty for early termination could be a disincentive to switching service providers. Nonetheless, the consumer would still have the choice of comparing the cost of early termination to the cost of not having access to the desired content and application. Moreover, contractual commitments are not a universal barrier to switching. The relative costs and benefits of early termination versus the relative costs and benefits of access to the desired content and application would vary depending on the length of the unexpired portion of the contract. The smaller the unexpired portion of the contract and the lower the penalty, the less important the contract would be as a barrier to switching. Service providers have routinely been known to induce consumers to switch by undertaking to reimburse early termination charges from the other carrier.
Van Schewick also argues that if the exclusionary conduct manages to drive the producers of the excluded application or content from the market, switching providers will not enable consumers to get access to the excluded product. As a result, fewer consumers will switch in response to the exclusion. (85) The argument assumes that one of the "duopoly" players (call it ADSL-SP) did successfully behave anti-competitively in the downstream high speed Internet access market--for instance, by blocking or degrading a third-party content, resulting in the content provider exiting the market. If that were true, the fact that consumers could switch from ADSL-SP to its cable competitor (call it Cable-SP) was of no effect because the content provider was already out of business. This assumption is typical of regulatory guarantee arguments in that it involves ignoring or underestimating the results that the weight of economic theories posit are to be expected from a competitive market. In a competitive market, the threat alone that ADSL-SP could lose the affected customers to Cable-SP should dissuade ADSL-SP from blocking or degrading access to the content. But if ADSL-SP does so, it is unlikely to cause the content provider to exit the market. Many ADSL-SP customers who need/demand the content could switch to Cable-SP and the content provider would still be able to reach them.
Van Schewick has also argued that despite such threat, consumers may in fact be incapable of voting with their feet due to information failure. If consumers are not aware that blocking or degradation is the result of behaviour of the access provider rather than the content provider, they will be inclined to blame the content provider. (86) Applied to more contemporary issues of P2P traffic shaping, van Schewick's point is that if consumers erroneously believe that the video content provider (or even BitTorrent software) and not the ISP is responsible for delayed packets, they would not feel any need to switch access providers. In short, inaccurate information could impede consumers' ability to rationally exercise their choice even where there is competition. One could go further to suggest that such information failure could be compounded if the ISP actively misled the consumer as to the cause and source of the traffic downtime, as the FCC concluded in the Comcast case discussed later.
Given the rapidly changing nature of the Internet, the ever-increasing pressure to engineer and optimize networks for increased bandwidth usage and the emergence of new network management practices, this paper agrees that downstream information failure and information asymmetry between consumers and service providers is a concern that merits attention. Indeed, Part II of this paper suggests that net neutrality should be redefined as transparency in network management practices.
(b) The Internet Content and Applications Market
There is no serious disagreement that the upstream content and applications market itself is highly competitive. A detailed examination of the evidence and sufficiency of market forces in this market is not necessary for the analysis in this paper. The focus of most economic, academic and business arguments in favor of net neutrality regulation is on the potential impact of downstream behaviour on competition upstream. However, some net neutrality proposals are also directed at decisions that strictly potentially involve the ISP, as well as content and applications providers. Lessig, for instance, is vocal that access tiering violates net neutrality principles because it could involve: (1) discrimination between two content and applications (and between two content and applications providers); and (2) discrimination between a content and application (and provider) and the ISP's own or affiliated content and application. As a result, it is useful to briefly examine conditions in this market to see the extent that an ISP, even the major ISPs like cable modem and ADSL providers, can impede competition, investment and innovation in this market. It is submitted that given the number and economic bargaining power of major industrial players in the content and applications market, it is unlikely that ISPs could engage in any significant and sustainable discrimination among these non-affiliated content and applications providers. Moreover, given evidence that cable modem and ADSL service providers are not widely integrating vertically up into the content and applications market, there is very limited opportunity or incentive to discriminate in favour of their own or affiliated content and applications.
The term "content and applications" is used in net neutrality discussions to generalize a variety of video, audio and text information ("content") available via the websites on the world wide websites, as well as a variety of technologies ("applications") available to facilitate access to and distribution of this information. Although the line between Internet content and application is often extremely blurred, there is no need to go beyond these generalizations for the purposes of this paper.
A scan of the landscape indicates a market with major industrial competitors. Players in the Internet content space in Canada include the biggest corporations or their affiliates and divisions, such as CanWest, CTVglobemedia, Quebecor, Yellowpages, Bell Canada's Sympatico, CTV/MTV/TSN/CHUM, Canoe, Canada.com, CBC, Torstar, Rogers, Astral and Corus. The top global four search engines--Google,Yahoo!, MSN.com (Microsoft), and AOL LLC--with combined market value of over $500 billion, are accessible to Canadian consumers and businesses directly and via their Canadian.ca URL extensions. (87) Their combined share of the search market in March 2008 was 95.3%. (88) They also dominate the online video viewing market, comScore reported that in February 2008, Americans watched more than 10 billion videos online, with Google and its YouTube.com asset accounting for 35.4%. Fox, Yahoo! and Microsoft sites ranked second, third and fourth respectively. (89) comScore also reported that in March 2008, the top three search engines and content websites used by Canadians are Google, Microsoft and Yahoo! (90) The most visited social networking site in Canada in March 2008 (and fourth most visited site overall) was www.facebook.com, while eBay was the fifth most visited site overall. (91)
eBay owns three powerful content and applications products and brands on the Internet: (a) the eBay online auction site available globally and under individual national domain URLs (It is accessible to Canadian consumers and businesses via its US URL www.eBay.com and its localized Canadian URL www.eBay.ca.); (b) Skype (a VoIP service acquired by eBay in 2005 for $2.6 billion and accessible globally for (steeply) discounted international calling. It is available to Canadian consumers via www.skype.com or www.skype.ca); and (c) PayPal, a virtual bank through which any one in the world can pay for purchases online. Canadian consumers and businesses can access the payment applications via its secured URL at https://www.paypal.com or https://www.paypal.ca.
In addition to audio, Apple has aggregated several video and gaming entertainment contents on its website, which consumers can download to their computer or iPod for a fee. Canadian consumers can access content on the Apple global e-commerce site at www.apple.com, or they can go to the Canadian content site at www.apple.com/ca/ (English language) or www.apple.com/ca/fr/ (French language). The global e-commerce site includes feature films and popular American television programs like Saturday Night live, South Park, Desperate Housewives and Lost. Apple has signed deals with several networks including ABC, Disney, MTV, NBC, NBC Universal and USA Network. On December 12, 2007, Apple announced that Canadians would now be able to download content from the Canadian version of iTunes. (92) The Apple site boasts of "more than 6 million songs" at $0.99 each; "more than 11,000 music videos" at $1.99 each; "600 TV shows" at $1.99 an episode; "countless hours of iPod Games fun" at $4.99 each; "new and classic films" at $9.99 each; and "more than 30,000 audiobooks." (93)
In the few areas where Canadian cable modem and ADSL service providers have integrated upwards into some of the content space, they are only minor players against the established mainstream providers. In video delivery, for example, both Rogers and Bell Canada have online video stores that potentially compete against video distribution sites, as well as video distribution applications like P2P file sharing. But the online video distribution market is decisively dominated by Amazon, Netflix, Blockbuster and Apple. In 2007, NBC Universal and News Corp. joined forces to establish www.hulu.com, which now consists of a large selection of videos from more than 50 content providers, including FOX, NBC Universal, MGM, Sony Pictures Television and Warner Bros. (94) In some cases, each of the cable modem and ADSL service providers in any local market have partnered with a separate major content and applications provider, thereby further neutralizing the ability to influence the upstream market. For example, Rogers is partnered with Yahoo! for searches, while Bell Canada is partnered with MSN. There are no cable modem or ADSL service providers competing directly.
One of the original sources of net neutrality concerns was the potential for cable modem and ADSL providers to block or discriminate against companies that were using VoIP applications to compete against the cable and telephone companies affiliated with local and long distance voice operations. The notorious Madison River case (95) in the US was a case of actual blocking of a VoIP provider (the only one documented to date), while Vonage 's argument in the Canadian Shaw Cable case discussed earlier illustrated the perception that the access provider could corder undue advantage on its own affiliated voice service. Perhaps due to the maturity of the voice service market, incidents and allegations of blocking or anti-competitive discrimination against non-affiliated VoIP service providers are very rare nowadays, although parties continue to argue for nothing short of a regulatory guarantee to eliminate the possibility. (96) Instead, the focus has now shifted to a new battleground--video applications. This is not surprising because unlike the matured voice market, the market for video content and video applications is barely emerging and is expected to be the key driver of the consumer telecommunications sector for the coming years. (97) This is also the area where cable modem and ADSL providers are most actively vertically integrating into the upstream content and applications market.
Accordingly, notwithstanding competition in this space, the argument is made that some scrutiny still needs to be placed on cable modem and ADSL service providers with respect to third-party video content and applications. The specific third-party video application that is getting the most attention right now is P2P applications like BitTorrent, which is a protocol used to distribute video content on the Internet. At a high level, P2P applications allow consumers to download videos from providers like Joost and not from online video services of cable and ADSL service providers. Just as it was with third-party VoIP, it is argued that cable modem and ADSL service providers have the incentive to block or interfere with consumers' use of P2P applications in order confer a competitive advantage on their own video offerings. These issues are examined in additional detail in the discussion of Free Press v. Comcast and CAIP v. Bell Canada later. (98)
In recognition of the competitive state of the content and applications market, proponents often shift the justification for regulatory guarantee to the small players and the start-ups. They argue that regulation is still needed because practices such as access tiering and network management is likely to stifle entry and innovation by the next Google. (99) It is ironic that these arguments are often made by established players like Google who really have no incentives to see the next Google. (100) Intuitively appealing as concerns about "morn and pop" Internet start-ups may sound, they are largely a distortion of the reality and are often generalized without sound economic analysis. No study has shown how and to what extent ISP decisions downstream impede entry upstream. It is to be noted that much of these types of conceptual arguments emerged during the dotcom bubble days when the Internet content and applications market was dominated by overnight sensations. The market is dramatically different today. The weight of anecdotal evidence is that barriers to entry for anyone with a bright idea is just as low today as they were back then, but there is clear evidence as discussed above that every conceivable segment of that market is now occupied by huge industrial players.
Popular anecdotes about start-ups do not provide any additional strength for the case for net neutrality regulation, and it is fundamentally inconsistent with the Policy Direction to impose the type of blanket network management and price regulation advocated by some forms of network neutrality proposals on ISPs for the sake of protecting a fledgling new entrant--that would be a classic case of disproportionate regulation. Moreover, start-ups are not protected by direct regulation in any other sector and there is no reason why they should be in the Internet market. If there is a concern about Internet start-ups, it is more efficient to devise other direct means (e.g. public grants) to empower Internet users (if deemed socially desirable to do so) to get established in the market.
In sum, despite concentration, the downstream high speed Internet access market is sufficiently competitive to ensure that ISPs do not engage in discriminatory behaviour that could impede competition in the upstream content and applications market. Further, Canadian cable modem and ADSL service providers are not vertically integrated into the upstream market, which means there is very little incentive to want to impede competition upstream. Moreover, the upstream market itself is occupied by very large industrial players who are dominant in the few spaces upstream into which cable modem and ADSL service providers have integrated.
3. The Benefits and Costs of Regulatory Guarantee: Innovation and Investments Within and at the Edges of the Network
The Government's Policy Direction requires an examination of whether a net neutrality regulation minimally interferes with market forces both upstream and downstream and whether its constraints are proportionate to its objective. The analysis in this paper suggests that the central objective of net neutrality regulation is regulatory guarantee--a guarantee that in spite of competition service, providers will not act anti-competitively. The Policy Direction requires weighing the benefits and costs of the measures proposed to achieve that regulatory guarantee. It should be emphasized that proponents are not saying that in the absence of the desired regulatory guarantee, market forces are not sufficient to yield optimal levels of investment and innovations in the content and applications market. Rather, the role of the regulatory guarantee is to ensure that this optimal performance is guaranteed. (101)
It is paradoxical to admit that market forces have produced optimal levels of innovations and investments, but that the market needs regulatory constraints to protect these results. Even if there is a quantifiable marginal benefit from such regulatory guarantee, it is unlikely that it could justify the intrusiveness of many forms of net neutrality being proposed to achieve it. (102) Several econometric studies have suggested that the welfare impact of several forms of net neutrality regulations even in a duopoly high speed Internet access market is negative to de minimis at best. For instance, with respect to the form of proposal that would prohibit access-tiering, Cheng, Bandyopadhyay and Guo modelled two content providers who desired to avoid network congestion and packet delay by paying a monopoly broadband access provider for preferential delivery. They found that social welfare was increased when this form of net neutrality regulation was removed and one content provider paid for preferential delivery, but social welfare remained unchanged when both content providers paid. Depending on the variables, consumers were slightly better off or unaffected. However, they found that incentives for the monopoly broadband access provider to invest in the network were reduced if net neutrality regulation was abandoned. (103)
Nicholas Economides and Joacim Tag performed a similar econometric analysis using a duopoly market assumption in which the competing high speed Internet access providers were prohibited from charging fees to content and applications providers. They found that, comparing unconstrained duopoly with duopoly under net neutrality, total surplus was higher in net neutrality and both the content sector and the broadband access providers had higher profits. However, the consumers were worse off under net neutrality. According to Economides and Tag, this is because in the absence of net neutrality regulation:
[C]ompetition for consumers is more intense since profits from content providers can be competed away. As a result, consumers enjoy lower prices and are better off under no regulation than under net neutrality. Net neutrality regulation relaxes price competition, leading to higher profits for platforms. Platforms are better off under net neutrality, which is the opposite to what was the case in the monopoly model. (104)
Although these types of studies are still in their infancy and similar studies have not been performed for other forms of net neutrality rules, the conclusion that one can draw at this point is that the marginal benefits of the form of rule that prohibits access tiering suggests that the net benefits of these rules are small to negative. Moreover, even when the net social welfare is positive, the positive benefits are weighted in favour of content providers at the expense of consumers. It is particularly necessary to evaluate the impact of net neutrality regulation on innovation and investments in the network itself. Regulatory guarantee arguments tend to focus on innovations and investments on the edges of the network in the content and applications markets. Even if regulatory guarantee results in incremental innovations and investments at the edge, it is not necessarily desirable if it reduces incentives to innovate and invest in the core network. Important innovations have occurred and continue to occur in the network, and without continuing innovation and investment by broadband network owners in the core of the network, many of the content and applications blossoming at the edge would most likely not have occurred. Investment in the physical network is the vehicle for innovation both at the edges and in the core. The evolution from analogue to digital technologies, the ongoing change from circuit-switched to packet-switched networks, the enhancement of cable networks to support digital voice telephony, and the ongoing extension of fibre closer to customer premises are all innovations occurring at the physical network layer and opening up vast opportunities for innovators at the content and applications layers. (105)
In its 2008 Communications Monitoring Report, the CRTC noted that Canadian broadband network operators are investing heavily in physical capacity expansion, bandwidth maximization technologies as well as intelligence engineering for QoS. Investments include:
[E]nhancements to wireless networks, expansion of fibre-to-the-node (FTTN) facilities as well as expansion of wireless and DSL capacity and coverage.... Bell Canada, [TELUS] and Rogers Communication Corporation (Rogers) continued to expand their high-speed mobile network coverage to additional urban centres and various cottage country locations.... Total telecommunications capital expenditures were $ 8.2 billion in 2007, an increase of 18.7% from the $6.9 billion in 2006. Wireless capital expenditures increased by 12.7% to $1.9 billion in 2007. Wireline capital expenditures, representing approximately 77% of telecommunications capital expenditures, increased from $5.3 billion in 2006 to $6.3 billion in 2007, an increase of 20.6%. In 2007 the incumbent TSPs' capital expenditures were approximately 73% of total wireline capital expenditure, compared to 75% in 2006. (106)
Broadband network operators could face reduced incentives and ability to innovate in their networks if they are handicapped from deploying efficient network management technologies, prohibited from recovering the costs of network investments and restrained from experimenting with efficient revenue models on both sides of the two-sided Internet market. The magnitude of the impact of net neutrality regulation on innovation and investment would depend on the specific forms of the regulation. The more intrusive and general in its scope, the more likely it is to impair innovation and investment. Regulations that prohibit or limit pricing actions as well as those that prohibit outright, or prescribe regulator-imposed network management practices are generally in this category and may particularly impede the kind of investment required to support emerging bandwidth-intensive content and applications.
Litan and Singer have suggested that, faced with net neutrality regulations that prohibit contracting with willing content and applications providers, a broadband network operator could take one of two directions on enhanced quality of service: (1) refuse to invest in intelligence engineering to support enhanced QoS; or (2) invest in QoS, but offer the service as a standard feature and spread the cost across all content and applications providers, including those who are not QoS-needy. (107) Litan and Singer argue that it is conceivable that broadband network operators will simply not invest or will do so only in small quantities. The implications of that decision could be negative for QoS-needy content and applications--the demand for the content or application could decline or disappear altogether. If broadband network operators had not developed QoS support to eliminate latency (jitters) in VoIP, it may never have attained its potential as a substitute for circuit switched services due to quality problems. The consequence would likely be worse for such services as online video games, streaming multimedia, video conferencing, digital surveillance and remote medical diagnosis and surgery.
Litan and Singer observed that rather than refusing to introduce enhanced QoS altogether, broadband network owners would likely comply with a net neutrality regulation prohibition against enhanced QoS contracting by offering a blended, one-size-fits-all QoS to all content and applications providers and then bundle this as a standard feature. But a standard one-for-all QoS that is not sufficient to meet the requirements of certain QoS-needy content and applications providers would harm them by depriving them of the QoS needed to make their applications function at the optimal level. Even worse, it could harm the remaining vast majority of content and applications providers, who have no demand for such blended QoS by forcing them to pay for what they do not need. (108) In a recent column in the Financial Post, Hal Singer noted two key harms that could result in Canada if net neutrality regulation prohibits access tiering and priority' contracting with content and applications providers:
One perverse effect of net neutrality, if adopted in Canada, would be to chill needed investment in the Internet to support the real-time applications that consumers are increasingly demanding. Another effect would be to ensure that Canadian consumers will be saddled with paying the full cost of enhanced Canadian broadband networks of the future, with no contribution from the predominantly U.S. content providers, whose wares demand the enhanced capacity and which stand to profit from it. (109)
Similar results are likely to occur with respect to forms of net neutrality regulations that would prohibit or limit ISPs from charging usage-based fees to consumers or implementing demand and usage differentiating schemes such as bandwidth tiers and caps. Indeed, such regulations may harm consumers. The current "all you can eat" flat-rate model allows high-volume users to impose costs on low-volume users, thus creating a subsidy system. This form of net neutrality entrenches the subsidy and prevents ISPs from creating efficiencies by ensuring users bear their own costs. This inefficiency will likely compound as more investments are made in the network and the blended, average cost per user increases.
Richard Clarke of AT&T estimated that the cost for a broadband operator like AT&T to provide the bandwidth required to support current Internet applications (e-mail, text, music and some videos) and recover the costs directly from end users without charging the content and applications providers would be up to $47 per month per customer. (110) The access cost jumps to $140 per month if existing broadband networks are augmented to provide sufficient bandwidth to support current applications and the online equivalent of two simultaneous standard definition TV channels at home, and $466 per month for the equivalent of two simultaneous HDTV channels. (111) Clarke concludes that these are daunting figures for the vast majority of consumers and that it is unlikely that there will be a market for a model in which the cost of next generation broadband networks is recovered solely from end users. (112) Clarke's estimates are open to debate (and may not be relevant to cable high speed Internet operators because they provide standard or hi-def TV signals through existing cable TV channels), but the point is that if the costs of next generation networks cannot be recovered from consumers, regulation that prevents other market-driven models for cost recovery poses a risk of impeding optimal levels of investment in the network.
III. TOWARDS A LIGHTER, FOCUSED, LESS INTRUSIVE NET NEUTRALITY POLICY
A. Regulatory Guarantee vs. Other Measures to Reduce Downstream Competitive Risks
To recapitulate, the few econometric studies available do not suggest that regulatory guarantee can achieve additional benefits significantly above and beyond those that market forces can produce in the competitive markets today. However, even if regulatory guarantee will result in some tangible net benefits, the Policy Direction requires that one examine alternative measures that could be less harmful and intrusive than the proposed regulation while still attaining similar results.
Alfred Kahn's recent paper on the economics of net neutrality suggests areas in which one might look for such alternative measures. Although Kahn opposes net neutrality regulation in general, he nonetheless leaves room for other possible measures which could be adopted to address the concern over the impact of concentration in the broadband access market. These include: (a) using antitrust laws to prevent anti-competitive blocking or discriminatory treatment of non-affiliated content and applications providers; (113) (b) relying on government subsidised broadband network; (114) (c) explicit facilitation of a third broadband technology to cable and ADSL, including spectrum set-aside; (115) and (d) using the antitrust doctrine of "essential facilities" to facilitate ADSL and cable resale competition. (116)
Similarly, Christopher Yoo observed in a recent paper that network neutrality proponents are focusing on the wrong policy problem. (117) "By directing their efforts towards encouraging and preserving competition in the market for application and content, they are concentrating on the segments of the industry that are already the most competitive and the most likely to remain that way." (118) Yoo suggested that the focus of intervention should be on facilitating additional entry into the downstream access market:
The market power exercised by DSL and cable modem providers exists because of the limited number of options that end users have for obtaining last-mile services. The number of options will remain the same regardless of whether or not last-mile providers hold ownership stakes in content and application providers or whether unaffiliated content and application providers are granted nondiscriminatory access. Vertical disintegration thus has no effect on last-mile providers' ability to extract supra-competitive returns. Consumers will receive benefits only by promoting entry by alternative network capacity. (119)
Yoo's contention is that if the choice is between regulating market decisions of the established duopoly players through forms of net neutrality regulation that constrain pricing and network management decisions, on one hand, or facilitating additional entry into the broadband access market, on the other hand, the latter constitutes a more efficient, proportional and minimally intrusive alternative. (120)
Measures similar to those highlighted by Kahn and Yoo are already established parts of Canada's telecommunications regulatory framework. First, Canadian competition law authorities have the power under the Competition Act to address anticompetitive behaviour by ISPs, including cases of abuse of dominance and misleading advertising. A discussion of the Competition Bureau's framework, including its application to the denial or disruption of service in the telecommunications industry (which by definition would include denial of access to third party content and applications), is discussed in detail in a recent publication by the Bureau, The Rode of the Competition Act in the Telecom Sector. (121) The Bureau indicated that "[t]o the extent that the CRTC has forborne from regulating conduct relating to a telecommunications service or class of services, anti-competitive conduct can be dealt with under the Competition Act." (122) It is not necessary for this paper to enter into detailed discussion of this power or how the Bureau might apply it to specific issues of net neutrality in the forborne downstream market for high speed Internet access services. It is sufficient to note that the Bureau lists the following as possible areas of investigation and enforcement: (a) "Margin squeezing"--defined in section 78(1)(a) of the Competition Act as "squeezing by a vertically integrated supplier, of the margin available to an unintegrated customer who competes with the supplier, for the purpose of impeding or preventing the customer's entry into, or expansion in, a market"; (b) Targeted Pricing"--defined as "[o]ffering certain customers better prices than those offered to other customers"; and (c) "Denial of access to a facility" which has been argued (123) to potentially include access tiering complaints under US antitrust laws. (124)
The Canadian government and the CRTC have also taken significant steps to increase competitive alternatives in the downstream high speed Internet access market level, both through promotion of facilities-based competition and resale competition. (125) On November 28, 2007, the federal government released details on how the auction for Advanced Wireless Services (AWS) spectrum, to be held on May 27, 2008, will be conducted. Notably', of the 105 MHz of spectrum to be made available, 40 MHz will be set aside exclusively for new entrants to bid on. The other 65 MHz will be available to all bidders. In addition, new entrants also receive mandated roaming and tower sharing rights. According to Industry Canada:
The government believes that new entry will further enhance competition, not only in the wireless market segment, but across all telecommunications markets in Canada pro viding new products needed to keep Canada at the forefront of innovation. Having low cost for data transmission and state of the art wireless devices are important to Canada's competitiveness. The measures being taken are intended to ensure an opportunity for entry into the marketplace. The department is satisfied that the potential benefits of new entry' warrant these measures. (126)
AWS promise access to a growing range of innovative wireless applications and enable the timely rollout of next generation technologies like high speed video and Internet. Globalive Communications Inc., which currently offers wireline Internet and home phone services to some one million customers under the "Yak" brand name, won enough of the set-aside licenses to give it the rights to provide AWS in all the provinces of Canada except the province of Quebec. In addition, three established cable companies also won regional set-aside licenses--Shaw Communications in Western Canada, Quebecor's Videotron in Quebec and Bragg Communications' Eastlink in the Maritimes. As a result, there is now a national facilities-based wireless high speed Internet competitor against the incumbent cable and telephone companies, as well as at least one new regional facilities-based wireless entrant in each local market where cable and telephone companies were erstwhile the two dominant players.
In evaluating the need for regulatory guarantee via new forms of net neutrality regulation, it should also be recognized that under Canada's current regulatory framework, both the telephone companies and the cable companies are mandated to make their respective ADSL and cable modem networks available for resale by third parties. The CRTC also ordered the telephone companies to permit a competitor to sell ADSL on unbundled loops in instances where the customer does not want voice telephony. This regime of mandated access by competitors to underlying broadband network facilities is a key distinguishing factor between Canada and the US, and one which proponents of net neutrality in Canada may have underestimated. It is sometimes argued that net neutrality regulation is especially necessary in the US because of the FCC's decision to remove broadband operators from Title II regulation. (127) The FCC completed its initiatives to eliminate mandated access to broadband networks when in its 2005 Broadband Policy decision it relieved broadband network operators of the obligation to offer the transmission component or "bit stream"--of their networks to competitors on a mandated wholesale basis. (128) In contrast, following the Policy Direction, the CRTC conducted a proceeding in the fall of 2007 to review the regime of mandated wholesale services. InTelecom Decision CRTC 2008-17, Revised regulatory framework for wholesale services and definition of essential service, the CRTC set out a framework to phase-out currently mandated wholesale services that were determined to be "non-essential." (129) In the result, approximately 66% of wholesale services will continue to be made available on mandatory cost-plus basis.
Finally, in Canada, the development of broadband infrastructure in both urban and rural communities is widely recognized as top priority pursuant to the express objective under subsection 7(b) of the Telecommunications Act. In Telecom Decision CRTC 2006-9, (130) the CRTC determined that communities located in rural and remote areas unlikely to receive broadband services from any service provider in the near future could receive broadband expansion through monies accumulated in the deferral accounts of the incumbent telephone companies. (131) Deployment was to be based on least-cost technology and include backbone and access facilities. Telephone companies are to make backbone facilities funded through the deferral accounts available to other competitors at a minimal rate, and any broadband services are to be made available on wholesale basis to competitors in all funded communities. (132)
While there are currently no studies modelling the exact impact of these initiatives on competition in the downstream high speed Internet access market, in combination, they are concerted efforts to infuse more competition into the already competitive downstream high speed Internet access market. It is submitted that they further undermine the argument for regulatory guarantee via extensive net neutrality regulations.
B. Redefining Net Neutrality as Transparency in Network Management Practices
The central argument in this paper has been that the upstream content and applications market and the downstream high speed Internet access market are sufficiently competitive that the potential behaviours that net neutrality regulations seek to prevent are unlikely to result in any significant reduction in competition. The classic economic explanation is that in a competitive market, consumers would simply vote with their feet when faced with such behaviours. Van Schewick has argued that despite such a threat, consumers may be, in fact, incapable of voting with their feet due to information failure. The argument is that if consumers are not aware that blocking or degradation is the result of behaviour of the access provider rather than the content provider, they will be inclined to blame the content provider. (133) Thus, if Bell Canada's ADSL customers erroneously believe that the video content provider (or even the BitTorrent's software being used to download the video) is responsible for delayed packets rather than Bell Canada, they would not feel any need to switch to Rogers cable modem service. In short, inaccurate information could impede consumers' ability to rationally exercise their choice even where there is competition. In Free Press v. Comcast, discussed below, the FCC also noted that:
Although Comcast and certain other commenters contend that competition among broadband Internet access providers is sufficient to address any concerns regarding net work management practices, they do not address the effects of this information asymmetry between the broadband Internet access provider and its customers and competitors. (134)
The rest of this paper will show why recasting net neutrality as transparency in ISPs' network management practices makes the ideal more relevant and more valuable in today's competitive environment. Viewing net neutrality as network management transparency is consistent with, and indeed complementary to, the competitive outcome approach to analyzing regulatory intervention that is at the heart of the Policy Direction, Canada's competition law principles, as well as the CRTC's current regulatory framework. However, this proposition does not require that the regulator prescribe ex ante rules or standards of transparency. It is submitted that such standards should be encouraged and permitted to be developed within the industry by the ISPs and content and applications providers themselves.
Two highly-publicized cases in the US and in Canada illustrate how information failure and asymmetry of information downstream could potentially impact competition upstream in the content and applications market, and why transparency, rather than the popular forms of net neutrality, should be the solution. In August 2008, the FCC released a decision in Free Press v. Comcast in which it found that Comcast had engaged in unreasonable network management practices in violation of the FCC's net neutrality policy. In the Canadian case of CAIP v. Bell Canada, the CRTC dismissed an application in which third-party ISPs (supported by content and applications providers and consumer groups) alleged that Bell Canada's shaping of P2P file sharing traffic violated general net neutrality principles and sections 27 and 36 of the Telecommunications Act specifically. Each of the cases is analyzed in detail below.
1. Free Press v. Comcast--the FCC's Relapse into Regulatory Guarantee On November 1, 2007, Free Press filed with the FCC a complaint against Comcast, and asked the Commission to declare "that an Internet service provider violates the [Commission's] Internet Policy' Statement when it intentionally degrades a targeted Internet application." (135) The applications at issue were P2P file sharing applications like BitTorrent. BitTorrent is an open-source, peer-to-peer networking protocol that was once notorious for being used for illegal piracy, but has become increasingly popular for quickly transferring large amounts of legitimate digital content. Unlike traditional methods of file sharing, which typically require establishing a single transmission control protocol (TCP) connection between a user's computer and a single server, BitTorrent employs a decentralized distribution model. Each computer in a BitTorrent peerage is able to download content from other computers in the peerage, and in turn, each computer also makes content available for those same peers to download, all via TCP connections. Furthermore, a computer can download different portions of the same content from multiple computers simultaneously, with each computer providing a different portion of the same content (For example, a computer could obtain different portions of a video file from several different other computers in the peerage.).
This very self-scalability that is the core strength of P2P networking applications is also the reason why it poses a significant network management challenge to ISPs. When a P2P application attempts to download data, it creates numerous connections to other nodes in order to download pieces of the data from multiple end nodes, reassembling the data upon successful receipt of all the pieces. Each node in the network that wants a file can download and store the content (or pieces of the content) and act as a server for the other nodes in the network. As such, end-users could select files to download and let the content download in the background, and often overnight or over many days, until the file is complete. Consequently, activities such as P2P file sharing could overwhelm other traffic on the network. A P2P application, rather than opening up only one TCP session, will open up 40 to 100TCP sessions in an effort to transfer data as fast as possible using multiple peer sources. By self-scaling to other P2P applications on peer computers, the user will continue to expand the number of active streams, eventually consuming all available bandwidth and resulting in congestion. Once all the available bandwidth is being consumed, P2P applications will use queuing technique for additional requests until more bandwidth becomes available, thereby perpetuating the congestion.
To better manage the potential impact of P2P file sharing applications on their networks, broadband access providers like Comcast deployed throughout their networks Deep Packet Inspection (DPI) devices. The main function of DPI in this regard is to identify P2P file sharing packets by looking at the packet headers for protocols that are unique only to P2P traffic. Once P2P file sharing packets are identified, the real question then is what is done with the packets. It is on this question that Comcast's network management practices drew the ire of consumers and the FCC.
The FCC found that Comcast deployed equipment across its networks to monitor its customers' TCP connections using DPI to determine how many connections are P2P uploads. When Comcast found that there were too many P2P uploads in a given area, Comcast's equipment terminated some of those connections by sending packet instructions to "reset packets" (called RST Injection). Specifically, once DPI identified an area with P2P uploads above a predetermined level, such as when one BitTorrent user in that area attempts to share a complete file with another user via a TCP connection, Comcast's servers (through which its users' packets of data must pass) sent to each BitTorrent user's computer an RST packet that looked like it came from the other user's computer (rather than from Comcast (136)), which in turn either slowed down or completely terminated the connection. (137) The FCC concluded that Comcast's practice constituted prima facie impediment of Internet content and application, contrary to its Internet Policy Statement. (138) The Statement instructed providers of broadband Internet access services that "consumers are entitled to run applications and use services of their choice" (139) and "to access the lawful Internet content of their choice," (140) subject to "reasonable network management" practices. (141) The question then was whether or not Comcast's practices were "reasonable."
The FCC determined that the practices were not reasonable. Before describing the rationale behind the find, there are key aspects of the decision that are particularly relevant to the discussion in this paper. First, it is important to note that the FCC explicitly refused to make a general finding that any ISP network management practice that "intentionally degrades a targeted Internet application" violates the Internet Policy Statement as requested by Free Press. Indeed, recognizing the fast changing and dynamic nature of the Internet, the FCC stated that:
We decline to adopt prophylactic rules at this time because confining our holdings to a particular set of facts should provide guidance to consumers and the industry without unduly tying our hands should the known facts change. (142)
Further, the FCC also recognized that while broadband access providers may generally deploy DPIs in their networks, how they manage P2P file sharing identified by DPI may be unique and different, and "so specialized and varying in nature as to be impossible to capture within the boundaries of a general rule." (143)
It is noteworthy that the FCC did not make a specific finding against Comcast's use of DPI alone without RST Injection. The FCC reserved judgment on the use of DPI technology in general use when it stated that:
We agree that Comcast's use of Deep Packet Inspection here was unacceptable. However, we make no judgment on the use of this method for different purposes, such as distinguishing legal from illegal content. (144) [Emphasis added]
Finally, the FCC concluded that a case-by-case approach is more consistent with Congressional directives to reduce and minimize regulation, noting that:
[A] case-by-case, adjudicatory approach comports with congressional directives and Commission precedents. As discussed above, federal policy advocates the preservation of the "vibrant and competitive free market" for Internet and interactive computer services, and the Commission itself has recognized that "broadband services should exist in a minimal regulatory environment that promotes investment and innovation in a competitive market." Deciding to establish policy through adjudicating particular disputes rather than imposing broad, prophylactic rules comports with our policy of proceeding with restraint in this area at this time. (145) [Footnotes omitted]
The FCC's intention to defer to minimal interference resonates with the Policy Direction issued to the CRTC.
The FCC's Granular "Means and End" Proportionality Approach
The FCC's Internet Policy Statement states that the principles therein are "subject to reasonable network management." The FCC began by determining that in order to pass the test, Comcast's "justification for its practice must clear a high threshold." (146) The FCC elaborated that this requires the same analytical rigour as it would apply under the "heightened level of scrutiny" standard, a standard commonly used to review whether government action violates constitutional rights of individuals but which is seldom used with respect to private actions of individuals and corporations. (147) Applying the standard of review to network management practices, the FCC enunciated the following requirements:
1) The ISP's network management practice must further a "critically important interest"; and
2) There must be a tight fit between the ISP's network management practices and the alleged interest; (148)
The FCC assumed without deciding that the purpose for which Comcast deployed its network management practices was of critical importance namely relief network congestion. However, the FCC concluded that the means by which Comcast sought to achieve that purpose were "unreasonable" because the manner it deployed DPI and RST Injection was not carefully tailored to the objective of easing network congestion for the following reasons:
(i) Comcast impeded customers who use very little bandwidth as long as they use P2P application; and (ii) RST Injection was not limited to periods of congestion or neighbourhoods with congested nodes; instead, anyone who used P2P was impeded even if they were using their own nodes were never congested. (149)
(i) customers using other applications get away with extraordinary bandwidth utilization, as long as they are not using the targeted application (P2P). (150)
It is submitted that by applying the "heightened level of scrutiny" standard of review, the FCC went beyond evaluating whether Comcast's practices were reasonable to making certain that the practices hurt no one. In other words, the FCC applied a standard of regulatory guarantee. (151) By adopting the so-called "heightened level of scrutiny" "means & end" standard, the FCC forced itself into a granular analysis, looking to see a fit between highly technical network engineering decision on the means chosen to resolve an equally highly technical network engineering problem of congestion. As a result, the FCC ended up doing exactly what it proclaimed it wouldn't do--micromanaging an ISP's network management decision and instituting an invasive regulatory mandate.
Although the FCC stated that it would not lay down ex ante prophylactic rules in this case, it is submitted that is the outcome of ruling that Comcast should have calibrated its P2P file sharing shaping at such level of granularity as specific cable nodes and specific times of congestion. The FCC may have severely constrained ISPs' freedom to manage their network with the most efficient tools and technology available. It is particularly noteworthy that the FCC neither considered nor received evidence on whether or not such granular level of traffic management is technically possible and if it is, whether the potential costs to consumers and society as a whole justified it. In the CAIP v. Bell Canada application discussed later, Bell Canada noted that several parties criticized its P2P traffic shaping because they claim it does not address congestion in a focused manner. (152) For example, critics proposed that Bell Canada should have considered shaping only the P2P traffic of ISPs whose customers are known to cause congestion; (153) targeting individual end-users according to their P2P usage; (154) and shaping P2P traffic only during periods of actual congestion and/or only in nodes where congestion occurs. (155) However, as with the Comcast proceeding, there is no evidence that these levels of disaggregation are technically possible, and even if they are possible, there is no evidence that they are efficient and would not increase access costs to consumers. In its Answer, Bell Canada gave some information on the technical viability, noting that:
[W]hile it may explore a more granular application of bandwidth management in the future, the management tools that would be economically and technologically suitable for a telephone company using DSL network equipment to perform network management at the level of granularity and in the dynamic fashion suggested by these parties are not presently available in the market. The Company's present network management practices continue to be the best practical alternative to address congestion in its network for both retail and wholesale high-speed Internet end-users. (156)
By stipulating that ISPs should deploy DPI at this level of granularity, it is arguable that the FCC has essentially imposed what amounts, in effect, to prophylactic ex ante liability rules. The FCC appeared to be searching for a guarantee that an ISP's network management practice would not disadvantage any user.
2. CAIP v. Bell Canada--The CRTC's Competitive Outcome Approach
By application dated April 3, 2008, the Canadian Association of Internet Providers (CAIP) filed a complaint with the CRTC alleging that Bell Canada was illegally "throttling" wholesale ADSL services that it sold to the CAIP members under its Gateway Access Service (GAS) tariff. (157) GAS is a wholesale access service that creates a high speed data access path between a customer's premises and the network of an ISP. ISPs who subscribe to the tariff can use these high speed data access paths to provision a wide variety of retail access and content services to their customers, including high speed Internet access service, remote LAN access, streaming audio and video services, dedicated data and Voice over Internet Protocol (VoIP) services.
In 2007, Bell Canada deployed DPI on networks as an additional measure to address congestion on its ADSL network during peak periods (identified on Bell Canada's network as 4:30 p.m. to 2:00 a.m.). In its Answer to CAIP's application, Bell Canada stated that its "DPI revealed that a small number of users were generating a disproportionate percentage of the total traffic on the network, and that a substantial amount of that traffic was P2P" file sharing, including BitTorrent. (158) It configured its DPI to identify such P2P file sharing packets by looking at packet signatures that are unique only to P2P file sharing, thus ensuring that other applications similar to P2P--such as Skype's VoIP service or Joost's video streaming service--were not caught. (159) "Once the P2P file sharing traffic is properly identified, the DPI equipment implements a selective traffic shaping function on the identified P2P traffic." (160)
In its Application, CAIP argued that Bell Canada was applying the DPI not only to its own retail customers' traffic, but also to the traffic going to the independent ISPs' retail customers. As a result, several ISP end users were no longer able to enjoy the same bandwidth throughput that they had previously enjoyed on flat rate plans. CAIP argued that Bell Canada's action, as far as it affects independent ISPs and their retail end users, was illegal for several reasons, including: (a) violation of net neutrality principles; (b) failure to provide advance notification to ISPs under CRTC regulation and the terms of the GAS tariff; (c) influencing the meaning of or controlling the content in violation of section 36 of the Telecommunications Act; and (d) unjust discrimination against ISPs' traffic and undue preference of Bell Canada's own traffic in violation of section 27 of the Act. (161) CAIP also alleged that, as a provider of online video services, Bell Canada acted on an incentive to harm competitive Internet video services that use P2P applications. (162)
Subsection 27(2) of Canada's Telecommunications Act provides that: No Canadian carrier shall, in relation to the provision of a telecommunications service or the charging of a rate for it, unjustly discriminate or give an undue or unreasonable preference toward any person, including itself, or subject any person to an undue or unreasonable disadvantage. Subsection 27(4) provides that: The burden of establishing before the Commission that any discrimination is not unjust or that any preference or disadvantage is not undue or unreasonable is on the Canadian carrier that discriminates, gives the preference or subjects the person to the disadvantage. Further, section 36 of the Act states that: Except where the Commission approves otherwise, a Canadian carrier shall not control the content or influence the meaning or purpose of telecommunications carried by it for the public. (163)
In addition to independent ISPs who purchased wholesale ADSL high speed Internet access from Bell Canada, major Internet content and applications providers intervened in the application, including Google (164) and Skype, (165) on grounds that Bell Canada violated some principles of net neutrality. The forms of net neutrality rules put forward by these content providers vary from outright prohibition on network management (166) to allowing network management practices that are proportional and least restrictive for addressing congestion. (167) The presumed goal of these restrictions is to ensure that the concentration in the downstream high speed Internet access market does not impede competition and innovation in the content and applications market. Hence, Google submitted that:
Like other internet-based companies, Google relies on the communications infrastructure provided by carriers, in order to reach its ultimate end users. In Canada, this means reliance on the ILECs and cable companies that control the means of broadband access by Google's customers. Google's business is premised on making its services, content and applications available through the internet to any end user who chooses them, without restriction by any gatekeeper. (168)
This control, Google asserted, gave the telephone companies and cable companies the ability to "'throttle' competition" by favouring their own applications and content and by blocking or degrading the content or applications of others. (169) As a result, innovation would be impeded, "as innovators will not be able to develop new tools secure in the knowledge that users on every network will be able to access and run them." (170)
The CRTC dismissed the application, finding that Bell's traffic shaping: (a) is permitted under its wholesale tariff; (b) does not violate section 27 of the Actas the traffic shaping is applied to both Bell's own retail customers and the customers of the ISPs; and (c) does not violate section 36 because there was no traffic blocking and any delay experienced by customers did not amount to editorial control or influencing the meaning of their communications. (171)
Although the CRTC cautioned in CAIP v. Bell Canada that its decision was specific to Bell's wholesale service and the evidence in the case, the decision clearly assumes that sections 27 and 36 of the Telecommunications Act together with the guidelines given in the Policy Direction provide a complete code that can satisfactorily address net neutrality issues in Canada. This is further demonstrated by the fact that in setting out the follow-up public hearing into ISP traffic management practices, the CRTC largely defines the issues in terms of whether current and emerging traffic management practices are "consistent with the Act" or "in accordance with the Act," (172) and comments are invited specifically with respect to sections 27 and 36 of the Act. (173) Moreover, it is submitted that the decision embraced the view that a pivotal test of whether an ISP's network management practices violate these sections is whether or not the practices result or could result in significant impairment of competition in the content and applications market.
These conclusions are consistent with the CRTC's historical approach to issues of net neutrality. While there is no single official statement of net neutrality in Canada like the FCC's Internet Policy Statement, the CRTC has used sections 27 and 36 to enunciate broad net neutrality principles in various adjudicative proceedings. In particular, in Telecom Decision CRTC 2005-28, Regulatory framework for voice communication services using Internet Protocol, the CRTC established a comprehensive net neutrality framework with respect to access to VoIP in Canada. (174) Previously independent ISPs were prohibited from using the wholesale high speed Internet services purchased from cable companies to provide various applications, such as IP-based voice telephony service, (175) and wholesale ADSL were not allowed to be re-sold to provide VoIP service to consumers. (176) Reiterating the principle that the terms on which access to broadband networks is provided to third-party service providers by cable and telephone companies must not result in a preference of discrimination that is contrary to subsection 27(2) of the Act, (177) the CRTC removed these restrictions in Telecom Decision 2005-28 with respect to VoIP applications. (178)
In the proceeding leading to Telecom Decision 2005-28, independent ISPs and content and applications providers demanded that the Commission establish and impose ex ante rules of general application that are specific to net neutrality. Specifically, the CRTC was asked to impose a condition that: prohibits broadband network operators from engaging in packet degradation and "deep packet inspection"; (179) prohibits cable and telephone companies from imposing contract restrictions that prohibit their end-user subscribers from accessing independent content and applications providers, such as VoIP and dial-around long distance service providers, and then blocking the subscriber in order to enforce such contracts; (180) and directs that any quality of service (QoS) enhancements introduced by ADSL and cable network owners, such as packet prioritization, should be made equally available to all VoIP service providers on an unbundled basis. (181)
The CRTC concluded that ex ante net neutrality regulation was not necessary in light of subsection 27(2) of the Act. On the issue of blocking and contract restrictions:
The Commission considers that it is unnecessary to impose the proposed contract restrictions on broadband service providers. The Commission considers that it can rely on subsection 27(2) of the Act, where appropriate, to prohibit a Canadian carrier from restricting its broadband customers from dealing with ah alternative service provider of the customer's choice. This issue can therefore be addressed by the Commission on a case-by-case basis, should it arise. Such competitive disputes ate likely to be resolved by the Commission in a timely manner, using its expedited procedures. (182)
On the issue of packet degrading, the CRTC stated that:
[I]f such ah issue were to arise, the Commission considers that it can rely on subsection 27(2) of the Act, where appropriate, to prohibit Canadian carriers from intentionally degrading traffic. In the Commission's view, the existing regulatory framework is sufficient for dealing with such anti-competitive behaviour by a broadband service provider. As in the case above, such competitive disputes would likely be resolved by the Commission, in a timely manner, using its expedited procedures. (183)
With respect to enhanced QoS, the CRTC held in essence that market forces should be allowed to dictate what service quality improvements ate introduced into the network and that innovations in QoS would be hindered if broadband network operators ate forced to make enhanced QoS available to content and applications providers on a mandated, price-regulated basis. (184) In a subsequent decision on reconsideration, the CRTC emphatically confirmed its position that subsection 27(2) of the Act provides an adequate tool for addressing net neutrality concerns. (185)
Subsections 27(2) and 27(4) involve ah analysis similar to that employed by the FCC in the Comcast decision. Subsection 27(3) requires that the Commission shall make a finding of fact as to whether or not the practice is prima facie discriminatory. Once a prima facie finding is made, the burden rests on the ISP to demonstrate that the discriminatory practice is not unjust of that the alleged preference of its own or affiliated content and application is not undue. As discussed below, the CRTC has determined on a number of occasions that competitive outcome is the standard by which to measure whether discrimination is unjust or preference is undue under subsection 27(2).
The CRTC has held that where a customer has access to competitive alternatives, service denial of differentiation does not constitute unjust discrimination. In Telecom Decision CRTC 2006-33, Part VII application by Superior Wireless Inc. against TBayTel alleging unjust discrimination, the CRTC denied an application by Superior Wireless Inc. (Superior) alleging that its customers had been subject to unjust discrimination by TBayTel. Superior argued that TBayTel was denying Superior's customers digital roaming while allowing its own and the customers of other competitors' digital roaming. The CRTC concluded that while TBayTel did discriminate against Superior's customers, the discrimination was not unjust, noting that:
In assessing whether the actions of TBayTel amount to unjust discrimination, the Commission also considers that the degree of competition in the wireless market is a significant consideration. In Application by Microcell regarding alleged contraventions of section 27(2) of the Telecommunications Act by Rogers Wireless and Bell Mobility, Telecom Decision CRTC 2003-26, 28 April 2003, the Commission considered that the wireless market was characterized by rivalrous behaviour and was robustly competitive. The Commission considers that this assessment continues to be valid with respect to the current state of competition in the wireless market. In this regard, the Commission notes that in its Report to the Governor in Council: Status of Competition in Canadian Telecommunications Markets, October 2005, the Commission reported that the wireless market continued to display strong growth and to be competitive. (186) The Commission considers that the robustly competitive nature of the wireless market suggests that the Commission should exercise restraint with respect to the application of its powers under subsection 27(2) of the Act.
Similarly, in Telecom Decision CRTC 2003-26, Application by Microcell regarding alleged contra ventions of section 27(2) of the Telecommunications Act by Rogers Wireless and Ball Mobility, the CRTC denied an application by Microcell, a wireless operator, alleging that Bell Mobility and Rogers were in violation of subsection 27(2) because they offered discounted pricing to Microcell's customers in order to induce them to switch from Microcell when those discounts were not made available to other customers. The CRTC concluded that given "the robustly competitive circumstances of the wireless market," the conduct might be discriminatory but it could not be said to constitute unjust discrimination. (187)
In Telecom Decision CRTC 2004-66, Application by Shaw Cablesystems GP. against TELUS Communications Inc. concerning high speed Internet promotions, (188) the CRTC denied an application by Shaw Cablesystems G.P. (Shaw) requesting that the CRTC impose restrictions on certain pricing activities engaged in by TELUS Communications Inc. (TCI) with respect to its ADSL higher speed Internet services. "Shaw argued that TCI promoted its higher speed Internet at unusually large discounts, in terms of both the magnitude and length of time for which the discount was available." (189) Shaw argued that TCI's promotions, which were in effect for as long as 36 months, "were designed to discourage entry and to force existing competitors to exit the market for high speed Internet." (190) Shaw submitted that:
[B]elow cost pricing for an extended period of time was inconsistent with telecommunications policy and the public interest.... [B]y engaging in a practice of offering broadband Internet services at prices well below Shaw's estimate of costs for a lengthy period of time, TCI was conferring an undue preference on itself and an undue disadvantage on other ISPs. (191)
This, Shaw argued, was contrary to subsection 27(2) of the Telecommunications Act. The CRTC denied Shaw's application. It concluded that the broadband Internet service market was very competitive and that in fact Shaw had a much higher market share in that market than TCI and, as such, differential promotional pricing by TCI was unlikely to lessen competition. (192) The CRTC stated that given "the extent of competition in the provision of higher-speed Internet access services," the targeting of discounted pricing by TCI "to new customers would not result in unjust discrimination and would not provide TCI with an undue preference." (193)
The principle that emerges from these decisions is that in order for differential treatment of customers to violate subsection 27(2) of the Telecommunications Act, it must have a potential to substantially lessen competition in the markets affected by the differential pricing. Specifically with respect to network management practices such as P2P file sharing, it is submitted that there is no violation of net neutrality principles under subsection 27(2) if the practice does not significantly impede competition in the upstream market for such applications. The CRTC reached a similar conclusion in CAIP v. Bell Canada when it stated:
There is no basis to conclude, based on the record of this proceeding, that the implementation by Bell Canada of traffic-shaping measures on GAS was intended, as alleged by CAIP, to secure sufficient bandwidth for its own services or to prevent ISPs from effectively competing against Bell Canada's introduction of usage-based charges for its retail service. Further, the Commission considers that there is no evidence on the record to establish that Bell Canada has benefited from the implementation of traffic-shaping measures with respect to GAS in the manner alleged by CAIP. The Commission notes that Bell Canada provided data on the growth rate for GAS, which indicated that there was no substantive change in the growth rate alter implementation of its traffic-shaping measures on GAS. The Commission further notes that there is no evidence on the record to demonstrate or even to suggest that competition has been reduced following the implementation of Bell Canada's traffic-shaping measures. (194)
In contrast, in Comcast, the FCC concluded that Comcast's practices could impede competition and innovation in the applications market, but this finding was not supported by evidence showing the potential scale of competitive impairment. Of course it is true that some users would decide to stop using BitTorrent if they were no longer able to enjoy their desired speeds. However, there is no evidence to suggest that this would cause BitTorrent to go out of business. To the contrary, it is highly likely that the threat of losing users would create an incentive for BitTorrent to make improvements to its software that would make it less burdensome on networks. Indeed, in the wake of the Comcast incident and while the FCC investigation was ongoing, several P2P applications companies, including BitTorrent and Pando, banded together with Comcast to create so-called "P4P" protocol which, among other things, promised new ways to optimize file sharing and improve P2P download performance. (195)
Prior to CAIP v. Bell Canada, the CRTC had never directly considered net neutrality related issues under section 36 of the Telecommunications Act. In a recent application, the CRTC was asked to use Section 36 to order ISPs to block access to certain allegedly hate-speech web sites. The CRTC ruled that the scope of its power under section 36 "has yet to be explored," but determined that the power is permissive and not prohibitive. (196) Similar to subsection 27(2), section 36 involves a two-stage analysis: a prima facie finding of violation which then shifts the onus on the respondent to establish that the conduct or behaviour should be "permitted" by the CRTC. In CAIP v. Bell Canada, the CRTC found that Bell Canada's P2P traffic shaping did not constitute editorial control over or influencing the meaning of communication. However, the Commission did not indicate what standards it would have employed to determine whether or not to "permit" the practice had it been found, prima facie, to involve editorial control over or influencing of the meaning of the affected communication.
In defining the issues for the public hearing into ISP traffic management practices initiated by PN 2008-19, the CRTC has suggested that comments be structured along this two-stage analysis. The CRTC inquired that:
Section 36 of the Act states that unless the Commission approves otherwise, a Canadian carrier shall not control the content or influence the meaning or purpose of telecommunications carried by it for the public.
a) What, if any, Internet traffic management practices employed by ISPs would result in controlling the content, or influencing the meaning or purpose of telecommunications?
b) For any Internet traffic management practice identified in Ca), what criteria should the Commission apply in determining whether to authorize such practice? (197)
When the issue is whether or not to permit a particular traffic management practice that has been found to prima facie violate section 36, a competitive outcome analysis of the practice should be one of the considerations. This is particularly so in cases in which the ISP's traffic management is alleged to be degrading the quality of a competitor's content or application to a level tantamount to blocking or influencing its meaning. In such a case, it could be argued that the practice is not a violation of section 36 unless the control/influence of content significantly impedes competition in the market for that application or content.
This is consistent with the fundamental policy rationale behind section 36. The original purpose of section 36 was to ensure separation of control over "carriage" and "content" and was used to address the competitive risks of network owners who also provide content and applications directly or through affiliates. Carriers had to obtain "permission" under section 36 in order to integrate vertically into the content and applications market. (198)
In instances in which the CRTC had permitted carriers to engage in content under section 36, it did so after concluding that the particular content market was sufficiently competitive and in essence that the carrier was unlikely to impede competition in the content market. (199)
In sum, section 36 allows for an analysis similar to what the FCC undertook in the Comcast case: (1) Does the network management practice result in significant degradation or even outright blocking as to constitute control or influence? If so, there is prima facie violation of section 36; (2) Can the ISP establish that the practice should be permitted? If the degradation or blocking has no significant competitive impact (for instance if the content could be communicated or received via alternative comparable means), it could be permitted under section 36.
3. Anti-competitive Incentive Under the Competitive Outcome Approach
One of the problems with the" regulatory guarantee approach to net neutrality issues is that it risks introducing extraneous factors into the analysis of whether or not a particular behaviour is net neutral, such as incentive to act anti-competitively. Incentive to act anti-competitively is not the same as the likelihood of competitive impairment. Such a presumption could lead to an erroneous conclusion that a particular discriminatory network management practice is not reasonable or just, simply because the carrier has an incentive to competitively harm the provider of the content or application that offers a competing product.
The FCC appears to have made this error in its Comcast decision. The FCC noted that P2P applications, including those relying on Bitmorrent, have become a competitive threat to cable operators such as Comcast because Internet users have the opportunity to view high-quality video with BitTorrent that they might otherwise watch (and pay for) on cable television:
Such video distribution poses a particular competitive threat to Comcast's video-on-demand ("VOD") service. "VOD ... operates much like online video, where Internet users can select and download or stream an), available program without a schedule and watch it an), time, generally with the ability to fast-forward, rewind, or pause the programming." Comcast has recently placed a significant emphasis on expanding its VOD business and its VOD revenues have experienced robust growth. Moreover, Comcast has "begun incorporating its VOD content online through sites competing directly with BitTorrent protocol sites." (200)
CAIP made a similar argument in its complaint against Bell Canada. CAIP alleged that Bell Canada initiated P2P traffic shaping to facilitate the launching of an IPTV service and to give an undue preference to Bell Canada's newly launched Bell Video Store. (201)
In his dissenting opinion in Comcast, Commissioner Robert McDowell was critical of the manner by which the majority approached the issue of incentive:
[T]he majority does not address the issue of motive. The allegations before us boil down to a suspicion that Comcast was motivated not by a need to manage its network, but by a desire to discriminate against BitTorrent and similar technologies for anticompetitive reasons. If Comcast intended to harm its competitors, would it not have targeted other online video providers? Americans download more than eleven billion Internet videos per month, yet the record contains no evidence that Comcast is interfering with sites like YouTube which do not use pipe clogging P2P software. The record also does not speak to the fact that other prominent video sites, such as Joost, use more efficient P2P software that does not cause the same congestion problems as BitTorrent. As a result of their use of software that works better on existing networks, virtually no network management is needed. The majority's silence on this key exculpatory point is deafening. (202)
Using the competitive outcome test, it is irrelevant that the broadband operator has an incentive to discriminate against a content or application provider or to favour its own content or application. What is relevant is whether doing so is likely to significantly impair competition upstream in that market. Answering that question requires a structural evaluation of competitive conditions in the upstream market. It has been shown in Part II that the content and applications market is very competitive. Indeed, as Commissioner McDowell's statement implies, even if BitTorrent is impeded there are prominent alternatives and substitutes. In fact, for the most part, cable or telephone companies providing high speed Internet access service are not vertically integrated upstream into all the content and applications, and those that do provide direct or substitute contents and applications are highly unlikely to hold a significant market share. In its Answer to the CAIP application, Bell Canada responded that allegations of an ulterior motive connected with the launching of its IPTV and online video store wrongly assumes that the Company has significant market power in the online content distribution market. Bell Canada indicated that Apple's iTunes is the clear market leader with 91% of the TV video download market and 42% of the movie download market. "Even if the P2P file sharing market were in competition with the video download market, it is highly doubtful that the bandwidth management during peak periods of P2P file sharing applications will lead to the attainment of significant market power in the online content distribution market for [Bell Canada]." (203)
4. The Competitive Outdoor Approach and Transparency of Network Management Practices
As discussed in Part I, information failure and asymmetry between ISPs and their customers could be a cause of potentially significant negative competitive outcome in the content and applications market despite the existence of market forces in the downstream access market. Cases such as Comcast and CAIP v. Bell Canada demonstrate that, regardless of the absence of animus on the part of high speed Internet access providers, there could be a failure in the communication with or comprehension by consumers of ISPs' network management practices. In its ruling, the FCC determined that Comcast's failure to disclose its network management practices, while not determinative of fault, was a material factor in assessing the reasonableness of its actions. (204) The FCC remarked that:
Many consumers experiencing difficulty using only certain applications will not place blame on the broadband Internet access service provider, where it belongs, but rather on the applications themselves, thus further disadvantaging those applications in the marketplace. On the other hand, disclosure of network management practices to consumers in a manner that customers of ordinary intelligence would reasonably understand would enhance the "vibrant and competitive free market ... for the Internet and interactive computer services" by allowing consumers to compare and contrast competing providers' practices. (205) [Footnotes omitted]
Similar concerns were voiced by parties that filed comments in the CAIP application. (206) Although the CRTC ruled that Bell Canada did not violate the CRTC's network change notification rules, it did admonish Bell Canada for failing to provide ISPs some advance notice of the impending launch of its traffic shaping policy. The CRTC directed that in the future, Bell Canada should give ISPs at least 30 days notification which should, at a minimum, "provide clear and meaningful information describing what the changes are, what traffic: can be affected, under what conditions, and for how long." (207)
Unfortunately, proponents tend to use information failure as yet another reason why regulatory guarantees should be pursued with disproportionate levels of regulation, (208) when instead it should be viewed as a factor in a competitive outcome analysis, such as when considering whether an impugned practice violates section 27 or 36 of the Telecommunications Act. It is submitted that the competitive outcome framework of analysis proposed in this paper can be supplemented with a regime of transparent network management practices to provide a complete framework for addressing all net neutrality concerns. Indeed, transparency and disclosure should be a factor in assessing the competitive outcome of a network management practice where there is potential for information failure or information asymmetry between the service provider and its customers. Because the focus of the competitive outcome is the actual or potential impact on the competition upstream of the network management practiced downstream by the ISP, information failure or asymmetry between the ISP and its customers is relevant to assessing that impact. The proof of competition is a consumer's ability to make a choice. Information failure corrupts market signals respecting competitive choice. In contrast, when consumers are aware of an ISP's network management practice, they can make fully informed decisions on a number of competitive possibilities. For example, a customer may decide to switch to another ISP that does not shape P2P file sharing; another customer may decide to stop using BitTorrent application and receive his or her video content using other methods that are not subject to shaping. Yet another customer may decide to make no changes to either the downstream or upstream provider and simply re-arrange his or her own consumption habit.
Economist Douglas Haas has suggested that meaningful self-disclosure by a broadband network owner of its "non-neutral" policies could achieve the best balance between market forces that drive innovation and important interests in consumer right of access. (209) Haas proposed the adoption of standard net neutrality information disclosure model legislation called the Traffic Control Disclosure Act based on the US Fair Credit and Charge Card Disclosure Act. As Haas explained it:
Internet service providers and content providers alike would consistently disclose the specifics of their service offerings and traffic control policies in a uniform table. If designed correctly, this disclosure would help consumers more easily compare different service offerings. Given the vociferous and vocal opposition to the most egregious differentiation policies, public disclosures would likely discourage all but a few standard classes of service differentiation. (210)
This paper is not in favour of regulators creating a binding blueprint of disclosure rules, as to do so would be impracticable and inefficient. The FCC explicitly stated that it was avoiding doing so in Comcast, although as argued previously, the FCC may have done exactly that by taking a regulatory guarantee approach in its determination of the case. (211) An appropriate role for the regulator in this instance is to indicate in broad terms the goals that an ISP's network management disclosure policy should aim to achieve and then leave it to each ISP (or the industry) to devise their own disclosure standards. It is relatively easy to identify the goal of disclosure under the competitive outcome standard--it is primarily to ensure that consumers are fully aware of the ISP's network management practice and have information to allow them to make a decision with respect to continuing a relationship with that ISP in the competitive market. Using this as the broad goal, a relevant industry working group such as the CISC in Canada could engage in industry consultations to formulate the minimum content of a disclosure standard.
[S]uch vague terms are of no practical utility to the average customer. Of course there are "limitations" on the speed and bitrate of a customer's Internet connection, but even the best-informed customer would not have inferred from these or Comcast's other terms of service that peer-to-peer protocols were disfavored on Comcast's networks. (213)
Such general references are unlikely to pass the transparency standard under the competitive outcome test, as they do not provide sufficient information to enable the customer to make an informed decision. Luckily, ISPs have tremendous incentive to make their network management practice transparent to consumers. They are well aware that it is simply good business ethic and a necessity in this highly competitive sector to be as open and informative with their customers as possible. In addition to competitive forces, there are general consumer disclosure laws that could be used to force ISPs to disclose specific network management practices. Furthermore, the misleading advertising provisions of Canada's Cattleman Act could also potentially be evoked.
In short, ISPs have sufficient incentive to make their network management practices transparent; the challenge is that the Internet has developed at too rapid a pace for lawyers and legalese to keep up. A review of the product literatures, including websites, of leading broadband network operators indicate that significant information is being disclosed to consumers with respect to pricing, terms and conditions of broadband access. (214) However, there is no standard form or content for disclosing technical and network engineering management information such as P2P shaping, and some ISPs go further than others in terms of disclosure. For example, in April 2008, Rogers mailed to each customer an explanation of the introduction of caps with additional usage charges above the cap and a measurement of their last three months usage to show them where they stood prior to the implementation of usage-based billing. In addition, Rogers has implemented a form of browser-based "cap alert" warning system that pops up randomly to warn the customer that he or she is at a certain percentage level of his or her bandwidth cap for the month. These are encouraging initiatives. Perhaps the industry can voluntarily adopt best practices such as these as minimum standard requirements.
This is exactly what the Competition Bureau has done with respect to prepaid calling cards. On December 6, 2007, the Bureau announced "an initiative to ensure that consumers receive proper disclosure of any terms and conditions that affect the value and use of prepaid long distance telephone calling cards." (215) The type of transparency complaints that the Bureau said it received from consumers are similar to those that net neutrality proponents make with respect to caps on bandwidth. The Bureau stated that specific complaints include cards that provide less minutes than advertised, charge hidden tees (e.g. administration fees), and charge higher per-minute rates than advertised (e.g. higher tees for using the card at pay phones, or for calls made to or from cell phones). The Bureau's new minimum disclosure standard requires calling card businesses to:
* Disclose the effective rate per minute and the number of minutes available on the card;
* Disclose any conditions that might adversely affect the advertised rate per minute and number of minutes near the main body of the representation this information should not be printed in a small font or appear on a background that obstructs its visibility;
* Discontinue the use of fine print disclaimers that contain information contradicting the main message--the main body of the representation should not be misleading when read alone. (216),
A similar minimum standard can be developed under the auspices of the Bureau or the CRTC to identify key information that content and applications providers and consumers deem to be material to ensure that consumers can make informed choices when faced with an ISP's network management practice.
In addition to transparency and disclosure to consumers, voluntary consultation with potentially affected content and applications providers, to the extent practicable, could further enhance the ISP's position under the competitive outcome test. This is not a necessary condition to passing the test, but a practical matter that can help defuse any temptation to ascribe ulterior motive to the ISP. In some cases, a legal obligation may' exist pursuant to contract, tariff or regulation for the broadband operator to engage affected content and applications providers before implementing a network management practice.
In the vast majority of cases, there is no contractual, tariff or regulatory basis for the high speed Internet access provider to engage content and applications providers because there is usually no service agreement between them. It would be useful for the industry to investigate the possibility of establishing guidelines as to when it is desirable that broadband access providers engage the content and applications providers before implementing certain network management practices. It is imperative to recognize that the proposal here is not that ISPs should notify content and applications providers prior to making or implementing any network management decision. Not only would that be disruptive and inefficient, it could also place ISPs in a competitive disadvantage by making them disclose potentially competitive plans. However, it is possible that such consultation could be harmlessly undertaken prior to embarking on management practices of general application, such as blocking or slowing content that is streamed via a protocol known generally to create network congestion or other network problems. In this regard, it is noteworthy that prior to the FCC issuing its ruling, Comcast announced that it was going to implement a "protocol agnostic" network management practice. As part of that practice, Comcast has engaged several P2P application companies to discuss how they could do their own part to improve P2P download performance while Comcast does it part to optimize backbone bandwidth.
Notwithstanding its sudden popularity in Canada, the paradox of regulatory guarantees underlying most forms of net neutrality regulation proposals calls into serious question their consistency with Canadian telecommunications policy. Canada's policy is committed to minimal interference in the telecommunications sector and any form of net neutrality regulation proposal must show significant benefits beyond what the competitive market can already offer in the sector. Put differently, unless the behaviour or conduct sought to be constrained by net neutrality regulation substantially impedes competition in the Internet content and applications market, such regulation ought not to be introduced in Canada.
In Canada, the prevailing evidence is that the high-speed Internet access market is sufficiently competitive and the CRTC has forborne from regulating the rates, terms and conditions of cable modem and ADSL service providers. Canada's competition law authority, the Competition Bureau, has found no reason to believe that market forces are not robust or that sublime behaviour such as tacit collusion is impeding competition. Moreover, the Internet content and applications market itself has transformed dramatically from the cosmos of fledgling dotcom start-ups that it was at the turn of the decade into a highly competitive beehive of activities occupied by multi-billion dollar industrial players, many of which are vertically and horizontally integrated into major media and communications fields. Unlike what was feared earlier by network neutrality advocates, cable modem and ADSL service providers have barely integrated vertically into this field in Canada.
Accordingly, the current state of the downstream and upstream Internet markets significantly undermines the policy rationale behind most proposals for net neutrality regulation. Economic theories dictate that it is impossible for a cable modem or ADSL service provider to sustain a behaviour that significantly or substantially impairs the content and applications market under these circumstances. Rather than create a new prescriptive code of net neutrality where none is compellingly required, the old structural concerns of net neutrality should continue to be addressed in Canada under the broad guidelines of sections 27 and 36 of the Telecommunications Act, on a case-by-case basis, as the CRTC currently does. There is no need to call these issues "net neutrality" as they are simply and truly" competition issues. The test of whether any particular downstream behaviour violates these sections and requires redress should be whether or not the impugned behaviour is likely to substantially impede competition in the upstream content and applications market.
In order to remain relevant and valuable as a public policy ideal in this competitive environment, "net neutrality" needs to be redefined more narrowly outside its overly broad regulatory guarantee box. A net neutrality policy should complement, in a way that does not substitute or undermine, existing market forces. Such a complementary role can potentially be played in cases where there is information failure or asymmetry in the downstream high speed Internet access market. One area where economic and business indicators show potential information failure is with respect to ISP network management practices. The new net neutrality ideal should be to ensure transparency in ISP network management practices in order to correct and minimize such failure. Net neutrality as network management transparency complements the competitive outcome framework of the Policy Direction and sections 27 and 36 of the Telecommunications Act in that the more transparent a network management practice is to downstream access consumers, the less likely it is to have a significant negative effect on competition upstream in the Internet content and applications market. As such, the more transparent that a network management practice or a pricing decision is to the affected consumers, the less likely it is to have significant competitive impact upstream, and the less likely that it would violate section 27 or 36 of the Telecommunications Act.
It is important that transparency be conceived as a goal and not a regulatory mandate. Neither the CRTC nor any other regulatory body should establish ex ante rules or standards for what constitutes transparent (or other reasonable) network management practices. Rather, consistent with Canadian telecommunications policy's deference to market forces, ISPs and the industry should be encouraged to develop minimum standards for transparency.
(1.) In addition to the rules described in this paper, proponents seek several new rules against wireless carriers, including the obligation to permit consumers to use their own handsets, establishing standardized platforms and interfaces for wireless content and applications, and imposing consumer disclosure rules on wireless carriers. The seminal intellectual work on this new development is a provocative paper by Professor Tim Wu of Columbia Law School in which he cited several vertical restraints that he alleged wireless carriers impose on content/applications developers as well as end-user consumers. Wu identified four areas of concerns that justify the application of a net neutrality regulation to wireless carriers: (i) discriminatory broadband access practices; (ii) control of network attachments; (iii) excessive control and burden on content/applications entry in the wireless applications market; and (iv) "dummying" down of product design and features. Tim Wu, "Wireless Carterfone" online: (2007) 1 International Journal of Communication 389 <http://ijoc.org/>. For a critique of Wu's paper, see Robert W. Hahn, Robert E. Litan & Hal J. Singer, "The Economics of 'Wireless Net Neutrality'" (2007) 3 Journal of Competition Law and Economics 399; and also George S. Ford, Thomas M. Koutsky & Lawrence J. Spiwak, "Wireless Net Neutrality: From Carterfone to Cable Boxes" (2007) 17 Phoenix Center Policy Bulletin No. 17, online: Phoenix Center for Advanced Legal & Economic Public Policy Studies <http://www.phoenix-center.org/>. This paper will not discuss wireless net neutrality issues, although the same analytical approach can be used and the conclusions will most likely be the same.
(2.) In general, ex ante regulatory regimes tend to set detailed and precise rules which a regulated entity must comply with before it makes any business decisions in the regulated field. In the area of telecommunications regulation, the classical examples were regulations under the Telecommunications Act, S.C. 1993, c. 38, as am. by S.C. 2005, c. 50 requiring carriers to obtain prior tariff approvals before providing a regulated service to customers or changing the price or conditions of offering an existing service. In contrast, expost regulatory regimes generally tend to set broad requirements and then evaluate the conduct of the regulated entity only in the event of a complaint and primarily in light of the impact of the conduct on competition. Reliance on competition principles under the Competition Act, R.S.C. 1985, c. C-34 is a classic example of ex post regulation. For the purpose of this paper, any attempt to prescribe what constitutes "net neutral" behaviour (including detailed prescriptions for what constitutes "reasonable network management practices") for Internet Service Providers (ISPs) is considered analogous to ex ante regulation.
(3.) The objectives in section 7 of the Telecommunications Act, ibid. are: "(a) to facilitate the orderly development throughout Canada of a telecommunications system that serves to safeguard, enrich and strengthen the social and economic fabric of Canada and its regions; (b) to render reliable and affordable telecommunications services of high quality accessible to Canadians in both urban and rural areas in all regions of Canada; (c) to enhance the efficiency and competitiveness, at the national and international levels, of Canadian telecommunications; (d) to promote the ownership and control of Canadian carriers by Canadians; (e) to promote the use of Canadian transmission facilities for telecommunications within Canada and between Canada and points outside Canada; (f) to foster increased reliance on market forces for the provision of telecommunications services and to ensure that regulation, where required, is efficient and effective; (g) to stimulate research and development in Canada in the field of telecommunications and to encourage innovation in the provision of telecommunications services; (h) to respond to the economic and social requirements of users of telecommunications services; and (i) to contribute to the protection of the privacy of persons."
(4.) See discussion in III.B. 2--CAIP v. Bell Canada--The CRTC's Competitive Outcome Approach.
(5.) See discussion in II.A--The Government's Policy for Economic Regulation in the Telecommunications Sector, and II.C--Evaluating Net Neutrality Proposals Under the Policy Direction and Canada's Telecommunications Regulatory Framework.
(6.) See discussion in II.B--Normative Goals of Net Neutrality Regulation.
(7.) Supra note 2.
(8.) Free Press v. Comcast, No. 08-183 (F.C.C. 2008), 2008 WL 3862114, online: Federal Communications Commission Daily Digest, Vol. 27 No. 162 <http://www. fcc.gov/Daily_Releases/Daily_Digest/2008/dd080820.html>[Comcast]. Comcast has filed an appeal of the FCC's decision to the Federal Court.
(9.) The Canadian Association of Internet Providers' application regarding Bell Canada's traffic shaping of its wholesale Gateway Access Services (20 November 2008), Telecom Decision CRTC 2008-108, online: CRTC <http://www. crtc.gc.ca/eng/archive/2008/dt2008_108.htm> [CAIP v. Bell Canada].
(10.) Telecom Public Notice GRTC 2008-19, Review of the Internet traffic management practices of Internet service providers (20 November 2008), online: CRTC <http://www.crtc.gc.ca/eng/archive/2008/pt2008-19.htm> [PN 2008-19].
(11.) J. Gregory Sidak, "A Consumer-welfare Approach to Network Neutrality Regulation of the Internet" (2006) 2 Journal of Competition Law and Economics 349 at 411.
(12.) There are works from the left to the right of the intellectual spectrum, including welfare economics (See e.g. ibid.; Barbara van Schewick, "Towards an Economic Framework for Network Neutrality Regulation" (2007) 5 Journal on Telecommunications and High Technology Law 329); game theory (See e.g. Hsing Kenneth Cheng, Subhajyoti Bandyopadhyay & Hong Gun, "The Debate on Net Neutrality: A Policy Perspective" [draft discussion paper] (2008), online: Social Science Research Network <http://ssrn.com/abstract=959944>; See also Douglas A. Haas, "The Never-Was-Neutral Net and Why Informed End Users Can End The Net Neutrality Debates" (2007) 22 Berkeley Tech. L.J. 1565); property rights (See ibid. at 370-83. See also Bruce M. Owen & Gregory L. Rosston, "Local Broadband Access: Prinam Non Nocere Primum Processi? A Property Rights Approach" [draft discussion paper] (2003), Stanford Institute for Economic Policy Research No. 02-037, online:
<http://siepr.stanford.edu/papers/pdf/02-37.html>); constitutional rights (See e.g. Randolph]. May, "Net Neutrality Mandates: Neutering the First Amendment in the Digital Age" (2007) 3 I/S: A Journal of Law and Policy for the Information Society 197); and anti-trust laws (See e.g. Christopher S.Yoo, "What Can Antitrust Contribute to the Network Neutrality Debate?" online: (2007) 1 International Journal of Communication 493 at 503 <http://ijoc.org/> [Yoo, "Antitrust"]; and Alfred E. Kahn, "Telecommunications: The Transition from Regulation to Antitrust" (2006) 5 Journal on Telecommunications and High Technology Law 159).
(13.) Order Issuing a Direction to the CRTC on Implementing the Canadian Telecommunications Policy Objectives P.C. 2006-1534, S.O.R./2006-355, C. Gaz. 2006.11.140.
(14.) Industry Canada, Telecommunications Policy Review Panel: Final Report (Ottawa: Industry Canada, 2006) [TPR Report].
(15.) Ibid. at 6-16.
(17.) Ibid. at 6-16 to 6-17.
(18.) Ibid. at 6-18.
(19.) Ibid. at 6-18 (Recommendation 6-5(b)).
(20.) Supra note 13, ss. 1(a)(i)-(ii), 1(b)(ii); see also s. 1(c)(ii). The guidelines in the Policy Direction that are relevant to the analysis in this paper are those directing the CRTC to:
1.(a)(i) "rely on market forces to the maximum extent feasible as the means of achieving the telecommunications policy objectives;"
1.(a)(ii) "when relying on regulation, use measures that are efficient and proportionate to their purpose and that interfere with the operation of competitive market forces to the minimum extent necessary to meet the policy objectives;"
1.(b)(ii) If using economic measures, ensure they "... neither deter economically efficient competitive entry into the market nor promote economically inefficient entry."
(21.) Action plan for the review of Commission regulatory treasures in light of Order in Council P.C. 2006-1534 (11 July 2007), Telecom Decision CRTC 2007-51, CRTC Telecom Decisions, Public Notices, Circulars and News Releases (Ottawa: CCG-Publishing, 2007) vol 15:8, online: CRTC <http://www.crtc.gc.ca/eng/archive/2007/dt2007-51.htm>.
(22.) See e.g. Sidak, supra note 11; Robert Hahn & Scott Wallsten, "The Economics of Net Neutrality," (2006) 3:6 The Economists' Voice, Article 8, online: The Economists' Voice <http://www.bepress.com/ev>; van Schewick, supra note 12, for a thoughtful article supporting Net Neutrality regulation on the basis of consumer welfare.
(23.) See e.g.Yoo, "Antitrust", supra note 12; and Kahn, supra note 12. For an earlier work, see Joseph Farrell & Philip J. Weiser, "Modularity, Vertical Integration, and Open Access Policies: Towards a Convergence of Antitrust and Regulation in the Internet Age," online: (2003) 17:1 Harvard J.L. & Tech. 2 <http://jolt.law.harvard.edu>.
(24.) For a sampling of US scholarship, see supra note 12. Canadian publications are also emerging, see e.g. Steven Globerman, "A Policy Analysis of Net Neutrality" (June 2008), online: Fraser Institute Digital Publications <http://www.fraserinstitute.org>; Craig McTaggart, "Net Neutrality and Canada's Telecommunications Act" (Paper prepared for the Fourteenth Biennial National Conference on New Developments in Communications Law and Policy, Law Society of Upper Canada, Ottawa, 25-26 April 2008), online: Social Science Research Network <http://ssrn.com/abstract=1127203> [McTaggart, "Net Neutrality"]; Craig McTaggart, "Was the Internet Ever Neutral?" (Paper prepared for the 34th Research Conference on Communication, Information and Internet Policy, George Mason University School of Law, Arlington, Virginia, 30 September 2006), online: The School of information at the University of Michigan <http://web.si.umich.edu/tprc/papers/2006/593/mctaggart-tprc06rev,pdf>; Neil Barratt & Leslie Regan Shade, "Net Neutrality: Telecom Policy and the Public Interest" (2007) 32 Canadian Journal of Communication 295. For emerging global and international perspectives, see e.g. Filomena Chirico, Ilse van der Haar & Pierre Larouche, "Network Neutrality in the EU" (TILEC Discussion Paper, Tilburg University, DP 2007-030, September 2007) online: European Parliament <http://www.europarl.europa.eu/document/activities/cont/200805/ 2008057ATT28494/20080507ATT28494EN.pdf>; Working Party on Telecommunication and Information Services Policies, "Internet Traffic Prioritisation: An Overview" Oganisation for Economic Co-operation and Development (6 April 2007), online: Organisation for Economic Co-operation and Development <http://www.oecd.org/dataoecd/43/63/38405781.pdf>; and Viktoria Kocsis & Paul W.]. de Bijl, "Network Neutrality and the Nature of Competition Between Network Operators" (Report prepared for the Directorate General for Energy and Telecommunications, Ministry of Economic Affairs, The Netherlands) (2007) 4 International Economics and Economic Policy 159.
(25.) For perspectives on this historical evolution, see e.g. Mark Cooper, "Open Communications Platforms: The Physical Infrastructure as the Bedrock of Innovation and Democratic Discourse in the Internet Age" (2003) 2 Journal on Telecommunications and High Technology Law 177 at 180-81 [Cooper, "Open Communications Platforms"]; Marjory S. Blumenthal & David D. Clark, "Rethinking the Design of the Internet: The End-to-End Arguments vs. the Brave New World" (2001) 1 ACM Transactions on Internet Technology 70; Mark A. Lemley & Lawrence Lessig, "The End of End-to-End: Preserving the Architecture of the Internet in the Broadband Era" (2001) 48 UCLA L. Rev. 925 at 930-34 [Lemley & Lessig, "The End of End-to-End"]; Tim Wu, "The Broadband Debate, A User's Guide" (2004) 3 Journal on Telecommunications and High Technology Law 69 [Wu, "The Broadband Debate"].
(26.) Asymmetry Digital Line Service (ADSL) technology enables the telephone companies to separate a data path on top of the voice channel in the traditional twisted copper pair used to provide voice service. The data channel is then used to provide services such as high speed Internet access.
(27.) Lawrence Lessig, The Future of Ideas: The Fate of the Commons in a Connected World (New York: Random House, 2001) at 43-48, 155-76 and 240-49.
(28.) Tim Wu, "Network Neutrality, Broadband Discrimination" (2003) 2 Journal on Telecommunications and High Technology Law 141 [Wu, "Network Neutrality"].
(29.) See ibid.; Wu, "The Broadband Debate", supra note 25; Letter by Timothy Wu & Lawrence Lessig, Ex Parte Submissions, CS Dkt. No. 02-52, 22 August 2003; Lemley & Lessig, supra note 25; Cooper, "Open Communications Platforms", supra note 25; Jim Chen, "The Authority to Regulate Broadband Internet Access Over Cable" (2001) 16 Berkeley Tech. L.J. 677; Mark Cooper, "Open Access to the Broadband Internet: Technical and Economic Discrimination in Closed, Proprietary Networks" (2000) 71 U. Colo. L. Rev. 1011 ; William P. Rogerson, "The Regulation of Broadband Telecommunications, the Principle of Regulating Narrowly Defined Input Bottlenecks, and Incentives for Investment and Innovation" (2000) 2000 U. Chicago Legal F. 119.
(30.) In Wu & Lessig, ibid., their 2003 ex parte submission stated: "The question an innovator, or venture capitalist, asks when deciding whether to develop some new Internet application is not just whether discrimination is occurring today, but whether restrictions might be imposed when the innovation is deployed" at 8.
(31.) One of the most comprehensive user-centric declarations was propounded by the then FCC Chairman Michael Powell in 2004 in what he called the "Four Freedoms": the unfettered ability of users to access content, use applications, attach personal devices, and obtain service plan information. Michael K. Powell, "Preserving Internet Freedom: Guiding Principles for the Industry" (Remarks at the Silicon Flatirons Symposium on The Digital Broadband Migration: Toward a Regulatory Regime for the Internet Age, University of Colorado School of Law, 8 February 2004), online: Federal Communications Commission <http://hraunfoss.fcc.gov/edocs_public/attachmatch/DOC-243556Al.pdf>. In 2005, under new Chairman Kevin Martin, the FCC formally adopted a version of the declarations as a "non-binding" policy statement (see U.S., Federal Communications Commission, Appropriate Framework for Broadband Access to the Internet over Wireline Facilities, F.C.C. 05-151 (2005) at 14986).
(32.) TPR Report, supra note 14 at 6-18 (Recommendation 6-5).
(33.) Part VII Application by the Canadian Association of Internet Providers Requesting Certain Oreders Directing Bell Canada to Cease and Desist from "Throttling" its Wholesale ADSL Access Services at para. 90 (3 April 2008), Telecom Decision CRTC 2008-108, online: CRTC <http://crtc.gc.ca/public/partvii/2008/8622/c51_200805153/885770.PDF>.
(34.) Quoted online: Net Neutrality Canada <http://www.neutrality.ca/>.
(35.) A recent net neutrality, legislative proposal in Canada, Bill C-552, proposes to allow, among other exceptions, ISPs to "offer directly to each user service at different prices based on defined levels of bandwidth or the actual quantity of data flow over a user's connection." Bill C-552, An Act to amend the Telecommunications Act (Internet Neutrality), 2d Sess., 39th Parl., 2008, cl. 36.1 (2)(d) [Bill G552]. World Wide Web inventor, Tim Berners-Lee, remarked that "Net Neutrality is this: If I pay to connect to the Net with a certain quality of service, and you pay to connect with that or greater quality of service, then we can communicate at that level.... Net Neutrality is NOT asking for the internet for free. Net Neutrality is NOT saying that one shouldn't pay more money for high quality of service. We always have, and we always will." Tim Berners-Lee, "Net Neutrality: This is serious" Decentralized Information Group (DIG) Breadcrumbs (21 June 2006), online: Decentralized Information Group (DIG) <http://dig.csail.mit.edu/breadcrurnbs/node/144> [emphasis in original]. Similarly, in his 2006 testimony before the Senate, Lessig stated that: "To oppose access-tiering, however, is not to oppose all tiering. I believe, for example, that consumer-tiering should be encouraged. Network providers need incentives to build better broadband services. Consumer-tiering would provide those incentives. Consumer-tiering, however, should not discriminate among content or application providers. There's nothing wrong with network owners saying 'we'll guarantee fast video service on your broadband account.' There is something wrong with network owners saying 'we'll guarantee fast video service from NBC on your broadband account.' And there is something especially wrong with network owners telling content or service providers that they can't access a meaningful broadband network unless they pay an access tax." U.S., Net Neutrality: Heating Before the Senate Comm. On Commerce, Science & Transportation, 109th Cong. (2006) at 2-3 (Lawrence Lessig), online: United States Senate Committee on Commerce, Science and Transportation <http://commerce.senate.gov/pdf/lessig-020706.pdf> [Lessig, 2006 Senate Testimony].
(36.) Lessig also testified to the Senate that charging providers content and application-specific fees (called "access-tiering") undermines net neutrality, regardless whether the fees are charged non-discriminatorily among providers. "By 'access-tiering,' I mean any policy by network owners to condition content or service providers' right to provide content or service to the network upon the payment of some fee. These fees are independent of basic Internet access fees. No one questions the right of network owners to charge Google for the bandwidth it uses. Instead, 'access-tiering' adds an additional tax on network innovators based upon the particular service being offered." Lessig, 2006 Senate Testimony, ibid. at 2, n. 2. See also Wu, "Network Neutrality", supra note 28 at 151-54.
(37.) Lessig, 2006 Senate Testimony, supra note 35.
(38.) Lessig testified in an earlier 2002 Senate Hearing that "[s]eparating control over the use of the network from ownership of the wires that make-up the network is a necessary step to restoring the growth and innovation of the original Internet." U.S., Heating on The Government's Role in Promoting the Future of Telecommunications Industry and Broadband Deployment Before the Senate Comm On Commerce, Science, and Transportation, 107th Cong. (2002) at 3 (Lawrence Lessig), online: United States Senate Committee on Commerce, Science and Transportation <http://commerce.senate.gov/pdf/lessig.pdf>.
(39.) For example, the Powell principles later modified into the FCC's Internet Policy Statement and Canada's Bill C-552, both of which are discussed in detail below.
(40.) Aesop, "The Dog in the Manger" in Aesop, Aesop's Fables (New York: Grosset & Dunlap, 2000) at 1, cited in Sidak, supra note 11 at 441, n. 380. The moral of this fable of Aesop's is that "Some begrudge others what they cannot enjoy themselves."
(41.) Lessig, 2006 Senate Testimony, supra note 35 at 5.
(42.) Wu & Lessig, supra note 29 at 8.
(43.) van Schewick, supra note 12 at 378.
(44.) Sidak, supra note 11 at 406-07.
(45.) Comments submitted by Skype Communications regarding Part VII Application by Canadian Association of Internet Providers Requesting Certain Orders Directing Bell Canada to Cease and Desist from "Throttling" its Wholesale ADSL Access Services (12 June 2008) at para. 31, Telecom Decision CRTC 2008-108, online: CRTC <http://www.crtc.gc.ca/public/partvii/2008/8622/c51200805153/920240.pdf> [Skype Submissions].
(46.) Comments submitted by Google regarding Part VII Application by Canadian Association of Internet Providers Requesting Certain Orders Directing Bell Canada to Cease and Desist from "Throttling" its Wholesale ADSL Access Services (3 July 2008) at para. 17, online: CRTC <http://www.crtc.gc.ca/public/partvii/2008/8622/c51_200805153_1/923481.pdf> [Google Submissions].
(47.) "Telus cuts subscriber access to pro-union website" Canadian Broadcasting Corporation (24 July 2005), online: CBC News <http://www.cbc.ca/canada/story/2005/07/24/telus_sites050724.html>.
(48.) TELUS, News Release, "Alberta court grants interim injunction against posting TELUS employee photos" (28 July 2005), online: TELUS <http: //about.telus.com/cgibin/ media_news_viewer.cgi?news_id=605&mode=2&news_keywords=website>.
(49.) TPR Report, supra note 14 at 16-18 (Recommendation 6-5(b)).
(50.) Bill G552, supra note 35 at s. 36.1(2)(a).
(51.) See below at III.B.1--Free Press v. Comcast--The FOC's Relapse into Regulatary Guarantee and III.B.2--CAIP v. Bell Canada--The CRTC's Competitive Outcome Approach.
(52.) Lessig stated in his testimony before Congress that "[a]ccess-tiering will create an obvious incentive among the effective duopoly that now provides broadband service to most Americans. By effectively auctioning off lanes of broadband service, this form of tiering will restrict the opportunity of many to compete in providing new Internet service. For example, there are many new user generated video services on the Internet, such as Google Video, YouAre.TV, and youTube.com. The incentives in a world of access-tiering would be to auction to the highest bidders the quality of service necessary to support video service, and leave to the rest insufficient bandwidth to compete. That may benefit established companies, but it will only burden new innovators." Lessig, 2006 Senate Testimony, supra note 35 at 2.
(53.) Lawrence Lessig & Robert W. McChesney, Editorial, The Washington Post (8 June 2006) A23.
(54.) Johanne Torres, "Vonage Canada Asks CRTC to Investigate Shaw's VoIP Tax" TMCnet (8 March 2006), online: TMCnet <http://www.tmcnet.com/news/2006/03/08/1440827.htm>.
(55.) Access to the Quality of Service Enhancement Service of Shaw Cablesystems G.P. (Shaw) and Packet Cable functionality of Rogers Communications Inc., Shaw, and Videotron ltee(21 September 2006),Telecom Decision CRTC 2006-61 at para. 24-25, online: CRTC <http://www.crtc.gc.ca/eng/archive/2006/dt2006-51.htm>. Vonage Canada's challenge of Shaw's enhanced QoS fee was raised in comments it submitted in support of a petition filed before the CRTC by another Canadian VoIP service provider, Cybersurf, in which Cybersurf requested that the CRTC direct Canadian cable modem operators to provide mandatory access to the enhanced QoS functionality of their packet cable. The CRTC rejected Cybersurf's petition and made no specific finding on Vonage's submission with respect to Shaw's enhanced QoS fee.
(56.) Wu, "Network Neutrality" supra note 28.
(57.) See e.g. Lessig on consumer access tiering in Lessig, 2006 Senate Testimony, supra note 35.
(58.) Bret Swanson, "The Coming Exaflood" The Wall Street Journal (20 January 2007), online: The Wall Street Journal <http://online.wsj.com/>. See also Bret Swanson & George Gilder, "Estimating the Exaflood: The Impact of Video and Rich Media on the Internet--A 'zettabyte' by 2015?" Discovery Institute (29 January 2008), online: Discovery Institute <http://www.discovery.org/a/4428>.
(59.) Litan & Singer use the term "QoS-needy" for these services: see Robert E. Litan, & Hal J. Singer, "Unintended Consequences of Net Neutrality Regulation" (2007) 5 Journal on Telecommunications and High Technology Law 533 at 536 [Litan & Singer, "Unintended Consequences"].
(60.) Estimates to build and connect consumers to Verizon's FiOS range from $1,200 to $1,700 per household, with a median of about $1,400 per household. See e.g. Bruce Meyerson, "Verizon discloses fiber-optic initiative" The Providence Journal (28 September 2006), online: The Providence Journal <http://www.projo.com/business/content/projo_20060928_vz28x.2d4bef0.html>; Daniel Marcus, "IPTV: The last mile, fiber or copper?" IPTV, online: Broadband Bananas <http://www.broadbandbananas.com/content/view/192/47/>; "Project LightSpeed" Broadband DSL Reports.com(21 October 2004), online: Broadband DSL Reports.com <http://www.broadbandreports.com/shownews/55799/>.
(61.) See e.g. Bell Canada Answer submitted by Bell Canada regarding Part VII Application by Canadian Association of Internet Providers Requesting Certain Orders Directing Bell Canada to Cease and Desist From "Throttling" its Wholesale ADSL Access Services (11 July 2008) at paras. 102 and 104, online: CRTC <http://www.crtc.gc.ca/public/partvii/2008/8622/c51200805153_1/926702.zip> [Bell Canada Answer], where Bell Canada initiated usage-based billing in December 2006 when it ceased marketing its service as "unlimited traffic" to new customers. Bell Canada stated recently that "[w]hile continued investments in network capacity must remain at the core of a carrier's network management solution, this type of incremental investment will not be sustainable without changes to existing revenue models and the introduction of new business models. Part of the issue is that current revenue models were not designed to handle the dramatic increase in demand due to online video.... [C]onsumers have been paying for services and not MB, and thus, consumers' willingness to pay for a service currently has no correlation with the amount of bandwidth that a service consumes.... As a result, in addition to expanding opacity, Bell Canada uses pricing-based mechanisms to manage network traffic such as bandwidth caps and usage-based pricing. This type of differential pricing is beneficial to both consumers and carriers because it matches prices to use, which provides the appropriate incentives for optimal network use and expansion."
(62.) See Jeffrey Lindsay, "Broadband Usage Caps--The Camel Gets Its Nose Under the Tent Flap" Berntein Research (5 September 2008) at 1-3, where Lindsay argued that "Like metered dial-up before it, many internet investors see metered broadband as unnecessary, anti-consumer and anti-Internet. Some see it as a clear indicator that there is insufficient competition in last mile connectivity." Lindsay further remarked that "Metered Pricing Would Likely Chill Internet Innovation ... Many think that unmetered usage has been a major factor in the U.S. Internet success story. It allows developers and users to try things without fear of running up a huge bill. In short, it enables serendipity. Imagine how MySpace or Facebook would have played out if users were constantly worrying whether their downloads and usage will blow this month's budget. How many parents might cancel a broadband account if they get an unexpected $200 bill from the cable company? Many think that metered broadband will actually accentuate the digital divide and will be felt hardest among the poorest families in America--almost the exact opposite of the fair use argument put forward by the cable MSOs." See also Om Malik, "Comcast Metered Broadband Official--Beware What You Download" GigaOM(28 August 2008), online: GigaOM <http://gigaom.com/2008/08/28/comcast-makes-metered-broadband- official-beware-what-you-download/>
(63.) Bill C-555, An Act to provide clarity and fairness in the provision of telecommunication services in Canada, 2nd Sess., 39th Parl., 2008, ds. 3(a)(i),(iv). Cf. Bill C-552, supra note 35, which will allow ISPs to "offer directly to each user service at different prices based on defined levels of bandwidth or the actual quantity of data flow over a user's connection" as an exception to net neutrality.
(64.) Review of Regulatory Framework (16 September 1994), Telecom Decision CRTC 94-19, online: CRTC <http://www.crtc.gc.ca/eng/archive/1994/DT94-19.HTM>, CRTC Telecom Decisions, Letter Decisions, and Public Notices (Ottawa: CCG-Publishing, 1994) vol 2:29.
(65.) "Last mile" and "access" are used as synonyms in this paper to refer to the service that connects the end user to the Internet.
(66.) Yoo, "Antitrust", supra note 12 at 512. At 512, n. 86, Yoo also noted that the two-sided model " ... is, of course, something of an oversimplification. In reality, the industry also encompasses backbone providers as well as last-mile providers serving business customers, as well as providers of hardware, software and numerous other complements."
(67.) Ibid. at 512-14.
(68.) TPR Report, supra note 14 at 6-18.
(69.) Comcast, supra note 8 at para. 44.
(70.) Competition Act, supra note 2 at ss. 1.1,7(1).
(71.) Competition Bureau Canada, Information Bulletin on the Abuse of Dominance Provisions as Applied to the Telecommunications Industry (Gatineau: Competition Bureau Canada, 2008), online: Competition Bureau Canada <http://www.competitionbureau.gc.ca/epic/site/cb-bc.nsf/en/02690e.hmal> [Draft TAB].
(72.) See generally Reply Argument of the Commissioner of Competition (7 December 2007), submitted regarding Revised regulatory framework for wholesale services and definition of essential service (3 March 2008), Telecom Decision CRTC 2008-17, online: CRTC <http://crtc.gc.ca/public/partvii/2006/8663/c12_200614439/839811.zip>. At para. 7, the Bureau argued that" ... mandating access to wholesale facilities and services should be the exception and not the rule. In particular, the Bureau submits that mandatory access should only be used as a regulatory tool only where (i) it is both necessary and the most effective means to control the market power of the owner of the facility in the relevant downstream market in which the facility is used as an input; and (ii) it is likely to have a substantial positive impact on competition in that market. Where this is not the case, the imposition of mandatory access will yield little or no benefit to end-users and may in fact prove harmful to competition, and ultimately consumers."
(73.) Argument of the Commissioner of Competition (23 November 2007) at para. 67, submitted regarding Revised regulatory framework for wholesale services and definition of essential service, supra note 72, online: CRTC <http://crtc.gc.ca/public/partvii/2006/8663/c12_200614439/839828.zip> [emphasis in original].
(74.) Canadian Radio-television and Telecommunications Commission, Communications Monitoring Report 2008 (Ottawa: Canadian Radio-television and Telecommunications Commission, 2008) at 207-08, online: CRTC <http://www.crtc.gc.ca/Eng/publications/reports/PolicyMonitoring/2008/ cmr2008.pdf> [Communications Monitoring Report].
(75.) Ibid. at 213. The price decline and its significance can be erroneously underestimated or overlooked if one ignores the dynamism of "high speed Internet" access service as a consumer product and the innovation occurring within the network. Hence, some critics would point out that the list price of any particular consumer high speed Internet access package has been increasing in the last few years for example, the basic entry level service has increased from approximately $39 in 2001 to $47 in 2008. However, the speed for the entry-level product has also increased three to four times, and bundling discounts of up to 15% are available on various packages. These discounts could be very large when high speed Internet service is bundled with other consumer offerings. For example, even at 5%, given that the discount applies to total billing, a $5 savings can be achieved on a $ 100 bill, which is typical for television and Internet bundles.
(76.) GMP Securities, "The Broadband Brief, Fourth Quarter 2007" (1 April 2008) at 13.
(77.) Communications Monitoring Report, supra note 74 at 215.
(78.) Ibid. at 213.
(79.) See e.g. Independent Members of the Canadian Association of Internet Providers Digital Subscriber Line Internet services by Bell Canada and Bell Nexxia (27 June 2002),Telecom Decision CRTC 2002-37, online: CRTC <http://www.crtc.gc.ca/eng/archive/2002/dt2002-37.htm>, CRTC Telecom Decisions, Letter Decisions, and Public Notices (Ottawa: CCG-Publishing, 2002) vol 10:6; IMCAIP against certain incumbent cable and telephone carriers--Provision of higher speed access and retail Internet services including Lite service (5 May 2004), Telecom Decision CRTC 2004-28, online: CRTC <http://www.crtc.gc.ca/archive/ENG/Decisions/2004/dt2004-28.htm>, CRTC Telecom Decisions, Letter Decisions, and Public Notices (Ottawa: CCG-Publishing, 2004) vol 12:3.
(80.) Telecommunications Act, supra note 2, s. 34. In particular, see s. 34(2): "Where the Commission finds as a question of fact that a telecommunications service or class of services provided by a Canadian carrier is or will be subject to competition sufficient to protect the interests of users, the Commission shall make a determination to refrain, to the extent that it considers appropriate, conditionally or unconditionally, from the exercise of any power or the performance of any duty under sections 24, 25, 27, 29 and 31 in relation to the service or class of services"; and s.34(3): "The Commission shall not make a determination to refrain under this section in relation to a telecommunications service or class of services if the Commission finds as a question of fact that to refrain would be likely to impair unduly the establishment or continuance of a competitive market for that service or class of services."
(81.) While it is generally agreed that collusion is impossible, in the recent CRTC's Essential Facilities hearings some content and applications providers argued that ADSL and cable operators have sufficient incentive and opportunity in any given local market to engage in behaviour that results in "coordinated effect." See e.g. Written Argument of Cybersurf Corp (23 November 2007) at paras. 31-49, submitted regarding Revised regulatory framework for wholesale services and definition of essential service, supra note 72, online: CRTC <http://crtc.gc.ca/public/partvii/2006/8663/c12_200614439/833799.zip>. Under the Competition Bureau's Merger Enforcement Guidelines, coordinated effects arise "when a group of firms ... is able to profitably coordinate its behaviour because of each firm's accommodating reactions to the conduct of others" (Competition Bureau Canada, Merger Enforcement Guidelines (Gatinean: Competition Bureau Canada, 2008), online: Competition Bureau Canada <http://www.competitionbureau.gc.ca/epic/site/cbbc.nsf/en/01245e.html>). However, the Bureau's expert witness testified that there is no evidence to support the kind of "coordinated effects" alleged by Cybersurf in the Canadian broadband access market (See e.g. Transcript of Proceedings Before the Canadian Radio-Television and Telecommunications Commission, Volume 2 at paras. 2529-47, from Revised regulatory framework for wholesale services and definition of essential service, supra note 72, online: CRTC <http://crtc.gc.ca/eng/transcripts/2007/tt1010.htm>).
(82.) M. Keith Chert & Barry Nalebuff, "One-Way Essential Complements" Cowles Foundation for Research in Economics (November 2006) at 3-4, online: Cowles Foundation for Research in Economics <http://cowles.econ.yale.edu/P/cd/d15b/d1588.pdf>.
(83.) van Schewick, supra note 12 at 376.
(84.) Transcript of Proceedings Before the Canadian Radio-Television and Telecommunications Commission, Volume 1, from Revised regulatory framework for wholesale services and definition of essential service, supra note 72, online: CRTC <http://www.crtc.gc.ca/eng/transcripts/2007/ttl009.htm>. See e.g. the cross-examination of the Competition Bureau witnesses by counsel for PRIMUS at paras 1814--28. Evidence was submitted suggesting that competition could still fail to discipline high speed internet access providers as a result of factors that limit consumers ability to switch from a cable company to a telephone company or vice-versa, including switching costs (e.g. having to buy new modems), loyalty programs (e.g. Air Miles reward), service bundling and inertia due to the need to change e-mail addresses. However, a Competition Bureau witness testified at para. 1815 that because cable modem internet service and ADSL service are in the same market, any switching costs must be relatively low as to have little or no impact on the operations of competitive forces
(85.) van Schewick, supra note 12 at 376.
(86.) Ibid. at 376-77.
(87.) Others active in the Canadian market are Branchez-Vous, CanadaWebDirectory, Francite Canada, InfoSpace Canada, La Toile, Lycos Canada and a number of other smaller regional search sites.
(88.) comScore, Press Release, "comScore Releases March 2008 U.S. Search Engine Rankings" (15 April 2008), online: comScore <http://www.comscore.com/press/release.asp?press=2185>.
(89.) comScore, Press Release, "More than 10 Billion Videos Viewed Online in the U.S. in February" (16 April 2008), online: comScore <http: //www.comscore.com/press/release.asp?ress=2190>.
(90.) comScore, Press Release, "comScore Releases Top Canadian Web Rankings for March 2008" (28 April 2008), online: comScore <http://www.comscore.com/press/release.asp?press=2200>.
(92.) See Matt Hartley & Grant Robertson, "iTunes Canada receives gift of TV" The Globe and Mail (12 December 2007), online: The Globe and Mail <http://www.theglobeandmail.com/servlet/story/RTG AM. 20071212.wapple 13/BNStory/Technology/? cid=al_gam_nletter_techweekly>.
(93.) Based on information on Apple's website (7 December 2007), online: Apple <http://www.apple.com/itunes>.
(94.) See online: Hulu <http://www.hulu.com/about>.
(95.) In March 2005, Madison River Communications LLC entered into a consent decree with the FCC in response to a complaint that it was blocking a customer's access to Vonage's Vol P service. See U.S., Federal Communications Commission, Madison River Communications, LLC and Affiliated Companies, Order, 20 F.C.C.R. 4295 (2005), online: Federal Communications Commission <http://www.fcc.gov/eb/Orders/2005/DA-05-543A2.html> .
(96.) See e.g. Vonage Submissions, supra notes 54 and 55; Skype Submissions, supra note 45.
(97.) Swanson & Gilder, supra note 58.
(98.) Further discussion of these cases will be found below at III.B.1 and III.B.2.
(99.) See e.g. Lessig & McChesney, supra note 53. The authors remarked that: "The implications of permanently losing network neutrality could not be more serious. The current legislation, backed by companies such as AT&T, Verizon and Comcast, would allow the firms to create different tiers of online service. They would be able to sell access to the express lane to deep-pocketed corporations and relegate everyone else to the digital equivalent of a winding dirt road. Worse still, these gatekeepers would determine who gets premium treatment and who doesn't." Echoing these sentiments, Senator Run Wyden (D-Oregon) stated in a testimony to Congress that "[c]reating a two-tiered system [enhanced QoS vs. standard] could have a chilling effect on small morn and pop businesses that can't afford the priority lane, leaving these smaller businesses no hope of competing against the Wal-Marts of the world.... "Senator Ron Wyden, Press Release, "Wyden Moves to Ensure Fairness of Internet Usage with New Net Neutrality Bill" (2 March 2006), online: United States Senator Ron Wyden <http://wyden.senate.gov/>.
(100.) David D. Clark, "Network Neutrality: Words of Power and 800-Pound Gorillas," online: (2007) 1 International Journal of Communication 701 at 702 <http://ijoc.org/>.
(101.) Wu & Lessig, supra note 29 at 4; see also van Schewick, supra note 12 at 335-36.
(102.) The potential scope of interference of many popular forms of net neutrality regulation has been recognized by notable net neutrality proponents who now suggest ways to lighten the elements of net neutrality regulation. Van Schewick has called for a "light form of behavioral regulation that narrowly targets the behavior identified as problematic and is far less intrusive" focusing on "rules that prevent network operators and ISPs from using their power over the transmission technology to negatively affect competition in complementary markets for applications, content and portals." See van Schewick, supra note 12 at 333-34. As noted above, Lessig recently clarified that Internet access providers should be able to apply access tiering and usage caps to consumers (but not to Internet content and applications providers). In his testimony to the US Senate, Lessig stated that "[t]o oppose access-tiering, however, is not to oppose all tiering, l believe, for example, that consumer-tiering should be encouraged. Network providers need incentives to build better broadband services. Consumer-tiering would provide those incentives. Consumer-tiering, however, should not discriminate among content or application providers." See Lessig, 2006 Senate Testimony, supra note 35 at 2-3.
(103.) Cheng, Bandyopadhyay & Guo, supra note 12.
(104.) Nicholas Economides & Joacim Tag, "Net Neutrality on the Internet: A Two-Sided Market Analysis" (2007) [working paper series] at 23, online: New York University Law and Economics Working Papers <http://lsr.nellco.org/nyu/lewp/>.
(105.) The issue of innovation within and at the edge of the network was thoroughly canvassed during the CRTC's hearings on wholesale service regulation: Revised regulatory framework for wholesale services and definition of essential service, supra note 72; for a specific example of the discussion of innovations within and at the edge of the network in Canada, see Testimony of Mr. Dale Hatfield, Rogers expert witness. Transcript of Proceedings Before the Canadian Radio-Television and Telecommunications Commission, Volume 4 at para. 7940 (12 October 2007), from Revised regulatory framework for wholesale services and definition of essential service, supra note 72, online: CRTC <http://www.crtc.gc.ca/eng/transcripts/2007/tt1012.htm>.
(106.) Communications Monitoring Report, supra note 74 at 181.
(107.) Litan & Singer, "Unintended Consequences", supra note 59.
(108.) Ibid. at 568: "In summary, blended QoS would likely harm end-users of content providers that require enhanced QoS (by reducing the quality of QoS-needy applications), and it would unambiguously harm end-users of content provider [sic] that do not value QoS (by increasing the price of an unnecessary component)." Ibid. at 562-63: Litan and Singer quantified the potential consumer welfare loss in an environment in which net neutrality regulation in the US forces broadband network operators to stop offering enhanced QoS to content and applications providers for online video gaming. They surmise that without enhanced QoS "purchased by content providers like Sony and Blizzard, online garners [consumers] could not experience the game as it was meant to be played." The consumer surplus from enhanced gaming QoS is the amount by which the price charged is exceeded by the consumer's willingness to pay. Online video game subscription is forecasted to reach 28.5 million gamers in 2009, with an average per-gamer monthly revenue of slightly less than $11 per month. Litan and Singer estimate that the surplus associated with consuming online video games in the US will be between $729M and $1.458B in 2009--all of which could be lost if broadband network owners stop offering differentiated enhanced QoS for a fee. Ibid. at 564.
(109.) Hal Singer, "Not neutrality; The concept of 'net neutrality,' if adopted in Canada, would chill needed investment in the Internet" Financial Post (29 March 2007) FP19.
(110.) Richard N. Clarke, "Costs of Neutral/Unmanaged IP Networks" (May 2006) at 19-20, online: NextGen Web <http://www.nextgenweb.org/wp-content/uploads/2007/08/ costs-of-neutral-unmanaged-ipnetworks.pdf>.
(111.) The access cost does not include additional fees that the user may have to pay the producer of the video content. See ibid. at 22.
(112.) Ibid. at 22-23. Of course, such cost estimates depend on a number of variables, including the assumed capacity and traffic volumes, and these estimates could be lower or higher under other sets of assumptions.
(113.) Kahn, supra note 12 at 182. At 187, Kahn calls for "especial alertness to the possibility of anticompetitive denial of access or vertical squeezing of independent suppliers of content."
(114.) Ibid. at 181. Kahn observed that "broadband facilities have to be created by investment, and applications requiring priority transmission impose opportunity costs on others; except as subsidized by government a possibility I do not exclude--those costs must be collected from users--subscribers to broadband services, providers of programming or content, or some combination of the two" and that "[t]he case for treating [Internet access] nevertheless as a public good, deserving of direct governmental subsidy or provision, must rest instead on the proposition, by no means unreasonable, that it provides benefits to the public at large-external to the direct transacters--sufficient to justify public subsidy." (emphasis added) (footnote omitted)
(115.) Kahn argued that if there is concern that the current competition among various broadband platforms is not sufficient to prevent the risks that net neutrality regulation is intended to address, it may be more appropriate for government to directly intervene by accelerating entry by other technological platforms. At ibid at 184, he remarked that "[t]here is clearly room therefore for agreement between proponents and opponents that ... de-regulation be conditioned on sufficient, independent competition from at least a third mode-presumably wireless, assured by freeing up more of the spectrum--while hoping for successful entry also of broadband over the ubiquitous power lines." In conclusion, at ibid at 186, Kahn recommended "... ensuring the availability of at least a third, independent broadband access option presumably wireless whether by application of the antitrust laws to intermodal mergers, opening up additional spectrum, subsidization or direct governmental provision as a necessary protector of both subscribers and providers of content...."
(116.) At ibid. at 186-87, Kahn calls for "an unwillingness to jettison the essential facilities antitrust doctrine-recalling, in particular, that the dominance of incumbent telephone and cable companies in the broadband Internet access market traces back to their original respective monopoly franchises." (footnote omitted)
(117.) Christopher S. Yoo, "Beyond Network Neutrality" (2005) 19 Harv. J.L. & Tech. 1.
(118.) Ibid. at 16.
(120.) See Christopher S. Yoo, "Would Mandating Broadband Network Neutrality Help or Hurt Competition?: A Comment on the End to-End Debate" (2004) 3 Journal on Telecommunications and High Technology Law 23 at 67 where he noted that "regulators can adopt a more humble posture about their ability to distinguish anticompetitive from procompetitive behavior and attempt to resolve the problem by promoting entry by alternative broadband platforms. Once a sufficient number of alternative last-mile providers exist, the danger of anticompetitive effects disappears, as any attempt to use an exclusivity arrangement to harm competition will simply induce consumers to obtain their services from another last-mile provider."
(121.) Competition Bureau, Speech, "The Role of the Competition Act in the Telecom Sector" (30 April 2007), online: <http: //www.competitionbureau.gc.ca/epic/site/cb-bc.nsf/en/02357e.html> [Competition Bureau, Speech].
(123.) See e.g. Brett Frischmann & Spencer Weber Waller, "Essential Facilities, Infrastructure, and Open Access" (2006), online: Social Science Research Network <http://ssrn.com/abstractid=942074> (proposing how the antitrust doctrine of essential facilities can be used to shape the parameters of the net neutrality debate).
(124.) Competition Bureau, Speech, supra note 121.
(125.) The government's decision to set aside spectrum for new entrants, and to mandate renaming and tower sharing, is surprising from a government whose avowed policy is to make the industry rely, to the maximum extent, on market forces. The effect of the policy is to facilitate additional competition at the broadband access level. However, for the purpose of this paper, this policy action further undermines the argument for yet another layer of regulatory guarantee by way of net neutrality regulation.
(126.) Industry Canada, News Release, "Government Opts for More Competition in the Wireless Sector" (28 November 2007), online: Industry Canada <http://www.ic.gc.ca/epic/site/icl.nsf/en/02062e.html>.
(127.) See also Nicholas Economides, "'Net Neutrality,' Non Discrimination and Digital Distribution of Content Through the Internet" (2008) 4 I/S: A Journal of Law and Policy for the Information Society 213 at 214-17.
(128.) U.S., Federal Communications Commission, Appropriate Framework for Broadband Access to the Internet over Wireline Facilities (Washington, D.C.: The Commission, 2005), online: Federal Communications Commission <http://fjallfoss.fcc.gov/edocs_public/attachmatch/FCC-05-150A1.pdf>. This is the culmination of a series of broadband deregulation decisions: U.S., Federal Communications Commission, Triennal Review Order (Washington, D.C.: The Commission, 2003), online: Federal Communications Commission <http://hraunfoss.fcc.gov/edoc_public/attachmatch/FCC-03-36A1.pdf>; U.S., Federal Communications Commission, MDU Reconsideration Order (Washington, D.C. :The Commission, 2004), online: Federal Communications Commission <http://hraunfoss.fcc.gov/edocs_public/attachmatch/FCC04-191A1.pdf>; U.S., Federal Communications Commission, FTTC Reconsideration Order (Washington, D.C.: The Commission, 2004), online: Federal Communications Commission <http://hraunfoss.fcc.gov/edocs_public/attachmatch/FCC-04-248A1.pdf>; U.S., Federal Communications Commission, Section 271 Broandhand Forbearance Order (Washington, D.C. : The Commission, 2004), online: Federal Communications Commission <http://hrauntoss.fcc.gov/edocs_public/attachmatch/FCC-04-254A1.pdf>.
(129.) Supra note 72, online: CRTC <http://www.crtc.gc.ca/eng/archive/2008/dt2008-17.htm>.
(130.) Disposition of funds in the deferral accounts (16 February 2006), Telecom Decision CRTC 2006-9, online: CRTC <http://www. crtc.gc.ca /eng/archive/2006/dt2006-9.htm>.
(131.) See generally Regulatory framework for second price cap period (30 May 2002), Telecom Decision CRTC 2002-34, online: CRTC <http://www.crtc.gc.ca/eng/archive/2002/dt2002-34.htm>; and Implementation of price regulation for Telebec and TELUS Quebec (31 July 2002),Telecom Decision CRTC 2002-43, online: CRTC <http://www.crtc.gc.ca/eng/archive/2002/dt2002-43.htm>, CRTC Telecom Decisions, Public Notices and News Releases (Ottawa: Canadian Government Publishing, 2002) vol. 10:9. The deferral accounts were created by Telecom Decisions CRTC 2002-34 and CRTC 2002-43. The incumbent TSPs were directed to place into those accounts amounts equal to the revenue reductions that would otherwise have resulted from an application of the price cap formula, in order to avoid a negative impact on local competition.
(132.) See Barratt & Shade, supra note 24 at 300-01. Interestingly, Barratt and Shade have also suggested municipal ownership of broadband infrastructure as an option in Canada. Their argument seems to be that municipal ownership of only the conduit allows for transparent and equal treatment of content and applications providers ("A city-owned fibre infrastructure could present the opportunity for true competition--again, separating carriage from service" at 301). Reverting to government ownership of telecommunications--again, networks is highly unlikely in this age, considering the competitiveness of markets. Government ownership of telecommunication networks is indeed a worse option than imposing net neutrality regulation. See McTaggart, "Net Neutrality", supra note 24 at 25, noting that public ownership of broadband networks in this day and age would amount to "a stunning reversal of industrial policy."
(133.) van Schewick, supra note 12 at 376-77.
(134.) Comcast, supra note 8 at 32, n. 242.
(135.) Ibid. at para. 10.
(136.) Because Comcast's method involves sending RST packets to both sides of a TCP connection, which is the same method computers connected via TCP use to communicate with each other, a customer has no way of knowing that it is Comcast (rather than the other peer) that is terminating a connection.
(137.) Comcast, supra note 8 at para. 41.
(138.) U.S., Federal Communications Commission, Policy Statement (Washington, D.C.: The Commission, 2005), online: Federal Communications Commission <http://hraunfoss.fcc.gov/edocs_public/attachmatch/FCC-05-151A1.pdf>.
(139.) Ibid. at para. 4.
(141.) Ibid. at 3, n.15.
(142.) Comcast, supra note 8 at para. 30.
(143.) Ibid. at para. 31.
(144.) Ibid. at 28, n. 217.
(145.) Ibid. at para. 32.
(146.) Ibid. at para. 47.
(147.) Ibid. at 28, n. 221. Cf. Dissenting Statement of Commissioner Robert M. McDowell, ibid. at 61.
(148.) Ibid. at para. 47.
(149.) Ibid. at para. 48.
(151.) Indeed, the FCC copiously referenced the testimonies of the key proponents of the regulatory guarantee approach, in particular Lessig and Wu, without appearing to weigh the competitive outcome of the impugned network management practice. See e.g. ibid at 13, n. 94 and 27 at n. 211 citing several testimonies by Lessig that "competition [is] assured" by imposing net neutrality regulation to maintain the Internet's traditional "open" architecture; ibid. at 25, n. 196 and 28 at n. 219, citing Wu in support of the finding that Comcast's practice impedes consumers Dora running applications of their choice without evaluating the impact of the impediment on competition. See also ibid. at 28, n. 220 citing the comments of Part15.org that "specifically targeting a specific type of IP protocol would not be in keeping with open-access," again without an analysis of the competitive impact of such targeting.
(152.) See Letter from Acanac Inc. to CRTC (3 July 2008) re: Part VII Application by the Canadian Association of Internet Providers Requesting Certain Offers Directing Bell Canada to Cease and Desist from "Throttling" its Wholesale ADSL Access Services at para. 32, online: CRTC <http://www.crtc.gc.ca/PartVII/eng/2008/8622/c51_200805153.hm>; Letter from Primus to CRTC (3 July 2008) re: Part VII Application by the Canadian Association of Internet Providers Requesting Certain Orders Directing Bell Canada to Cease and Desist from "Throttling" its Wholesale ADSL Access Services at paras. 41-42, online: CRTC <http://www.crtc.gc.ca/PartVII/eng/2008/8622/c51_200805153.htm> [Primus Submissions]. See also Comments of Canadian Internet Policy and Public Interest Clinic on behalf of the Campaign for Democratic Media at para. 24 (3 July 2008), online:
<http://www.crtc.gc.ca/PartVII/eng/2008/8622/c51_200805153.htm> [CIPPIC Submission]; and Google Submissions, supra note 46 at para. 23(d).
(153.) Primus Submissions, ibid. at paras. 41-42.
(154.) CIPPIC Submissions, supra note 152 at para. 25.
(155.) Google Submissions, supra note 46 at para. 23(d).
(156.) Bell Canada Answer, supra note 61 at para. 118.
(157.) Canadian Association of Internet Providers' request for interim relief regarding Bell Canada's Practice of "throttling" its wholesale ADSL services (14 May 2008), Telecom Decision CRTC 2008-39, online: CRTC <http://www.crtc.gc.ca/eng/archive/2008/dt2008-39>. While the specific allegations and relief sought in CAIP's application were directed at Bell Canada, several general references were also made to the network management practices of other high speed Internet access providers in CAIP's comments and those of intervening parties. In particular, Rogers Cable, a major high speed cable modem Internet access provider, was alleged to engage in shaping P2P file sharing traffic. In its comments, Rogers described its network management practice as follows:
Like Bell Canada, Rogers' network management equipment uses Deep Packet Inspection (DPI). Unlike Bell, however, Rogers only manages upstream P2P traffic (that is, uploading of P2P traffic from a Rogers high speed Internet subscriher's CPE) and not downstream traffic (that is, downloading of P2P traffic onto a Rogers high speed Internet subscriber's CPE). DPI identifies incoming P2P traffic.... If a packet is identified as being a P2P file sharing protocol packet, it gets routed into a portion or allocation of the network dedicated tn P2P traffic. All other traffic is delivered using an allocation of upstream traffic dedicated to that traffic.... Rogers focuses on managing upstream P2P traffic because that is where the con gestion impacts our network the most. Unlike upstream traffic, a Rogers customer's downstream capacity is not allocated between P2P and other applications. Of course, because some P2P applications (Bit Torrent for example) restrict download speed to the maximum upload speed provided by" the user, a customer's P2P download speed can be limited by the upstream cap but that is a result of the business decision taken by the P2P applications provider.
See Letter from Rogers to CRTC Re: Canadian Association of Internet Providers (CAIP)--Application requesting certain orders directing Bell Canada to cease and desist from throttling its wholesale ADSL Access Services (3 July 2008) at paras. 4-5, online: CRTC
(158.) Bell Canada Answer, supra note 61 at para. 111.
(159.) Ibid. at para. 115.
(160.) Ibid. at para. 114.
(161.) Based on the same state of facts, on May 9, 2008, the Canadian Internet Policy and Public Interest Clinic (CIPPIC) filed an application with the Privacy Commissioner under the Personal Information Protection and Electronic Documents Act (PIPEDA) alleging that Bell Canada is violating consumers' privacy when it uses DPI to identify P2P file sharing traffic. The complaint raises interesting privacy issues which are independent of net neutrality arguments. They will not be addressed in this paper, although it is arguable that the fact that redress could potentially be sought under PIPEDA further weakens the need for a new' net neutrality legislation or regulatory regime. See Letter from CIPPIC to Privacy Commissioner of Canada (9 May 2008) re: Bell Canada/Bell Sympatico Use Deep Packet Inspection: PIPEDA Complaint, online: CIPPIC <http://www.cippic.ca/uploads/Bell-DPI-PIPEDAcomplaint_09May08.pdf>.
(162.) Letter from CAIP to CRTC re: Part VII Application by the Canadian Association of Internet Providers Requesting Certain Orders Directing Bell Canada to Cease and Desist from "Throttling" its Wholesale ADSL Access Services (24 April 2008) at para. 26, online: CRTC
<http://www.crtc.gc.ca/public/partvii/2008/8622/c51_200805153/895702.pdf> [CMP Interim Reply].
(163.) Telecommunications Act, supra note 2.
(164.) Google Submissions, supra note 46.
(165.) Skype Submissions, supra note 45.
(166.) See Google Submissions, supra note 46 at para. 2 where Google declared that "providers of broadband internet access services, including Bell, should be prohibited from throttling lawful applications." (footnote omitted)
(167.) Skype Submissions, supra note 45 at para. 5.
(168.) Google Submissions, supra note 46 at para. 7.
(169.) Ibid. at para. 17(b).
(170.) Ibid. at para. 17(c).
(171.) CAIP v. Bell Canada, supra note 9.
(172.) PN 2008-19, supra note 10 at para. 3.
(173.) Ibid. at paras. 9(4) & 9(5).
(174.) (12 Ma), 2005) Telecom Decision CRTC 2005-28, online: CRTC
<http://www.crtc.gc.ca/eng/archive/2005/dt2005-28.htm>, Telecom Decisions, Public Notices, Circulars and News Releases (Ottawa: Publishing and Depository Services, 2005) vol 13:3 [Regulatory framework].
(175.) Terms and rates approved for large cable carriers high speed access service (21 August 2000), Telecom Order CRTC 2000-789, online: CRTC <http://www.crtc.gc.ca/eng/archive/2000/02000-789.htm>.
(176.) Digital subscriber line services providers' access approved for unbundled loops and co-location (27 October 2000), Telecom Order CRTC 2000-983, online: CRTC <http://www.crtc.gc.ca/eng/archive/2000/O2000-983.htm>.
(177.) Regulatory framework, supra note 174 at para. 417.
(178.) Ibid. at paras. 429, 446.
(179.) Ibid. at, paras. 458-65.
(180.) Ibid. at paras. 452-57.
(181.) Ibid. at paras. 466-73.
(182.) Ibid. at para. 475.
(183.) Ibid. at para. 478.
(184.) Ibid. at para. 480.
(185.) Reconsideration of Regulatory framework for voice communication services using Internet Protocol (1 September 2006), Telecom Decision CRTC 2002-13 at paras. 132-34, online: CRTC <http://www.crtc.gc.ca/eng/archive/2006/ dt2006-53.htm>.
(186.) Part VII application by Superior Wireless Inc. against TBayTel alleging unjust discrimination (25 May 2006) at paras. 29-30,Telecom Decision 2006-33, online: CRTC <http://crtc.gc.ca/eng/archive/2006/dt2006-33.n>.
(187.) Application by Microcell regarding alleged contraventions of section 27(2) of the Telecommunications Act by Rogers Wireless and Bell Mobility, (28 April 2003) at para. 58, Telecom Decision CRTC 2003 26, online: CRTC <http:// www.crtc.gc.ca/eng/archive/2003/dt2003-26.htm>, CRTC Telecom Decisions, Public Notices and News Releases (Ottawa: Canadian Government Publishing, 2003) vol. 11:3.
(188.) (8 October 2004),Telecom Decision CRTC 2004-66, online: CRTC <http: //www.crtc.gc.ca/eng/archive/2004/dt2004-66.htm>, Telecom Decisions, Public Notices, Circulars and News Releases (Ottawa: Publishing and Depository Services, 2004) vol. 12 : 14.
(189.) Ibid. at para. 10.
(191.) Ibid. at paras 10, 12.
(192.) Ibid. at para. 31.
(193.) Ibid. at para. 33.
(194.) CAIP v. Bell Canada, .supra note 9 at paras. 45-46.
(195.) Comcast supra note 8 at 49.
(196.) See Letter from CRTC to Edward Antecol regarding Richard Warman Application (24 August 2006) online:
<http://www.crtc.gc.ca/eng/archive/2006/lt060824.htm>, where the CRTC noted that:
The Commission notes that section 36 of the Act would not allow it to require Canadian carriers to block the web sites; rather, under section 36 of the Act, the Commission has the power to permit Canadian carriers to control the content or influence the meaning or purpose of telecommunications it carries for the public. The scope of this power has yet to be explored.
(197.) PN 2008-19, .supra note 10 at para. 9.
(198.) See Regulation of broadcasting distribution undertakings that provide non-programming services (9 July 1998), Telecom Derision CRTC 96-1, online: CRTC
<http://www.crtc.gc.ca/eng/archive/1996/DT96-1.htm>, CRTC Telecom Decisions, Letter Decisions and Public Notices(Ottawa: CCG--Publishing, 1996) vol. 3:48 (Using section 36, the CRTC has determined that broadcast carriers that offer Internet services on their broadband facilities are by definition controlling the content of that particular telecommunications (the Internet service)); see Regulation under the Telecommunications Act certain Telecommunications Services offered by "Broadcast Carriers" (9 July 1998), Telecom Decision CRTC 98-9, online: CRTC <http://www.crtc.gc.ca/eng/archive/1998/DT98-9.htm>, CRTC Telecom Decisions, Public Notices and News Releases (Ottawa: CCG-Publishing, 1998) vol. 6:7 (The CRTC has then permitted broadcast carriers to engage in content services by granting approval to their Internet and non-programming, services, and has approved the provision of content and applications by, telephone companies pursuant to section 36); see Review of Regulatory Framework (16 September 1994),Telecom Decision CRTC 94-19, online: CRTC <http://www.crtc.gc.ca/eng/archive/1994/DT94-19.HTM>, CRTC Telecom Decisions Letter Decisions, and Public Notices (Ottawa: CCG-Publishing, 1994) vol. 2:29; and see Stentor--Request for Approval under Section 36 of the Telecommunications Act (31 March 1999), Telecom Decision CRTC 99-4, online: CRTC
<http://www.crtc.gc.ca/eng/archive/ENG/Decisions/1999/DT99-4.HTM> CRTC Telecom Decisions, Public Notices and News Releases (Ottawa: CCG--Publishing, 1999) vol. 6: 38.
(199.) See .supra note 187.
(200.) Comcast, supra note 8 at para. 5. (footnotes omitted)
(201.) CAIP Interim Reply, supra note 162 at paras. 26-28.
(202.) Comcast, supra note 8 at 65 (Dissenting Statement of Commissioner Robert M. McDowell).
(203.) Bell Canada Answer, supra note 61 at para. 164.
(204.) Comcast, supra note 8 at paras. 52-53.
(205.) Ibid. at para. 52.
(206.) See generally Skype Submissions, supra note 45 and Google Submissions, supra note 46. As well, in Canada on May 28, 2008, a Private Member's Bill (Bill C-552) was tabled in the House of Commons to amend the Telecommunications Act:
To prohibit network operators from engaging in "network management practices that favour, degrade or prioritize any content, application or service transmitted over a broadband network based on its source, ownership or destination," subject to certain exceptions, and also prohibits network operators from preventing" a user from attaching any device to their network" and requires network operators to make information about the user's access to the Internet available to the user. Bill C-552, supra note 35 at cls. 36.1 (1), (3) and (4).
Bill C-552 would add a new section 36. I, titled "Network Management," to the Act:
36.1 (1) Network operators shall not engage in network management practices that favour, degrade or prioritize any content, application or service transmitted over a broadband network based on its source, ownership or destination. Bill C-552, supra note 35.
The proposed legislation would permit several exceptions, including the right of ISPs to "manage the flow of network traffic in a reasonable manner in order to relieve congestion" and to "offer directly to each user service at different prices based on defined levels of bandwidth or the actual quantity of data flow over a user's connection." Bill (:552, supra note 35 at cls. 36.1 (2)(a) and (d). Also, on June 2, 2008, David McGuinty, a Liberal MP, introduced another Private Member's Bill in the House of Commons--Bill C-555 which requested the government to direct, among other things, "an assessment of network management practices that favour, degrade or prioritize any packet transmitted over a broadband network based on source, ownership or destination." Bill C-555, supra note 63 at cl. 3(c).
(207.) CAIP v. Bell Canada, supra note 9 at para. 74.
(208.) See van Schewick, supra note 12.
(209.) Haas, supra note 12 at 1628 35.
(210.) Ibid. at 1630.
(211.) Comcast, supra note 8 at para. 36.
(212.) Ibid, at para. 53.
(214.) A review of the websites of major Canadian broadband network operators reveal varying degrees of disclosures regarding traffic management, bandwidth caps, contractual limits etc. See e.g. "Acceptable Use Policy", online: Rogers <https://your.rogers.com/about/legaldisclaimer/Unified-AUPEng.pdf>;" Acceptable Use Policy", online: Shaw
<http://www.shaw.ca/en-ca/AboutShaw/TermsofUse/ AcceptableUsePolicyInternet.htm#q8>;"Bell Internet Service Agreement" (updated 11 August 2008), online: Bell Sympatico
<http://service.sympatico.ca/ index.cfm?method-content.view&category_id=550&contentid=11013>; "Policies & Terms," online: Telus <http://www.mytelus.com/internet/policies/display.do>.
(215.) Competition Bureau Canada, News Release, "Competition Bureau Calling on Prepaid Phone Card Providers to Disclose More Information" (6 December 2007), online: Competition Bureau Canada <http://wwvv.competitionbureau.gc.ca/epic/site/cb-bc.nsf/en/0048le.html>.
ALEXANDER J. ADEYINKA *
* Alexander J. Adeyinka is a Toronto-based telecommunications regulatory lawyer. He has done extensive business and academic work in Canadian and International telecommunications law, policy and regulation. All opinions expressed in this paper are strictly those of the author and are not attributable in any manner or form to his employers or their affiliated businesses.
|Printer friendly Cite/link Email Feedback|
|Author:||Adeyinka, Alexander J.|
|Publication:||Ottawa Law Review|
|Date:||Dec 22, 2008|
|Next Article:||La norme esclavagiste, entre pratique coutumiere et norme etatique : les esclaves panis et leur statut juridique au Canada (XVIIe-XVIIIe s.).|