III-1.6: LA NEUTRALITÉ D’INTERNET [Net Neutrality], by Nicolas Curien and Winston Maxwell

 

1. The authors define Net Neutrality as “equal access by all Internet users to all of the Web’s content, services, and applications.” This principle must coexist with the necessary constraints related to bandwidth management, financing innovation, privacy and intellectual property protection, etc.

 

2.  Neutrality must therefore be ‘hybridized’ with reality. Therefore, the authors show a preference for ‘the greatest possible amount of neutrality’, which they call ‘quasi-neutrality’. Forbidding all forms of discrimination is unrealistic, and therefore neutrality must become a sort of second-tier optimum. Nonetheless, quasi-neutrality does not mean that all forms of discrimination should be allowed. The authors therefore plead for efficient discrimination, which should be allowed because differences in treatment are justified by the pursuit of technological and economic efficiency, unlike anticompetitive discrimination, which should remain forbidden. Even efficient discrimination must be transparent in order to prevent it from hiding anticompetitive discrimination. Therefore, users must be informed of any bandwidth or access restrictions that might be linked with their subscription.

 

3. Another issue is that of cost sharing. Indeed, the considerable expense of building a platform that other economic agents profit from should not be borne by only one category of agents. The Internet is like a city or an ecosystem requiring co-regulation between public and private agents. Regulation must always remain modest, because competition law is often sufficient to curb the excesses of agents in very strong competition with one another.

 

4. In order to explain these issues, and to allow readers to form their own opinion on the current debate, the authors set forth the ‘ingredients of the debate’. They emphasize the three aspects of neutrality: the financial aspect (i.e. who should share revenue and what the financial counterpart between ISPs and content providers should be); the technological aspect, on bandwidth management practices; and the deontological aspect, on ISPs’ non-discrimination between the way they treat different content.

 

5. Furthermore, neutrality has a downstream aspect (the ways consumers access content), and an upstream aspect linked to the repartition of costs and revenue. The ways these issues will be resolved will depend upon the behavior of ISPs, search engines, browsers, and computer manufacturers. The use and regulation of their economic power will be decisive.

 

6. The authors colorfully and limpidly point out that this debate is between the ‘technology experts’ (meaning ISPs), who are most concerned with competition and cost- and revenue-sharing; the ‘architects’ (meaning the programmers who created Internet), who refuse all forms of regulation out of hand, and who do not believe there is any commonality between telecommunications and the Internet: indeed, they believe that the rigid methods applied in telecommunications hinder innovation. They are, by definition, in favor of a broad principle of neutrality. The third category are ‘citizens’, who fight for freedom of speech and protection of their privacy, and they are wary of any corporation or government, claiming their right to access content for free, and even the right to download intellectual property-protected works for free. They become hostile towards the architects when the latter form an alliance with telecommunications corporations (contract between Verizon and Google in 2010).

 

7. The issue of Net Neutrality began with a few court decisions in the year 2000, which established the principle that ISPs cannot discriminate between the content users access. Then, in 2005, the Federal Communications Commission laid out the Internet’s ‘four freedoms’ with its Madison River case. These freedoms are: the freedom for internet users to access all legal content, to use all legal applications and services, to connect all equipment to the network, and to have a choice between multiple, truly competitive ISPs. However, the legal ramifications of this decision have never been clearly understood, and it is unknown what mandatory force must be given to the regulator’s declarations. This proves that the principle of Net Neutrality has never been legally established.

 

8. In 2007, the FCC handed down a decision concerning Comcast, an ISP and cable operator that prevented some of its subscribers from using a peer-to-peer file-sharing service. In this case, the FCC demanded that the ISP justify this discrimination. Unable to do so, the company was fined. The regulator used this case to establish a test of proportionality in order to determine whether or not the bandwidth management techniques used by ISPs are reasonable. In other words, the discriminatory practice must have a legitimate goal (such as preventing network saturation at peak hours, or protecting the network against viruses or illegal content), and the discriminatory practice must be proportional to the goal pursued.

 

9. But, this raises the question as to whether or not the regulator is legitimate to intervene, and its decision was overturned by the Appeals Court in 2010 on the grounds that the FCC’s legislative mandate only concerns telecommunications, which has nothing to do with Net Neutrality. So, in the current state of affairs, the industry is trying to create a code of good practices. We observe that even if there is not a legal void, there is at least a great amount of legal uncertainty in this area. Indeed, no measures have been taken concerning mobile networks and neutrality (unless we consider that the government might associate the requirement for respecting neutrality when it assigns frequencies: thereby, operators could pay more for a frequency that was not subject to the neutrality requirement, because they would be able to manage bandwidth as they see fit). The FCC reacts very quickly to operators’ behavior, and therefore wishes to adopt mandatory rules on the transparency of bandwidth management and a strict non-discrimination policy, even though the operators themselves believe that only anticompetitive forms of discrimination should be punished, and argue that the agency is not legitimate to make rules in this area.
It is true that the contract signed between Verizon and Google in January 2010 could be seen as a setback for neutrality, or could be seen as an affirmation that net neutrality can be dealt with through self-regulation.

 

 

10. The most important question to be resolved is whether or not mandatory Net Neutrality will dissuade ISPs from investing in infrastructure. The other question is whether of not competition between ISPs will be enough to create innovation, in which case self-regulation must be preferred to regulatory intervention.

 

11. The authors point out that although Net Neutrality has a hugely economic aspect, in that it implies new forms of revenue sharing, it also directly concerns fundamental citizens’ liberties, such as the freedom of speech and the right to privacy. Regulatory authorities, both American and European, have said that the right to communication means that end users must be able to freely access information on the Internet.

 

12. Neutrality’s most fervent supporters conclude that all forms of discrimination must be forbidden. ISPs retort, and regulators agree, that networks have to be efficiently managed, and only legal content may be freely accessed. In the United States, this is seen as self-evident, but in France, the debate over the HADOPI (illegal downloading watchdog authority) showed that it might be acceptable to illegally download copyrighted works. This would mean that ‘neutrality’ would be a ‘new sort of basic right’. The authors point out that this is evidently not the case, because what is illegal in the real world is also illegal in the virtual one. Similarly, the restriction placed on certain content does not imperil freedom of speech, because this is not an absolute freedom, and judges have blocked access to websites that, for example, preach racial hatred.

 

13. When the HADOPI law was voted in 2009, the French Constitutional Council tried to strike a balance between the protection of copyright and the right to free communication by using the principle of proportionality. It ruled that allowing a simple agency to restrict Internet access was disproportionate. Similarly, in 2010, an American Court of Appeals ruled that freedom of speech is protected by the Constitution, but that this freedom does not authorize violating copyright.

 

14. The right to privacy can be jeopardized by the communications tool of unprecedented scale that is the Internet. Laws have begun to be voted in various countries in order to limit the time that personal data can be stored by operators. Furthermore, certain methods used to filter website content can constitute a breach of privacy, which is why, in order for such filters to be legitimate, there has to be a public interest, and the principle of proportionality has to be respected. If the filtering is the result of a contract between Internet users and their ISPs, the principle of contractual liberty allows this greater breach of privacy, so long as the protected person has expressed his free and enlightened agreement.

 

15. Privacy watchdogs (the CNIL in France) intervene in order to define these rules, especially in situations where data are stored within centralized networks. The danger is when information on websites visited (for targeted advertising purposes) is stored by ISPs themselves, and not just by a particular website.[1]

 

16. The authors then point out that Europe began to discuss the question of Net Neutrality long after the United States. The European directives in the 2009 ‘Telecoms Package’ simply obliged operators to inform subscribers on bandwidth management practices and provided regulators with the possibility to set mandatory minimum standards for the provision of Internet service. Member States are therefore quite free to make their own rules in this area, despite the fact that competition-oriented European Community legislation stated in 2002 that in the absence of a manifest need for ex ante regulation, the system must be based on ex post competition law. But, the authors believe that while ‘asymmetrical regulation’ exists in order to build a competitive market, there also exists ‘symmetrical regulation’, intended to pursue other objectives, such as consumer protection. Asymmetrical regulation deals with dominant market agents, while symmetrical regulation, unconcerned with market strength, addresses all agents equally.

 

17. Therefore, wholesale markets, such as the wholesale market for landline telephony, must still be regulated asymmetrically. Similarly, the historical telecommunications provider is still often in a dominant position on the market for DSL high-speed Internet access, and therefore must be regulated asymmetrically. Similarly, the incumbent will be the only provider for access to end-clients by new competitors, at least until new fiber optic local loops compete with the traditional copper-wire local loop.

 

But, the situation is complex, for technological innovations might lead to competition over the technology adopted by each ISP, and it is unknown whether or not regulators ought to intervene in these choices. Similarly, the access price for the local loop is fixed in order to provide incentive for investment or for other reasons. Asymmetrical regulation is practiced in this way in Europe, but not in the United States, where the FCC does not have such powers, and where, for example, mobile operators are not obliged to provide network access to their competitors.

 

18. The 2009 European Directives stated that the transmission of content must not distort competition, which amounts to distinguishing between anticompetitive discrimination (forbidden) and technological discrimination (authorized[2]). The legislation requires regulators to defend citizens’ interest in accessing and sharing information, and in using the applications of their choice. The authors believe that regulators such as the ARCEP in France are legitimate to resolve disputes on this subject between an ISP and a company, even when they are not contractually bound to one another, so long as they are indirectly linked by an interconnection or an access point.

 

19. Specific European legislation does not answer the question of whether an ISP can or cannot practice upstream discrimination between various content providers in order to benefit some and disadvantage others (i.e. exclusivity clauses). However, competition law, which is naturally polyvalent, could deal with the question through abuse of dominance and anticompetitive agreements.

 

20. The authors believe that it is difficult to prove abuse of dominance, because it is hard to identify the relevant market on which the company is supposed to have a dominant position. This is because it is exceedingly difficult to calculate each ISP’s share in the market for content transmission. These probative difficulties explain why Competition Law mainly manifests itself through antitrust suits. This might be applied to a contract between an ISP and a content provider under which the ISP will exclusively give priority access to certain content providers or search engines. Exclusivity is not forbidden per se, because it often stimulates innovation, and exclusivity is common in this industry in order to make investments in new services profitable, such as the 3D video developed by one of Google’s affiliates. Competition authorities perform analysis on a case-by-case basis, and exclusivity can be justified if it remains strictly proportional to what is necessary for innovation.

 

21. This very instructive book’s conclusion is in the form of questions and answers. To the question ‘what is neutrality and what are its limits?’, the authors reply that as much as technologically possible, net neutrality must fulfill the double standard of non-interference and equivalency. As for the question on the issues surrounding neutrality, the authors point out that these issues are both economic (value sharing) and societal, linked to the freedom of speech and innovation. They then insist again upon the difference between efficient, technological discrimination, which must be authorized under the principle of ‘quasi-neutrality’, and anticompetitive discrimination, which must be punished. The authors posit that in a free-market economic model, competition law comes first, and regulation, whether symmetrical or asymmetrical, must come second only when necessary. Regulators intervene ex ante, but also ex post to resolve disputes. Net Neutrality is not only an issue to be decided by infrastructure regulators (in France, the ARCEP), but also by content regulators, since they are linked with ISPs in co-regulatory structures.

 

 

 

 

---