This article examines the relationship between zero-rating and net neutrality in light of the latest case law of the Court of Justice of the European Union, namely the Vodafone case (Case C-854/19). The aim is to determine whether the UK should follow the EU’s example in banning zero-rating practices as seen in Vodafone. It is argued that this question should be answered in the affirmative. Specifically, three key rationales are advanced to support the current ban on zero-rating pursuant to the Vodafone decision. First, it is submitted that this is necessary for the protection of individual freedoms and rights. Second, it is also necessary for the protection of the ‘informational ecosystem’. Third, the economic development of the Internet requires a neutral environment and hence a ban on zero-rating is needed. Therefore, although it is unclear how and if the rules on zero-rating will change now that the UK is not a member of the EU, a normative analysis illustrates that there are important reasons not to deviate from the European framework.
Introduction
Zero-rating is defined as the process by which an Internet Service Provider (‘ISP’) ‘applies a price of zero to the data traffic associated with a specific application or a class of applications’ and as such, ‘the data does not count towards any data cap in place’.1 Zero-rating practices have sparked considerable debate amongst academics, lawyers, and governments. It is important to note from the outset that the debate on zero-rating is intertwined with the concept of net neutrality. ‘Net neutrality is the principle according to which data packets on the internet should be moved impartially’.2 As Maniadaki notes, net neutrality is about ‘the principle of equal treatment’.3 Therefore, zero-rating, by offering some services for free and others only upon payment, creates an imbalance whereby some content providers are favoured over others. Such an imbalance is a threat to net neutrality, which is a concept developed to prevent differential treatment, or in other words, discrimination.4
In light of the recent Vodafone5 decision of the Court of Justice of the European Union (‘the Court’ or ‘the ECJ’) that led to a ban on zero-rating throughout the European Union, this article first draws an overview of the latest case law of the Court, before proceeding to a normative analysis as to why the UK should not change the applicable rules (i.e. the European framework) post-Brexit. It is submitted that there are three key rationales that support the current ban on zero-rating and its implementation in the UK as well. These are: first, the protection of individual rights; second, the protection of the ‘informational ecosystem’;6 and third, the preservation of the economic development of the Internet.
Overview: The Latest Case Law of the Court of Justice
The European legal framework on zero-rating has developed incrementally, starting with the 2015 Open Internet Access Regulation (‘the Regulation’).7 Prior to that, there was no set of comprehensive rules for internet access.Instead, other areas of law would regulate internet access, such as consumer law and competition law. The Regulation is important because it encouraged openness, plurality of content and non-discrimination, and also introduced the principle of net neutrality. Specifically, Article 3(1) promotes open standards, and Articles 3(2) and 3(3) ensure that discriminatory traffic management (such as throttling8) is prohibited. These are followed by the exception of ‘reasonable traffic management’.
Since 2015, the applicability of the Regulation on zero-rating has been determined through case law of the Court as well as the relevant guidelines issued by Ofcom (the government-approved regulatory and competition authority for broadcasting, telecommunications, and postal industries of the UK).9 There have been two central cases on zero-rating practices. The first preliminary ruling addressing the issue was the Telenor judgment.10 This involved app-specific zero-rating applicable beyond the data cap. The Court decided that the carrier was not allowed to block or slow down traffic because it was not zero-rated; rather, according to the Regulation, all traffic must be treated equally. Thus, the Court adopted a broad interpretation of the net neutrality principle in the Regulation. Yet, this did not bring an end to zero-rating. While it was made clear that it is against EU law to block or slow down traffic-related applications other than zero-rated ones when the data cap is reached, the judgment did not prevent ISPs from offering zero-rated applications if they too are blocked upon reaching the data cap.11 Thus, whilst the Regulation applied, the judgment was inevitably vague as to the ‘type’ of zero-rating banned because zero-rating practices were not mentioned specifically in the Regulation.
This changed with the Vodafone case12, after which the question of whether zero-rating occurs within or beyond the data cap practically became irrelevant. The decision has been described as a ‘bombshell judgment’13, where the Court ‘plunges [a] dagger’14 into zero-rating practices. Behind this sensationalist language lies a simple truth: the Court determined that zero-rating is incompatible with EU law, namely the Open Internet Access Regulation.15 The reasoning of the Court was based on the idea that zero-rating inherently represents a commercial incentive. As such, it cannot satisfy the general obligation in Article 3(3) of the Net Neutrality Regulation, which stipulates that ‘providers of internet access services shall treat all traffic equally, when providing internet access services, without discrimination, restriction or interference’.16 The Court drew a much-needed distinction between commercial considerations and technical considerations. This distinction hinges on the differences between considerations with which an ISP must necessarily engage and considerations that are avoidable (in the sense that the only purpose they serve is commercial in a way, meaning that it is profit-oriented). This is crucial because it also relates to a discussion on what ‘normal’ internet access entails, in contrast with discriminatory commercial practices which arguably do not fall within such a ‘normal’ access. Furthermore, the judgment made clear that the incompatibility of zero-rating with the Regulation is not affected by limitations such as roaming.17 Therefore, it is evident that the Court adopted a rather stringent view, according to which the incompatibility of zero-rating with EU law remains even when there are limitations on zero-rating (e.g. roaming or tethering).18 As will be shown below, such a perspective largely corresponds with the argument advanced in this contribution, since its key ambition is to demonstrate that a ban on zero-rating is more than beneficial – it is required.
The conclusion reached from this brief overview of the latest case law of the ECJ is that most zero-rating practices are banned following Vodafone. Even though this is a huge step, there is one remaining question pertaining to whether zero-rating is allowed in relation to a specific class of apps. The judgment left this unanswered. For example, at the moment, Vodafone Voxi offers a package of ‘Endless Video’, including apps such as Netflix, TikTok and Youtube.19 It remains to be seen whether this type of zero-rating will also be banned. It is likely that it will, since it creates an economic incentive to use certain applications over other ones.
Reform in the UK: Why the UK Should Follow the EU’s Example
The Vodafone judgment sets an important precedent against zero-rating throughout the EU and binds regulatory bodies in their assessments.20 Yet —at least in principle— the judgment does not bind the UK. Therefore, the key question is whether the UK, and more specifically Ofcom, will follow the EU’s approach in expanding the ban on zero-rating occurring within, not merely beyond, the data cap.
The UK has fully implemented the Regulation through the Open Internet Access (EU Regulation) Regulations 2016.21 Thus, the framework remains the same. The remaining question pertains to the interepretation provided by the courts that differs between the EU and the UK. In practice, it is far from clear how and if the rules on zero-rating will change now that the UK is officially not a member of the EU. Nevertheless, it is important to examine whether the UK should follow the EU’s example in banning zero-rating practices. It is submitted that there are three key rationales justifying a ban on zero-rating that push for this to be implemented in the UK as well. These are the following: a) protection of individual rights, b) protection of the quality of the internet, and c) protection of innovation in the interent. All hinge on a common aim: maintaining net neutrality.
A. Protection of the Individual
The central rationale supporting the Court’s decision is that zero-rating has detrimental effects on individual freedom, because ISPs are essentially allowed to make choices on behalf of the user. This directly links to the fact that ISPs are private, unaccountable bodies22 that are solely focused on maximising their profits and are thus indifferent to the protection of rights. Such incentives are what the Court in Vodafone interpreted as inherently commercial considerations, which meant that zero-rating cannot fall under the ‘reasonable traffic management’ exception. This interpretation is justified and in fact, necessary. Relying on ISPs’ willingness to behave in a fair and legitimate manner in the name of the general good would be quite naïve given that they are driven by their own financial ambitions.23
Yet, in the contemporary age, where the Internet is integral to the everyday life of all modern individuals, it is unfair not to treat the Internet as a public, open and free good, to which everyone has a right of access. This element of unfairness was pointed out by Audibert and Murray. They suggest that such a right encapsulates two elements: first, ensuring that nobody is excluded from the functions that the Internet serves and second, ensuring that the consumption of one does not reduce the consumption of another.24 Practices such as zero-rating, which are inherently discriminatory, pose a threat to this right of access to a neutral internet. This is because some services or apps get ‘preferential access’ from consumers.25 While the change is only in the price, the fact that most consumers’ choices are largely affected by that factor means that their own freedom to choose between services has been skewed.26
Hence, to protect this right effectively, appropriate regulatory mechanisms must be put in place against a strict legal framework. Some have advanced similar arguments on the basis that such rules would be in the consumers’ interests.27 Yet, it is argued that it is preferable to pay attention to their rights instead. Consumers’ interests are not easy to discern due to the fact that the role the internet plays in consumers’ lives, and thus their interests in it vary from culture to culture and depend on the quality of the Internet in each country. Reliance on consumers’ interests should occur once more holistic research has been undertaken.28 On the contrary, individuals’ rights are enshrined in fundamental conceptions of democracy. Specifically, rights such as the right to privacy and the right to freedom of expression are essential if the internet is to function in a democratic way. Net neutrality is the way to uphold ‘participative democracy’ because it protects these aforementioned rights.29 Without net neutrality, ISPs are able to discriminate and censor, effectively going against these rights. This is because inidviduals who do not agree with exceeding their data cap by viewing the non-zero rated content, are stuck into a ‘walled garden’ in which the only content available is the content decided by ISPs.30 How then, can such content be described as the product of a democratic process? Ultimately, the goal is to make access to the Internet open, just, and equal. Without sufficient protection of rights, this will arguably never be achieved.
B. Quality of the Internet
Interestingly, this discussion can move beyond protecting individual rights. The protection of the quality of the Internet itself must be considered as another compelling reason for preserving net neutrality by banning zero-rating practices. This argument was one that the court missed in its judgment. Rochel, largely influenced by Floridi’s work on information ethics, argues that ensuring diversity is important in terms of the informational ‘ecosystem’s quality’.31 This requires a ‘global governance’ understanding of the Internet, whereby the Internet is understood as something so valuable that needs protection in itself. A useful illustration is provided by way of Recital 11 of the Open Internet Regulation, which introduces the idea of innovation and states that protection of net neutrality is important in terms of maintaining and promoting innovation in general.32 In the context of the Regulation, innovation is not clearly defined, but is likely to be understood as economic innovation, which will be discussed below. Yet, in this section, it is suggested that innovation may be something larger than that. Innovation is conceptualised not only as promoting commercial incentives for investment and growth, but also as promoting diversity which is enriching for the internet per se.33
The issue is that with zero-rating, certain applications are incentivised, while others are discouraged.34 This is harmful because the diversity of the informational sphere is eroded. In other words, the Internet should be viewed as part of a broader ecosystem in which information is exchanged. The quality of such an ecosystem needs protection, because as Turilli, Vaccaro and Taddeo argue ‘when information flow is degraded, blocked, or of a quality no longer compatible with environmental requirements, portions of the Internet cannot flourish’.35 Under such an understanding, zero-rating is clearly harmful to the Internet due to its inherent capacity to degrade the information flow by implicitly creating preferences of some ‘portions of the Internet’ over others.
C. Economic Development and Innovation
Another argument that can be advanced against zero-rating is economic in nature. While diversity is arguably, as discussed above, needed for the Internet to flourish in a more abstract or philosophical sense, it is also needed for economic development. Economic development is hindered through practices such as zero-rating because they have a negative effect on competition between applications and ISPs. Zero-rating works because of agreements between content providers and ISPs whereby content providers pay ISPs to give their content to consumers without cost. This leads to a situation where a few powerful actors control the network, with the rest left in a disadvantaged position. Thus, zero-rating acts as a deterrent to economic innovation and the development of the Internet.36 The Vodafone decision paves the way for such regulation to take place because, after all, it is the neutrality of the Internet that creates incentives to invest as it fosters competition between applications and diminishes monopoly power.37 Of course, it is not only the neutrality of the Internet that is a source of investment incentives. Indeed, its large consumer base and the potential to expand out of geographical borders are also crucial. However, many remain unconvinced that regulation can be advantageous to economic innovation.38 In fact, Hahn and Wallsten argue that regulation reduces investment incentives.39 While theyargue that competition law and antitrust enforcement are adequate means to encourage economic innovation, their argument is based on a false presumption that consumers have the ability to recognise and choose less discriminatory ISPs. This is a rather idealised version of consumers’ abilities and general understandings.40
Perhaps Hahn and Wallsten’s ideas stem from the works of thinkers such as Barlow, who contended that the internet is a phenomenon of global character which cannot be placed within the bounds of nation-state authority.41 Yet, it seems that Barlow’s interpretation has been misplaced over time. Regulation will not necessarily do away with this global character. It will simply need to be adapted to a global framework, where national and global are complemented by each other. Such a framework could potentially identify certain general principles to be followed globally, serving as a rights-based model of regulation corresponding to Barlow’s ‘global character’, whilst at the same time leaving other issues to the discretion of each nation. UK-based regulation could be part of that.
Interestingly, Tim Wu, who coined the term ‘network neutrality’ itself was also biased against regulation. This is because he believed that the market of the wireless industry cannot flourish with such constraints.42 However, in light of the power the Internet has grown to assume, it seems anachronistic to keep such views alive.43 Government intervention has an important role to play in terms of regulation because without it, an unregulated Internet would grow to become a dangerous platform with the potential to harm individuals instead of benefitting them.
While the arguments above are credible, one possible problem arises in applying them to the UK context. It could be argued that zero-rating in the UK calls for a different solution than the one given in the EU. This is because the European setup amplifies the need for competition,44 whereas in the UK, the need for competition between internal actors is still important, but not as much as in the EU. That is, assuming that the EU is more than likely to have at least double the number of ISPs as the UK. However, the English competition law system is mainly based on the EU’s system, so while perhaps competition is less intense in the UK than in the EU, it is not much less important. Therefore, considering all the above factors, from a normative lens, the UK should adopt the Court’s measure in banning zero-rating practices.
The Missing Piece from the Current EU framework
This article intended not only to applaud the Court’s latest decision on zero-rating, but also to point out the important rationales underpinning its approach. Nevertheless, it does not follow that the current framework is flawless. There is one significant issue with how the rules have unfolded, stemming from the fact that zero-rating was not even mentioned in the Regulation. This led to two problems: Firstly, the guidelines for the Regulation in terms of zero-rating are particularly complicated and lengthy,45 and hence hard to understand. Secondly and subsequently, the Court was forced to intervene in order to fill the gaps. While this is not necessarily problematic given that the Court’s intervention was beneficial in the end, it could be argued that for the UK, the problem of zero-rating could be resolved once and for all through statutory reform, rather than through judicial means. By enacting specific legislation on zero-rating, the rules would be clearer and easier to understand by both ISPs and consumers. However, it should be noted that this contribution does not intend to discuss the particularities and challenges a legislative motion to regulate this matter would have. Instead, its focus is on the broader question of zero-rating.
Conclusion
As discussed in this article, zero-rating is an intricate practice to regulate, mainly because it is particularly difficult to agree on how such regulation should take place. The fact that there are various interests at stake should not be underestimated. Zero-rating, closely connected to net-neutrality, involves many different actors and requires one to consider various issues such as market competition, rights discourse and social concerns relating to equality.46
The conclusion of the present analysis can be summed up in the idea that neutrality is needed for justice in both the society we live in and the informational society we access. Furthermore, neutrality fosters, rather than impedes, economic flourishing due to the levelling of the field exercise it performs. Thus, as the Court explicitly decided in Vodafone, zero-rating practices should be strictly regulated. The only question in the development of the current framework is whether such a decision should have been left at the judges’ discretion, instead of being dealt with at the supranational level. One thing is clear: when applying this decision to the UK, specific legislation suitable to the domestic context should be a key consideration.
[1] Andrew Murray, Information Technology Law: The Law and Society (4th edn, Oxford University Press 2019) 41.
[2] ibid 37.
[3] Katerina Maniadaki, ‘Net Neutrality Regulation in the EU: Competition and Beyond’ (2019) 10(7) Journal of European Competition Law & Practice 479.
[4] Johan Rochel, ‘Towards a just Internet access: republican net neutrality for the EU’ (2020) 28(2) International Journal of Law and Information Technology 112–138.
[5] Case C-854/19 Vodafone GmbH v Bundesrepublik Deutschland [2021] EU:C:2021:675.
[6] Maniadaki (n 3).
[7] European Parliament and Council Regulation 2015/2120 of 25 November 2015 laying down measures concerning open internet access and amending Directive 2002/22/EC on universal service and users’ rights relating to electronic communications networks and services and Regulation (EU) No 531/2012 on roaming on public mobile communications networks within the Union [2015] OJ L310 (‘Open Internet Regulation’)
[8] Throttling is defined as the process during which an “Internet Service Provider deliberately restricts […] the internet bandwith or speed” without informing the user. This “results in speeds slower than what […] the ISP should be serving” Deepan Ghimiray, ‘ISP Throttling: What It Is and How to Stop It’ (Avast Academy, 2 August 2021) < https://www.avast.com/c-how-to-stop-isp-throttling#gref> accessed 7 February 2022.
[9] Ofcom, ‘Ofcom’s approach to assessing compliance with net neutrality rules’ (Ofcom, 16 May 2019) <https://www.ofcom.org.uk/__data/assets/pdf_file/0014/148100/ofcom-approach-net-neutrality.pdf> accessed 12 January 2022, page 3.
[10] Case 807/18 Telenor Magyarország Zrt. v Nemzeti Média [2020]. EU:C:2020:708
[11] ibid.
[12] Vodafone (n 5)
[13] Jesper Lund, ‘CJEU in surprise judgment: zero rating is illegal under EU law’ (EDRi, 8 September 2021) <https://edri.org/our-work/cjeu-in-surprise-judgment-zero-rating-is-illegal-under-eu-law/> accessed 30 October 2021.
[14] David Meyer, ‘In boost for net neutrality, top EU court plunges dagger into controversial ‘zero rating’ practice’ (Fortune, 2 September 2021) <https://fortune.com/2021/09/02/in-boost-for-net-neutrality-top-eu-court-plunges-dagger-into-controversial-zero-rating-practice/> accessed 30 October 2021.
[15] Open Internet Regulation (n 7).
[16] ibid art 3(3).
[17] Lund (n 13).
[18] ibid.
[19] Voxi Vodafone, ‘Choose your plan’ <https://www.voxi.co.uk/plans> accessed 30 October 2021.
[20] Lund (n 13).
[21] Open Internet Regulation (n 7),
[22] Lucie C. Audibert and Andrew D. Murray, ‘A Principled Approach to Network Neutrality’ (2016) 13(2) SCRIPTed 118.
[23] Mark A.Lemley and Lawrence Lessig, ‘The End of End-to-End: Preserving the Architecture of the Internet in the Broadband Era’ (2001) 48 UCLA Law Review 925.
[24] Audibert and Murray (n 22).
[25] Ellen P. Goodman, ‘India’s ban on Facebook’s free service is an overreaction’ The Guardian (London, 9 February 2021) <https://www.theguardian.com/technology/2016/feb/08/indias-ban-on-facebooks-free-service-is-an-overreaction> accessed 12 January 2022.
[26] ibid.
[27] Torsten J. Gerpott, ‘Zero-rating arrangements of mobile Internet access service providers – An analysis of main factors shaping the need for regulatory interventions’ (2018) 42 Telecommunications Policy 489.
[28] René Arnold and others, ‘The value of network neutrality to European consumers’ (26th European Regional Conference of the International Telecommunications Society, Madrid, June 2015).
[29] Audibert and Murray (n 22).
[30] Doug Brake, ‘Mobile Zero Rating: The Economics and Innovation Behind Free Data’ (Information Technology & Innovation Foundation, 2016) 1 – 23.
[31] Rochel (n 4).
[32] Open Internet Regulation (n 7) Recital 11.
[33] Rochel (n 4).
[34] Murray (n 1).
[35] Matteo Turilli, Antonino Vaccaro and Mariarosaria Taddeo, ‘Internet Neutrality: Ethical Issues in the Internet Environment’ (2012) 25 Philosophy & Technology 133, 143.
[36] For an economic analysis of net neutrality see: Sebastien Broos and Axel Gautier, ‘The Exclusion of Competing One-way Essential Complements: Implications for Net Neutrality’ (2017) 52 International Journal of Industrial Organization 358.
[37] Helani Galpaya, ‘Zero-rating in Emerging Economies’ (2017) 47 Global Commission on Internet Governance 11.
[38] Robert Hahn and Scott Wallsten, ‘The Economics of Net Neutrality’ (The Berkeley Electronic Press, June 2006) <https://courses.ischool.berkeley.edu/i205/s10/readings/week12/hahn-wallsten-neutrality.pdf> accessed 30 October 2020.
[39] ibid.
[40] Audibert and Murray (n 22).
[41] John Perry Barlow, ‘A Declaration of the Independence of Cyberspace’ (1996) 18 Duke Law and Technology Review 5.
[42] Christopher T. Marsden, ‘Net Neutrality: Towards a Co-regulatory Solution’ (1st edn, Bloomsbury Academic 2010).
[43] Audibert and Murray (n 22).
[44] Rochel (n 4).
[45] Lund (n 13).
[46] Galpaya (n 37).
Aikaterini Pampouki
LLB (LSE) ’22