After a long period of calm, a number of recent events have reignited the debate around net neutrality and the commercial relationship between ISPs and content providers. Overarching regulatory frameworks and the approach to zero-rating in particular are under question across the EU, South Korea, UK, and US.
COVID-19 highlighted the importance of having a flexible approach to zero-rating access to certain content. This consideration is at the heart of a review Ofcom is currently carrying out in the UK. Now that the country no longer must follow the European framework a conversation has been started between regulator and industry on whether current arrangements are fit for purpose.
Recent rulings of the European Court of Justice have made a strict interpretation of the Open Internet Regulation, and put the practice of zero-rating at risk. BEREC will now need to review its guidance on this commonplace practice, with ISPs and Big Tech keen to maintain a flexible approach. The worst case scenario is that the Regulation may need to be rewritten.
In South Korea, the balance of power between ISPs and content providers is shifting in favour of the former. The presence of network usage fees could see Netflix having to pay in excess of £100m to SK Broadband. Operators elsewhere will be watching with interest, although the Korean case is likely to remain an anomaly for now.
So far, consumers have generally fared well, with little evidence of harm. This is mostly as a result of strong competition between ISPs, rather than the presence of regulatory safeguards. Possible harm may arise where zero-rating schemes exclude smaller content providers. This possibility justifies ongoing monitoring of net neutrality, and the need for case-by-case assessments.
Things look very different compared to a decade ago, when net neutrality was really only an issue in the US
The debate around net neutrality has been reignited, after a long period of quiet. 10 years ago, the US was really the only country where net neutrality was seen as a significant issue, due to less competition between ISPs at the retail level and some blocking of content and applications (e.g. Comcast blocking peer-to-peer services such as BitTorrent). Under the Obama administration, the FCC regulated net neutrality with two orders (the Open Internet Report and Order of 2010, and the Open Internet Order of 2015 after the first was defeated by legal challenges). In Europe, it was largely a non-issue. Some regulators such as Ofcom in the UK opted for a hands-off approach, monitoring the issue without finding a need for intervention. The only problems arose from the treatment of VoIP and P2P applications, which in 2012 the EC found to be restricted for a large number of users (20–50% of all EU mobile customers). At the time, zero-rating was not yet common practice. In the years that followed, the debate around the impact of zero-rating on competition and consumer welfare remained open – so much so, that the EU Regulation of 2015 and several other frameworks around the world continued to allow it, or left regulators to decide on it case-by-case. In the vast majority of cases where regulators have investigated tariffs with a zero-rated element, they’ve allowed them to continue. Our Net Neutrality and Zero-Rating Tracker shows that only in 4 out of 22 cases were offers blocked. In 8 cases they were approved without changes, and in another 10 they were approved with some minor tweaks.
Things now look somewhat different (see Figure 1). COVID-19 highlighted the importance of having guaranteed access to certain resources online at all times, which in many countries led to the zero-rating of health and educational websites. This issue is at the heart of Ofcom’s current review of the net neutrality framework in the UK. Following its exit from the EU, the country has inherited the EU Regulation of 2015, but is now free to chart its own course – and could revert back to a more hands-off approach. In the meantime, recent rulings of the European Court of Justice could result in the opposite happening in the EU. According to these rulings, any traffic management based on commercial considerations is incompatible with the Regulation. As a result, BEREC is now reviewing its guidelines on zero-rating. A more restrictive approach is in the cards – unless the EC steps in and changes the Regulation to provide more clarity. In South Korea, two court rulings this year ordered Netflix to pay network usage fees to SK Broadband – something that will prick the ears of ISPs elsewhere.
Content providers in South Korea pay network usage fees, but this is likely to remain an anomaly for now
Since 2016, the Korea Communications Commission (KCC), allowed ISPs to collect ‘network use fees’ from some content providers. In practice, this marked the introduction of a sending-party-pays model for internet traffic, which differs from the typical bill-and-keep arrangement used worldwide. Until recently, this model was applied selectively (mainly between ISPs), however amendments to the country’s Telecommunications Act in 2020 explicitly required content providers with at least 1m users and those responsible for more than 1% internet traffic to share the responsibility (and the cost) of providing a stable service to users.
Some content providers, such as Netflix, have avoided paying these fees, but are now being asked to contribute. In 2019, SK Broadband asked the KCC to broker an agreement between the two parties. Netflix argued it did not have an obligation to pay and filed a lawsuit in April 2020. In June this year, the court sided with SK Broadband and ruled that Netflix is receiving network services such as network quality management, which come at a cost. The value of the fee will be left to commercial negotiation, although the court estimated that the fee to be paid to SK Broadband would be around KRW200bn (£125.3m). Despite the ruling, Netflix has confirmed it will continue its partnership with SK Broadband and both parties have called for cooperation. However, this didn’t stop SK Broadband from filing a new lawsuit in October, due to the surge in traffic caused by the popularity of the show ‘Squid Game’. SK Broadband argues that an additional KRW27.2bn (£22.8m) is due. Netflix is again questioning that it has such an obligation and worries it could be replicated elsewhere.
Despite Netflix’s concerns, the Korean case is likely to remain an anomaly, since legislators presumably acted to protect their national champions. In 2012, they did not side with the largest ISP, KT, when the operator tried to block Samsung’s ‘smart TVs’ after Samsung refused to pay for the surge in network traffic they generated. At the time, KT’s move was criticised by the KCC which eventually arbitrated an agreement between the two sides. The Korean case is even more unique considering that the KCC adopted net neutrality guidelines enshrining the principles of no blocking, no throttling, no unreasonable discrimination, which hardly square with a framework where content providers have to pay ISPs. While there have been cases in other countries, these have been few and far between. For example, Netflix maintains it no longer pays Comcast in the US, and has settlement-free interconnection with more than a thousand ISPs worldwide through its ‘Open Connect’ programme.
The future of zero-rating is uncertain in both the EU and US
In the absence of net neutrality rules, ISPs in the US have been able to extract revenue from content providers by charging them to participate in zero-rating programmes – such as AT&T’s Sponsored Data programme. Under the scheme, AT&T zero-rated video services as it does with its own HBO Max. After the approval of California’s net neutrality law (which came into force in March 2021), AT&T ended the scheme, noting that the Californian law would forbid it and arguing that a state-by-state approach to net neutrality is “unworkable”. However, the law does not forbid zero-rating per se, so long as it's applied to an entire category of service, and does not amount to paid-for prioritisation. The operator also allowed end users to opt out from the Sponsored Data programme, which suggests it could have excluded users in California rather than terminating the scheme nationwide. How zero-rating will be treated in the future of the US is currently uncertain. The Biden administration is expected to reintroduce net neutrality rules, although there is currently no indication of the shape these will take. Most likely they will be inspired by the same principles of the 2015 rules, and allow zero-rating while leaving the FCC the power to decide on specific cases.
In the EU, recent rulings of the European Court of Justice have created significantly more uncertainty around the practice of zero-rating. In September 2021, the ECJ issued three rulings making a restrictive interpretation of the EU Regulation on Open Internet Access, and argued that any form of traffic management based on commercial considerations is incompatible with Article 3(3) of the Regulation. This has prompted BEREC to revise the guidelines for the enforcement of the Regulation for the part that relates to zero-rating. A preliminary consultation was held in October 2021, and reopened the debate on zero-rating along the usual lines i.e. whether or not it is good for consumers and compatible with net neutrality.
Consumer bodies across Europe have come out against zero-rating. The European Consumer Organisation (BEUC) is arguing that zero-rating is a “clear violation” of the principle of net neutrality and of the EU Regulation. Operators are defending the practice and are pushing for an interpretation of the rulings that will allow it to continue. This is best summed up in the position expressed by ETNO and the GSMA, which argue that zero-rating should be seen as a practice of differentiated billing rather than one of traffic management. As such, it cannot be the source of unequal treatment of internet traffic. Given the favourable treatment of their applications and services, Big Tech is perhaps unsurprisingly also advocating for a continued flexible approach to zero-rating. Google believes the ECJ’s ruling would still make zero-rating of entire categories of services possible, and that such an approach would align the EU with the framework currently in place in the US since California’s net neutrality law has been in place. We expect BEREC to take on board the views of the industry, and to adjust its guidelines in a way that still makes zero-rating possible. However, the Regulation may need to be rewritten to better clarify the circumstances under which zero-rating is allowed. Without that, future ECJ rulings could confirm the interpretation of the recent ones and undo the work that BEREC is now carrying out.
What has the consumer experience been so far?
While consumer bodies have generally spoken in favour of protecting net neutrality, the impact of the rules on consumer welfare is not always clear cut. In many cases, good consumer outcomes are more often linked to how competitive and mature a market is, than to any safeguards regulators may have put in place. There is generally a mutual interest of ISPs and content/service providers to cooperate with one another and ensure end users can access the content they value. As a result, ISPs have little incentive to block any lawful content, since this could result in upsetting customers who could feel they are not getting value from the connectivity they purchase. This is especially true in competitive markets, where there are many alternatives and switching providers is relatively easy and where operators risk losing customers to their rivals. Even in the US, in the absence of net neutrality rules, no cases of blocking have emerged.
One of the arguments often made against the practice of zero-rating is that it excludes small, newer content and applications providers, who could struggle to gain scale if consumers face a barrier to adopting their services over larger more established players who have a relationship with telcos. A number of smaller content providers have lamented the obstacles they face in accessing operators’ zero-rating programmes. In response, the 2020 update of BEREC’s guidelines encouraged regulators to assess how ‘open’ a zero-rating programme is, and whether its terms are transparent and non-discriminatory.
The evidence on these concerns holding true isn’t necessarily forthcoming. When EE launched 4G in the UK back in 2012, it did so with a partnership with music streaming service Deezer, that was barely known at the time. The use of Deezer did not come out of a consumer’s data allowance, and was one of the first examples of zero-rating. It remained part of EE’s Music Data Pass for several years, alongside more popular services such as Apple Music and Spotify. Similarly O2, includes Soundcloud (another relatively small content provider) in its Music Allowance.
