Last week, after 18 months of negotiations, the Council, Parliament and Commission of the EU have reached an agreement about the telecoms regulation which should, finally, end roaming and safeguard net neutrality across the EU. Or will this deal actually end roaming and net neutrality?
This is what Commissioner Oettinger tweeted about the agreement:
“Here are the details on the end of
#roaming & #netneutrality: http://europa.eu/rapid/press-release_IP-15-5265_en.htm … http://europa.eu/rapid/press-release_MEMO-15-5275_en.htm … #tsm political deal reached”
Although Oettinger probably did not mean ‘the end of net neutrality’, digital civil rights organizations fear that this is precisely what the agreement will achieve. Furthermore, the Dutch government (represented in the Council by Minister Henk Kamp of the liberal party) publicly expressed disappointment with the result of the negotiations in the Council regarding net neutrality. According to Kamp the position of the Council (informally contained in this leaked ‘non-paper’) would result in a substantially lower standard of protection of net neutrality than the rules in force in The Netherlands since 2013. If this is true and if the regulation is to achieve its purpose of complete harmonization, the Dutch stand to lose some of their net neutrality protection. How much exactly? Well that remains to be seen.
The strongest net neutrality protection in the world?
Thus far there is no official or definitive text of the agreement published, but an elaborate press release intends to shed light on the key questions. According to the press release, the net neutrality rules provide the strongest protection in the world. Is this a gross exaggeration? Have there been substantial improvements after the Council’s position? Or have civil society, the Dutch government, and 126 EU Parliamentarians cried wolf? It actually appears to be non of the above, or perhaps a little bit of everything.
No blocking or throttling
Judging from the press release and the Council’s non-paper, the regulation will essentially be quite similar to the Dutch law on net neutrality, and prohibit any blocking, throttling, degradation or discrimination of Internet traffic by Internet service providers, except if this is necessary and proportionate for a narrowly circumscribed legitimate aim in the network management sphere (e.g. to preserve the security and integrity of the network and to mitigate network congestion). The general format thus resembles that of article 10 ECHR, as is advisable.
A comparison between the Council’s non-paper and the Dutch law further shows some differences in wording which may support the press release’s claim that the proposed EU net neutrality rules would be even stronger than those of The Netherlands. For example:
“Providers of internet access services may implement reasonable traffic management measures. In order to be deemed reasonable, such measures shall be transparent, non-discriminatory, proportionate and shall not constitute anti-competitive behaviour, and shall be based on objectively different technical quality of service requirements of specific categories classes of traffic. Such measures shall not be maintained longer than necessary. Providers of internet access services shall not engage in traffic management measures going beyond the reasonable measures set out in sub-paragraph 2, and in particular shall not block, slow down, alter, degrade or discriminate between specific content, applications or services, or specific categories or entire classes of traffic, except as necessary, and only for as long as necessary, to:…” (followed by the list of legitimate aims)
appears on the whole to be (even) stronger and more explicit than:
“Providers of public electronic communication networks which deliver internet access services and providers of internet access services do not hinder or slow down applications and services on the internet, unless and to the extent that the measure in question with which applications or services are being hindered or slowed down is necessary to:…” (followed by the list of legitimate aims)
(Excerpt from the unofficial translation of the Dutch law by Bits of Freedom)
So far so good then! (Although it could be regarded as a slight omission that providers of the underlying physical and virtual communications networks used by ISP’s have not explicitly been included in the scope of the EU net neutrality provisions.)
Paid prioritization will be banned
In the non-paper text there is no legitimate aim which would allow paid priority business models over the Internet, and the press release explicitly confirms that paid prioritization in the open Internet will be banned. Another checked box for strong net neutrality protection in the EU.
Specialised services are allowed, provided this is not to the detriment of Internet access quality
According to the press release and the non-paper, specialised services would be allowed. Specialised services are described in the press release as “services like IPTV, high-definition videoconferencing or healthcare services like telesurgery”. The non-paper avoids the term ‘specialised services’, but does allow them:
“Providers of electronic communications to the public, including providers of internet access services, and end users, including providers of content, applications and services shall be free to agree on the provision of services which are distinct from internet access services and which are optimised for specific content, applications or services, or a combination thereof, in order to meet their requirements for a specific level of quality”
Specialised services are called ‘separate services’ in the Dutch law, which does not contain a formal definition of the term. The explanatory memorandum provides examples such as IP-TV and VoIP (as a separate telephony service and not an online application such as Skype). Providers are allowed to offer separate service
s, but may not offer packages or bundles to access a part of the Internet. The Dutch law does not explicitly provide that separate services may not be to the detriment of the quality of Internet access.
The non-paper does provide for this restriction, but adds (compared to the Parliament’s version) the proviso that this applies only to other Internet users. Thus: your IP-TV could take up some capacity from your Internet connection (and possibly even be prioritized over it) but not from your neighbor’s connection. And your provider will be obliged to inform you about the impact of this. E.g. so that you can choose not to turn on your TV if you have to make an important Skype call. (Although it would be even better if users could themselves control, in real time via their router’s settings interface, which of their own IP-based services have priority in case of congestion).
I fully agree with the press release that this does not result in a ‘two tier Internet’. It merely allows for separate, non-Internet lanes to exist next to the open Internet lane, as has always been the case. What is important is that it is ensured that those separate lanes do not capture capacity from the Internet lanes or take general priority over the Internet (of other end users). Therefore, another checked box for strong net neutrality protection in the EU.
Zero rating could be allowed
The Dutch law and the ACM’s interpretation are quite clear about zero rating (also known as sponsored connectivity). It is not allowed to exempt the data traffic of particular online services from an Internet subscription with a capped data allowance. This in order to prevent that ISP’s would still get to pick the winners and losers in the online sphere through restrictive data caps and carefully chosen exemptions (rather than by discriminatory traffic delivery). Separate (and non-metered) subscriptions for non-Internet IP-TV and VoIP, or even Spotify or Netflix, are possible, however, as long as they are not tied to an Internet subscription.
There is no mention of zero rating in the non-paper, while the press release is quite vague about this. It appears that zero rating would not be prohibited, at least initially, but regulators should keep a watchful eye to prevent that zero rating effectively restricts freedom of choice.
Perhaps the agreement after the trialogue is not quite the strongest net neutrality protection in the world then, or at least not yet. But it is not far off, and it is to be hoped that perfect does not become the enemy of good enough here. The European public have waited long enough for strong net neutrality (even if not the very strongest). And this certainly also goes for the end of roaming charges within the EU.