Don’t harm the best effort Internet to compensate for the loss of roaming income

This post has undergone various limited revisions since its original publication. When not purely cosmetical, these revisions are signaled through footnotes.

The European Parliament is presently in the last stages of first reading for a revision of the telecom regumation. This revision addresses two facets: roaming costs for international mobile communications and network neutrality. Cancelling or strongly decreasing (depending on cases) roaming costs is useful, but it cannot be done at the expense of Internet users and of the rights of all to impart content and services on the Internet.

Telecommunication companies, in particular incumbents, once drew large revenues from fixed line voice communication. This income is now severely reduced due to VoIP and multiplay offers. Internet Service provision is a profitable business, but its predominant model of fixed cost per month “illimited” traffic has kept its revenues well below the old pay-per-call model of voice communications.

This is excellent news for consumers and for the economy. If there is one thing that created growth and jobs in Europe, it is the illimited (though bandwidth constrained) lump sum subscription to the Internet, the generalized household equipment that it enabled and the possibility for all Web service providers to reach potential consumers on the internet. Even more importantly the development of the open neutral Internet has led to an immense development of non-market (not-for-profit, non-commercial) practices of individuals, empowering them in all domains: literacy, culture, democracy. We are just at the beginning of this process, which is fragile because of the trend towards recentralization of Web services by large companies. Keeping the Internet and the Web open and non-discriminatory is thus a duty for all reasonable policy-makers.1

The loss of revenues from fixed voice communication was partly compensated by revenues from mobile phone communications. The Commission and National regulatory authorities have tried to inject more competition in the mobile telecommunications. They succeeded partially in national markets and the Commission now implements a policy to cut roaming costs. This is a valuable objective.

However, if the MEPs listen too uncritically to the discourse of telecommunications operators (in particular ETNO), the cost to pay for this progress will be to hurt an even more important common good: the open best effort Internet (see definition). There is a hidden deal being pushed by some, which would permit the European Commission and the other European institutions to terminate the present legislature with the perspective of a wonderful gift to consumers2. The other side of the deal is that there is a compensation for the pains suffered by telcos due to losing the roaming income. We take away one rent from you but we will safeguard another one by letting you charge for “added-value” services on the Internet. Anyone who has really worked in telecommunication innovation knows that for 20 years, the telcos and ISPs have dreamt of only one thing: money for nothing in the Internet world. They first went for selling contents when they believed that Content was King. When this proved to be balloonish, they moved to another idea: let’s add new services to the Internet that have specific bandwidth requirements,3 such as HDTV distribution, VOD. Those services, were charged asides of the subscription, partly to pay providers, and partly to offer a specific bandwidth on specialized networks. At this stage, nobody complained.

That’s were the whole debate on network neutrality in Europe comes in. Because the real dream of telcos and ISPs is to charge for things that are already or would naturally develop on the Internet. In order to do so, they must be able to claim that it is “necessary” to obtain the needed “quality of service”. There it is worth travelling back in time, as far as the 1960s, when the telecommunications monopolies considered that among those things that needed “quality of services” than no packet network could deliver were the telephone and telefax. Still in the 1990s, I heard repeatedly stated by their representatives that that no truly large scale voice communication system would ever work on the best effort Internet. It is also worth considering that there are 4 types of Internet applications and services depending on their real-time and synchronization constraints :

  • those that are impossible because of minor physical constraints such as the speed of light ;
  • those that work perfectly well on the best effort Internet provided enough bandwidth exists in the network and enough computing power at the periphery, which represent the vast majority of useful applications and services,
  • those that require only sufficient bandwith reservation, buffering, and optimized routing without discriminatory priorization, and,
  • some examples, such as telesurgery, that present a need for special treatment in cases where the “no guarantee” property of best effort protocols would become inacceptable, even if in practice best effort networks work better in the great majority of cases. Unfortunately for the promoters of these examples, they have nothing to do with what is under discussion in the network neutrality debates: telesurgery is deployed entirely specialised networks, does not interfere with the open Internet in any way, and would not be prohibited by any of the amendements that are presently under discussion4.

Amendments have been proposed by NGOs defending the right to access and impart information and to innovate in information services. They have been tabled in a different form but with a similar spirit in the ITRE committee of the European parlimament by Catherine Trautmann and others and are opposed by rapporteur Pilar del Castillo Vera. These amendments have only one aim: to prevent ISPs to artificially limit the open Internet in such a way that they could sell as “added-value” or “specialized” services applications that would normally work on the open best effort Internet. The particular provision requiring that the “application layer [of specialized services] is not functionally identical to services and applications available over the public internet access service;” has particularly infuriated Mr. Jens Rohde of the ALDE group who stated in a Danish TV channel that “This is the ACTA movement again. But I will not be controlled by this.”5 But these amendments are not extremistic, they only protect one of our most precious common goods against the unfair competition of rent-seekers.

  1. Last two sentences added to stress that the Internet should not be regarded only through a market-based lense. []
  2. Meaning those consumers who travel a lot accross Europe and whose children live or study in other countries. []
  3. Or absence of jitter, but this can be coped by bufferization and other techniques provided that there is enough bandwidth. []
  4. This and the previous bullet points have been modified on the day of publication at 23:00 to benefit from remarks made by Benjamain Bayart. []
  5. informal translation, please check the Danish source: Så kender vi ACTA-bevægelsen igen. Men den lader jeg mig ikke styre af.. []

Comments are disabled for this post