Home > Media and Analysts > Speeches and Presentations > 2007 > Feb > Net Neutrality Briefing

26|02|07

Joint CEPS and Progress for Freedom conference

22 February 2007

Given by Tom Kiedrowski on behalf of David Currie

1. Introduction

• The term 'net neutrality' is frequently used in a number of ways, often depending on who is arguing the case. Broadly, it is the debate around whether and where there should be a principle of non-discrimination regarding different forms of internet traffic carried across networks.
• What I’d like to do today is:
  
  
  • Spend a short time on some of the different flavours of net neutrality;
  • Explain how Ofcom thinks this debate applies to the situation in Europe – there are some important differences between Europe and the US which we believe distinguish our approach; and
  • Explain Ofcom’s position is on this issue – and particularly emphasise some of the consumer issues relating to net neutrality.

2. What is net neutrality?

• This debate is much more subtle than the often-polarised views would make out. There are a number of reasons for this.
  
  
  • First, some have argued that to date, the net has treated different types of traffic equally, and has been free. Neither of these is true. Application providers have always had to pay for their content to be hosted. Different hosting arrangements cost different amounts, and offer differential quality of service to the end-user.
  • Second, the debate is typically argued from a supply-side perspective; that is, whether network operators should be allowed to block, degrade, or charge for prioritising, application providers’ traffic. Yet there is a consumer dimension to net neutrality. Prioritising traffic allows network operators to offer tailored packages to end-users, for example which were optimised for gaming or VoIP. This means that services can be delivered at a quality appropriate to the application. I’m going to discuss some of these consumer issues in more detail later on.
  • Third, net neutrality is best thought of as a continuum. The most extreme definition of net neutrality is that there should be no prioritisation of any types of traffic by network operators – that is, ‘all bits are equal’. As I’ve said, this is an extreme definition – and this isn’t a position that is widely argued in Europe at present. However, there are a range of approaches that network operators may adopt. Some have important consumer benefits, like degrading certain types of traffic, such as peer-to-peer traffic, which would otherwise cause the network to gridlock. However, it is the prioritisation or degradation of traffic, based on identity of application provider, that is most contentious in the European context.
    
• Given these subtleties, it is important to define net neutrality tightly. We think it is most useful to define net neutrality as any situation where the network operator does not:
  
  
  • prioritise one application, or one application provider’s traffic over another;
  • deliberately degrade any application, or application provider’s traffic;
  • charge any application provider for providing a higher quality of service; or
  • deliberately block any one application or application provider’s traffic.

3. Differences between the Europe and US

• The net neutrality debate originated in the US and was triggered by a change in the way that broadband networks are regulated. The obligation on incumbent telcos to offer unbundled access was recently removed - something which we in the UK see as a key tool in addressing competition problems.
• In fact, most incumbents in Europe are required to offer unbundled and/or bitstream access to broadband. As a result, the retail ISP market has lower barriers to entry than in the US and it remains relatively competitive. This is the case in the UK for example, where no ISP has been found to have SMP in the retail market for broadband internet access.
• The lack of unbundling in the US is likely to mean rapid concentration in the retail ISP market. The result is generally that, at the retail level, consumers only have a choice of broadband operator to the extent that they are able to choose between operators with lines connected to their home – a choice between cable and telcos, rather than multiple ISPs offering bitstream or LLU access.
• This lack of choice, it is argued, could lead to abuse of the operator’s resulting market power. And this, it’s then argued, needs to stopped by outlawing any “non-neutral” behavior using new legislation. This is an important point because it is the retail ISP, operating at the IP layer of the network, which can have most impact in prioritising traffic.
• The differences between the UK and US have been noted by the pro-net neutrality lobby, in the form of Vint Cerf, Google’s Chief Internet Evangelist. In his testimony to a US Senate committee, he used the separation of BT’s wholesale and retail operations as an example of the non-discrimination absent in the US.
• As the UK has such strong retail ISP competition, these problems, and therefore a strong case for new ex-ante legislation, are much harder to imagine.

4. The debate in the Europe

• Despite these very important differences, we’ve seen a lot of discussion about net neutrality in the Europe. It is an important debate because it is about a change in the kind of network the internet is – and this is as true in Europe as it is in the US.
• Because this debate concerns such fundamental issues about the future of commercial relationships and payment flows, we think it is unlikely to go away. And three long-term trends are driving this change:
  
  
  • Rapidly increasing traffic on the internet. People are using the internet for applications that generate vastly more traffic than before. For example, three years ago a small email would be considered light usage of the internet and a graphics-heavy website heavy usage. Now, a small email is still light usage but an HDTV film would be considered ‘heavy usage’. This rapid increase in traffic is generating substantial congestion in some parts of the internet.
  • Change in the nature of applications on the internet. People are increasingly using the internet for time-sensitive applications, such as VoIP, as well as activities such as downloads which are much more tolerant of delay.
  • Increased intelligence in network routers. The internet was designed to be a network with most of its processing power at the edge; internet routers were kept simple in order to cope with large volumes of traffic. But today’s routers are capable of identifying packets associated with different applications, and prioritising them accordingly.
    
• All of these trends create a desire, and deliver an ability, to change the internet from a network that delivers all traffic on a best-efforts basis, to one that prioritises traffic according to particular criteria. We believe that they are all trends that are likely to continue, if not accelerate, in 2007.

5. What is Ofcom's position

• While this is an important debate, we believe that market structure and regulatory policy differentiate Europe from the US, particularly because regulators already have extensive powers to address net neutrality concerns under the existing European Framework.
• One of the important principles underlying the way Ofcom regulates in the UK is that we have said we will act with a bias against intervention. That principle is important in the case of net neutrality because – were we to intervene if we were to follow the path suggested in the US – we would be dictating a new market structure – a very dangerous place for a regulator to be.
• As I’ve already mentioned, the UK has strong retail ISP competition. In a competitive market, with consumers that are well informed on the activities of different ISPs and who can easily move to competitive providers, competition itself can provide a constraint – consumers who are aware of the issue and can choose to change retail ISP could be expected to do so. That is why competition, consumer information and customer migrations are important issues for the net neutrality debate. I shall expand on these issues further in a moment.
• But we also believe that we already have the tools, in the form of the European Framework, to address issues in relation to net neutrality as they arise.
• If a network operator with SMP was charging for prioritisation, or blocking or degrading traffic, in ways which we considered anti-competitive, the current European Framework gives us the tools to put in place the necessary remedies.
• Remedies that might be developed include obligations to supply, charge caps, the potential to impose minimum quality of services obligations or mandate ISPs to provide particular information to consumers such as whether they block access to certain ports or websites. However, net neutrality rules might become one of the remedies imposed to address that SMP. Although net neutrality remedies are not explicitly described in the Framework, they could be easily developed as an iteration of the existing non-discrimination rules.
• If a network operator without SMP was charging for prioritisation, or blocking or degrading traffic, then because of the operator’s lack of SMP, this activity would be unlikely to be anti-competitive, and not therefore need to be regulated.
• If, however, ISPs were prioritisating, degrading or blocking traffic in a way that caused concern, there could be three possible ways to address this problem through ex ante regulation:
  
  
  • the ISPs could be found to have joint SMP. However, the hurdle for finding joint SMP is rightly very high.
  • we could apply Article 5.1 of the Access and Interconnection Directive, which allows NRAs to impose ex ante obligations on operators to ensure end-to-end connectivity, without the need to find SMP.
  • we could use our powers derived from the European Framework which enable us to require suppliers (even non SMP suppliers) to comply with various general conditions in order to take part in the market to address particular issues, for example, in relation to information transparency.
    
• So at this stage we don’t consider that the European Framework requires adjustment in this area but we will continue to monitor developments in this area, and to keep our position under review.
• These arguments have focused on net neutrality from a competition perspective. But I haven’t yet mentioned the arguments about innovation. The argument goes that the most fundamental innovations on the internet have come from start-ups; from teenagers in garages. If these innovators had to pay to allow end-users to access their applications at the required quality of service, the rapid innovation on the internet would effectively end.
• The other side of that argument is that differential quality of service levels may also allow increased scope for innovation – VoIP is a good example of this. Without prioritisation, no consumers would be likely to take up these kinds of innovative services because their quality wouldn’t be good enough. So the innovation argument potentially runs both ways through this debate.
• But this brings me to the important point that consumer transparency and ease of migration are vital in the net neutrality debate.

6. Consumer information and migrations

• There are two areas that we remain concerned about in relation to net neutrality: consumer information transparency, and service migrations. We think these are increasingly important.
• We remain concerned to ensure that there is:
  
  
  • Transparency. So that consumers are fully informed of any traffic prioritization, degradation or blocking policies being applied by their ISP (and well enough in advance of any changes) so that competition is effective at the retail level. To take an extreme example, if Tiscali blocked its subscribers from accessing Yahoo, this would need to be well-known if this action was to act as a competitive disadvantage to Tiscali in the retail market.
    
    In Ofcom and the UK Government’s joint submission to the Framework Review, we wholeheartedly agreed with the Commission’s objective of increasing transparency and facilitating access to third-party price information for comparison purposes and we look forward to seeing the precise proposals to be put forward by the Commission later this year.
  • Ease of migration. So that industry migration processes are industrialised, smooth and low cost. Ofcom has been particularly focused on broadband migrations – having recently undertaken a review of this issue. In response to evidence that consumers were finding it difficult to switch between broadband providers, we have introduced new rules about broadband switching to require all providers to comply with a migration authorization code (MAC) and to put in place a code of practice and high level obligations designed to ensure consumers are able to switch broadband providers easily.
    
• Provision of this kind of consumer information in a way that consumers can understand is likely to be a particular challenge. To some extent, broadband and internet users have shown themselves to be capable of detecting technical restrictions introduced by ISPs and using collective power to lobby for removal of restrictions they do not want. This was true, for instance, when ISPs introduced download caps to curb heavy peer-to-peer usage.
• But for the mass market, broadband is a complex enough consumer proposition without having to consider the extent to which certain application providers are prioritised or de-prioritised by ISPs.
• It is a supplier’s responsibility to ensure that information about these kinds of considerations is communicated to consumers in a way that is understandable and we’re encouraged that ISPA has published a best common practice document for their members, encouraging them to provide clear information about internet filtering practices.
• But if there was evidence that the market wasn’t doing so adequately, we may need to revisit this issue. And the powers that I have already mentioned under the general conditions to require suppliers – even non SMP suppliers – could be used to address, for example, the information that providers supply consumers about their products.

7. Conclusion

• In summary we don’t believe that net neutrality is as significant an issue in Europe as it has been in the US.
• While the debate is important and is likely to continue for the reasons that I’ve outlined, we believe that the existing market structure and regulatory policy in Europe, combined with powers to address net neutrality concerns under the existing European Framework will be sufficient to address issues that arise in relation to net neutrality now and in the future.
• We will be closely monitoring developments in this area and will be particularly concerned to ensure that consumer information transparency exists, and service migrations are simple, easy and effective for consumers.

Back to top