'The Access Revolution: an evolution of regulation for competition'
KPN Annual Event
Brussels, 14 January 2008
Ladies and Gentlemen,
This conference addresses the central concern of telecoms sector today: the coming access revolution in high speed broadband. I have just been in Las Vegas for the annual Consumer Electronics Show and it is clear that convergence is on the march – converged high definition displays were on view everywhere. This is clearly the future. But to get to this future we need: first, services we can trust; second, open access to rich professional and user generated content; and third – the subject of today's conference – open, reliable and cheap bandwidth.
The state of play in high speed internet connectivity
Firstly, where are we today? After my visit to the USA last week, I am more than ever convinced that Europe is doing well on broadband availability and take-up. In July 2007, we had over 90 million connections in the EU25, 10 million up on January 2007. The average penetration rate of the 27 is at 18% and rising fast. Five European countries are world leaders (Denmark, Sweden, the Netherlands, Finland and Belgium) ahead of even South Korea and Japan, while the USA has slipped back to 15th place.
It has to be recognised that the EU average is being brought down by the slower moving EU markets. More than 30 percentage points separate the leading and the last-placed Member States. The single most important factor explaining this gap is lack of effective competition on the market because access regulation has not been effectively implemented.
In December, in its latest Lisbon Strategy update, the European Commission laid down a challenge to all Member States to move ahead on broadband take-up, to achieve a 30% penetration rate by 2010 and to embrace this target in their National Lisbon Strategies this year.
To assist this process, by summer in the mid-term review of the i2010 strategy, I will publish a new indicator of broadband take-up in Europe that compares national performance, not only on broadband penetration but also geographic coverage, speed, competition and price.
Advanced roll-out
As the structure of our future broadband performance indicator implies, the game is no longer solely about broadband penetration but about bandwidth, quality and openness to new services.
Content is becoming an essential factor, with the development of services such as IP-TV and mobile TV, in allowing market players to differentiate offers by developing bundled products. Further service development is likely to result in the need for significantly higher broadband speeds of up to 100 megabit per second or more.
On bandwidth speeds in Japan and Korea are top, based on extensive fibre optic networks. It is true in this respect we lag behind, but it should be noted that we doing at least as well as the USA as regards speed, with average connection rates of between 1 and 2 Megabits per second and 12 million fibre connections in both the US and the EU.
In my discussions last week, I was surprised how critical the industry is of broadband services in the US. I found a widely held view that the European regulatory framework and its emphasis on access obligations to open up competition is not at all the impediment to investment and innovation that some market players claim, but actually the main reason Europeans are ahead of the USA in terms of usage of new web services such as social networking sites and virtual worlds.
The regulation of access networks keeps open the potential bottleneck to competition in broadband markets. European rules have provided a real choice for consumers.
The European model is empirically proven to promote not just choice, competition and innovation but also investment: in 2006, investment in the EU telecom sector reached another peak of over €47 billion, 5% up on 2005. This was the fourth year on year increase since 2003. By the way, European investment is at least at the same level as other major regions (Asia Pacific: €44.3 billion and North America: €43.7 billion).
Investment
But there is an important new question to answer as we look at next generation networks offering high bandwidth, available everywhere, cheap and open to innovation and competition. Where will the investment come from to move us from legacy copper networks to new high bandwidth networks?
It is clear that next generation core networks are interesting to market investors: we have seen investments in such networks across most of Europe. But access networks are more difficult for two reasons. The cost of the upgrade is high (between 300€ and 1000€ per household on average, in many places even more!). Second, access is the bottleneck and therefore it creates competition problems. If one player can get hold of this bottleneck he will have the means and motivation to shut off the market to other players – or at least give access to who he wants and on the terms he wants.
How we treat next generation access is therefore the single most important policy question in the telecoms sector today. We have to create incentives for investment whilst making sure that no-one (and I insist on this no-one), can be in a position to foreclose the market.
What are the incentives to invest? The key is a stable and predictable regulatory environment. Investment decisions are finely balanced. If we are to encourage market players to invest in upgrades to the access networks, they will need to be able to offer a long term return on investment to their investors.
Regulation has a role to play here. I see it as entirely appropriate for regulators to allow infrastructure providers to make a reliable return on next generation access investments in return for testable guarantees of non-discrimination and an agreed plan for infrastructural investment that will lead to an open, high speed infrastructure. By the way, one of the potential attractions of functionally separating access networks is to make this incentive structure clearer and more operational.
The traditional revenues from voice calls are still large but year on year they are declining; there has been increasing revenue from broadband services, but not at a level that can offset this decline. Incumbents are searching for alternative revenue streams, bundling multiple services on their infrastructure is one of those.
An EU survey (1) showed that 18% of EU households subscribe to at least one bundled service, and that the most frequently purchased service package is a 'double play' offer of fixed voice telephony and internet. My worry is that such bundling will, de facto, stifle choice and innovation. The traditional operators have every reason to try to channel traffic through their own services and to close off the open nature of the internet, especially if they are faced with demands from investors for more revenue to pay back major investments.
(1) e-Communications Household Survey (July 2006), Special Eurobarometer, European Commission
How will we get there?
Let me be very direct: except where the structure of the market has non-discrimination built into it such as in a well designed system of functional or structural separation the incentive of the telecom company is to design new infrastructures in a way that controls or chokes off competition.
If we look from this perspective at the three different models of network upgrade that are being envisaged for next generation access we see some very interesting developments.
The first model involves VDSL deployment whereby the local loop is upgraded to fibre between the local exchange and the street cabinet. At the street cabinet, higher capacity VDSL modems are installed. Such deployments are relatively quick and cheap to deploy but face capacity constraints in the medium term as bandwidth demand rises up. In terms of open competition however there are serious concerns that VDSL could be attractive to incumbent telecom operators, because they require competitive market entrants to substantially scale up their investment in switching capacity. In even a small country such as the Netherlands a shift from ADSL to VDSL demands about a tenfold increase in the number of switches. Competitors that are unable or unwilling to scale up their investment would be forced out or forced back on the incumbent networks. I wonder, therefore, if VDSL is not a dead end street for both future capacity and competition.
The second model involves deployment of passive optical networks. This brings fibre-to-the-building using existing ducts, since a single fibre is used to carry the traffic to and from multiple users. In the short term it will deliver much higher bandwidths than ADSL or VDSL and is cheaper than end to end fibre to the home. But the flexibility in the medium term may be more limited, not least because the end user equipment and the equipment in the network have to be compatible. Unbundling these passive fibre networks is therefore more difficult and the incumbent increases control.
The third model is point-to-point fibre deployment. This is in fact the only fully future proof approach in terms of ability to deliver more and more capacity as techniques improve and as demand grows. Moreover it is the only approach to next generation access that permits a completely open access policy with the unbundling that has put Europe in the lead today. The difficulty here is cost: existing ducts are often too small to allow multiple fibres to pass through and therefore major construction spending is required. This is by far the most expensive option.
Will regulation persist?
We can see that the choice of which type of Next Generation Access is offered is not neutral. Market investors have a strong incentive to keep investment costs low and where possible increase their control over the access bottleneck. That is why in my opinion regulation will and must persist, so that open access is guaranteed through effective non-discrimination obligations.
The remedies will however have to be adapted to the access solution that has been implemented. In the case of VDSL, unbundling requirements at street cabinet would have to continue to allow competitive access operators to stay in business.
It is unclear that passive optical networks can be unbundled in the way that we see today on copper networks. This requires close attention and probably experimentation with novel architectures, using wave division technology to offer virtual unbundling as a more flexible alternative to bitstream access.
Point-to-point fibre deployment, meanwhile is rarely being deployed by private market investors. Certainly, this is due to its high cost, but it is also probably due to its openness. Where we do see it being used is in open access schemes initiated by municipalities, in cities such as Stockholm and Amsterdam. These schemes are local partnerships that take a pure "infrastructure utility" approach by building ducts and end to end dark fibre and then leasing access to service providers. Clearly by so doing these cities have created for their business and citizens a future proof network infrastructure and for the investors in the networks a very long term stable return on their investment given that ducts and dark fibre have a potential operating life of several decades. Under these conditions of guaranteed open access circumstances, perhaps, infrastructural competition is less important than an open and high performance platform.
However, the municipal solution seems unlikely to be relevant for all of Europe and could lead to a very fragmented landscape. Nevertheless: the industry and regulators, should consider well these models because they represent a legitimate concern by local representatives who want to have a future-orientated and open network in their city.
Whichever infrastructure route we take forward, my conclusion is clear: regulation will have a role to play to keep networks open and to guarantee progress, efficiency and choice. We are approaching a revolution in terms of network access; we have to make sure that the revolution delivers not just high speed access but choice and innovation.
Thank you for your attention.
