Competition and investment in superfast broadband

08 November 2011

Speech to Total Telecom World Conference

By Ed Richards

Chief Executive, Ofcom

8 November 2011

I want to share with you the regulator's approach for developing superfast broadband in the UK.

We have important duties and objectives as the independent regulator in the UK, established over many years and both supported and respected by governments of different political hues.

Nevertheless, today, in a period of very considerable technological but also social and economic change, our regulatory approach needs to be seen alongside the context of the public policy objectives set and pursued by government.

The government wants the UK to have the best broadband network in Europe by 2015.

The European Commission thinks all European homes should have access to at least 30 mbps by 2020, with half of those households subscribed to 100 mbps services.

These are ambitious goals. With our duty to encourage the availability and use of high speed data services, these are objectives of which we are obviously very supportive.

But we also have a unique regulatory task.

We need to consider a long term perspective which ensures that regulation meets today's goals but also ensures there are positive incentives in place for the longer term too.

Our approach must support the investment intensive transition from current generation to next generation broadband, while ensuring both copper and fibre networks work for consumers.

In addition to considering the next three years, we must also anticipate, to the extent that it is possible, the opportunities and risks of at least a five to ten year timescale.

To be effective, this approach has to survey the broad communications landscape, taking in mobile, other wireless and fixed networks. It must be alert to developments in devices, in software and applications, in the cloud and in content of many different forms.

It must understand how these different technologies and services interact, how they both complement and compete with each other, and what this means for both companies and consumers alike.

Some of the developments in the mobile and wireless area are going to be of huge importance in the next few years. For example, the auction next year of 250 MHz of released spectrum for 4G will dramatically increase the availability of next generation broadband services and make those services accessible on the move.

I could spend much of the time I have with you today on these kinds of developments and their importance to the overall mix of next generation broadband. However, given time constraints, today, I intend to focus my remarks on fixed networks.

Aims and objectives

Let me start with what we are trying to achieve. There is a perennial risk of trying to overcomplicate objectives - and while this is a complex area we should aim to set out what we are trying to achieve as simply as possible.

What we are trying to achieve is relatively straightforward:

  • We want to see widespread availability - of both current and next generation broadband services based on efficient investment and sustainable returns;
  • We want to see effective competition between many providers as deep within the infrastructure as is economically sustainable;
  • We want to see high levels of innovation in networks, devices and services.
  • And as a result of these three things, we would expect UK consumers to benefit from:
  • meaningful choice between suppliers, and a range of different products and services
  • Low prices and good quality of service - which compare well internationally in the value they deliver
  • High levels of take up arising from the nexus of price, quality of service, choice and innovation

Where we are today

That's what good looks like to us. So where are we today?

The UK currently has nearly 20 million active fixed broadband connections, up from 2 million when Ofcom came into being.

Three in four homes and businesses are connected. Nearly two-thirds of these fixed connections (12 million) are provided by BT's competitors.

We have a diverse range of copper based services, with improving levels of performance. The average speed is now around 8Mbps, compared with a maximum of 512kbps in 2003.

Superfast broadband is now becoming widely available.

Virgin Media has upgraded its cable network to DOCSIS3, offering speeds up to 100Mbps within its footprint, as well as trialling a 200Mpbs residential service and a 1.5Gbps service in Old Street, London.

BT is upgrading its access network using a combination of FTTC and FTTH with an ambition to reach 66% of the country by 2014.

BT's fibre to the cabinet speeds are also increasing as a result of changes to the access network frequency plan. FTTC speeds will rise for many customers - up to 80Mbps during next year. And with technologies such as vectoring, these higher speeds should be maintained as take-up increases, and may be improved still further.

This is good progress and, by any account, positions us well in relation to comparable European countries, if not among the world's most advanced connected societies, many of whom have considerable economic and geographical advantages in relation to the cost of deployment.

The challenge of risk and uncertainty

However, for superfast broadband, subscriber numbers are still low, perhaps because the nearest thing we have found to a 'killer app' so far is the demands of the multi-user household.

Amid a cornucopia of entertainment and information services, and the promise of advanced telemetry, e-health and interactive education, it is interesting that the only 'killer app' we have so far is the presence of teenage children.

Social networking, streaming and sharing from the teenage bedroom, leading to local contention, the victim of which is the person typically paying the bill, seems to be among the strongest reasons for adopting superfast broadband.

But as an approach to promoting superfast broadband take up, 'having more teenage children' seems a little long term, and a little distant from reality.

So there remains uncertainty within the sector about the level of demand for superfast services and of course willingness to pay. Yet it is clear that consumption of data continues to surge.

Figures we published as part of our first ever Infrastructure report last week showed that residential fixed broadband customers are using 17GB of data each month. London Internet Exchange data also shows that traffic over its network routers has increased seven fold in the past five years.

This trajectory seems certain to feed through to a demand for increased speed - the fact that we cannot identify specific 'killer apps' beyond bandwidth hungry teenagers is in some ways beside the point.

The industry recognises this almost inexorable process and the transition from copper to fibre is therefore already well underway.

This fundamental network upgrade involves a very different challenge to that faced at the introduction of current generation broadband - an innovation at the heart of what future technological archaeologists may one day call the ‘late copper era’.

For current generation broadband, the overall scale of investment was in the low hundreds of millions - important, but comparatively modest, capital expenditure.

The cost of building super-fast broadband networks by contrast will be several billion in order to reach a good majority of the population - with many more billions to take coverage to what many regard as the logical conclusion of near universal fibre to the premise.

In contrast to the era of 'late copper' the 'early fibre' era involves considerably greater risk for investors and companies, as well as for bigger calls for governments and regulators.

Compared to the transformation wrought by moving from dial up to basic broadband using an existing copper network, there is considerably more uncertainty in almost every respect. This includes

  • alternative deployment technologies;
  • consumer demand and willingness to pay;
  • capital expenditure and deployment costs;
  • end user equipment;
  • and of course in the nature of competition.

Regulating with more risk and uncertainty

In the face of such uncertainty and risk, we need a more flexible and a more long-term strategy than the era of 'late copper' required.

We need an approach that provides clarity and focus for the near term when important decisions have to be made, while recognising that we will also have to respond to how different technologies fare and to how the market and competition develops.

We must accept that this phase of development is different to its predecessor and, most fundamentally, we must recognise that we need to aim to combine essential investment with effective competition.

Promoting investment and competition

Our role is to be as consistent as we can be and by doing so create a climate which supports a significant degree of predictability in our approach. If successful, this will promote a climate for private sector investment and sustainable long term competition.

Since 2005 Ofcom has cleaved to the clear strategic principle that regulation should promote competition at the deepest level of infrastructure at which it will be effective and sustainable.

We believe that this remains the right approach for the transition from copper to fibre.

It also underpins our practical approach:

First, identifying the critical interventions that would support both investment and competition in the immediate period. Physical infrastructure access and virtual unbundled local access are the products of this approach;

Secondly, responding to developments that enable technical innovation, further investment and which will help maintain effective competition in the present and longer term.

Recent examples of this include our support for changes to the access network frequency plan to enable faster FTTC speeds and in promoting 'wires only' wholesale products. In the longer term our plans to support standardisation of wavelength division multiplexing on PONs would fall into the same category.

Thirdly, complementing this with selective interventions to promote competition in the context of converging technologies and services. Our work to open access to premium content, to make switching easier and to provide reliable information to consumers are examples of this approach.

I would like to talk a little more about some of these areas.

Today Local Loop Unbundling is the primary basis of competition in current generation 'late copper' services. Industry has invested in infrastructure and we have focused on functional separation in BT to support this form of competition.

We expect that to continue over the next few years.

For superfast broadband the picture is subtly different. Our strategic principle, when applied, does not lead us to conclude that competition based on multiple providers deploying their own fibre networks in the same area will happen, or that it would be effective and sustainable.

Industry players seem to share this view.

The next deepest level of competition is therefore through a bit stream type remedy - although one that is far superior to its 'late copper' antecedents.

Virtual unbundled local access - VULA

VULA allows other communications providers to use BT's new fibre networks, between the end user and the local exchange, so that they can compete in the supply of downstream superfast services.

We are looking to open up as much of the network, control and ultimately value as possible to competition in order to replicate many of the LLU-driven benefits. VULA must therefore meet a number of key requirements, such as local physical interconnection, service agnostic un-contended capacity and place the maximum control possible in the hands of customer CPs.

However, we have allowed BT flexibility over the way it sets its prices for VULA. We’ve done this for two reasons:

First, the investment case for NGA is uncertain and risky. Heavy-handed regulation could have serious effects in such a nascent market.

Second, while consumers might be prepared to pay a premium for superfast, we believed that the price for superfast services would be constrained by current generation services - whose prices are either subject to competitive pressures or regulated.

Indeed, looking at the current retail prices and consumer demand this constraint appears to be even more effective than we imagined.

In order to establish a level playing field for superfast services, we have required BT to offer VULA products on the same basis to all providers (including its own downstream divisions).

And we have set out clearly the approach we would take in relation to any potential margin squeeze.

This, as I have noted, is a different phase in investment and competition. So we can't guarantee the same or bigger margins than those enjoyed today - nor would it be right to try to do so. But, we will ensure that an efficient operator is able to compete in the downstream market.

Competing using VULA

We recognise the challenges of providing superfast broadband. The costs of fibre deployment are substantial but, as yet, consumers appear to be unwilling to pay very much of a premium for the service it delivers.  This puts pressure on the wholesale-to-retail margins.

The answer to this is unlikely to be to put retail prices up - that would simply reduce demand. At a time when disposable incomes are being squeezed generally, consumers are more demanding about value for money not less.

The answer is more likely to lie in rising to this challenge by developing new services or combinations of services that exploit the higher bandwidth of fibre and so improve the overall value proposition to consumers.

Industry has to innovate and deliver a service that consumers' value at the price point that works for consumers, rather than one which works for the industry.

We would expect this to involve not just speed, but devices, services and applications; a bundled proposition which exploits different competitive advantages and seeks out different competitive positions.

Fundamentally, it's not about regulation, it’s about innovation.

In parallel, we hope that companies will compete to be more efficient to deliver greater value. This will mean looking for ways to reduce costs.

A couple of examples illustrate the point:

There should be scope for alternative installation processes where the provider doesn't need to rely on Openreach for the consumer installation - often referred to as wires-only. This is the kind of thing we would like to encourage.

In the longer term there is also scope for a simpler network configuration where voice and broadband are delivered over a single converged network, rather than having a separate network for each.

It’s possible that cost could be removed if fibre broadband were provided on its own with voice over the top.

This would of course be a significant change for both BT and competitive operators alike and would have to be considered very carefully.

The final third

VULA is only an option where BT has deployed FTTC or FTTH. It can’t be the primary answer outside the two thirds of the UK where BT plans to deploy NGA.

Given that commercial deployments will only cover two thirds of the UK, the government plans to invest £530m to ensure superfast broadband coverage is extended to as much of the country as possible.

It estimates that this central funding, with additional local funding and commercial investment, should extend NGA coverage to 90% of the population.

In the final 10% of the UK, where neither commercial or government funding is currently available for NGA deployment, a creative mix of technologies is likely to be needed to deliver NGA services in due course.

This policy is mainly for government, but we have worked with the government to identify some key criteria to ensure that public investment is the friend of competition and not inadvertently its enemy.

We believe that open access must be provided as part of any publicly financed initiative, together with interoperable standard systems which maximise the potential for effective downstream competition.

The danger otherwise is that we end up with single vertically integrated providers in a range of areas and an absence of choice, innovation and competition. This would be a particularly egregious outcome if delivered by virtue of taxpayers money rather than private risk capital.

Duct and pole access

Given our focus on VULA it stands to reason that we do not expect that duct and pole access - PIA as it is known - will be used as a main basis for competition in areas susceptible to purely commercial deployment.

Because BT and other providers will compete with one another on an ongoing basis using VULA, we have required it to be provided on an equivalent basis.

Conversely, because we do not expect BT and others to compete with one another on an ongoing basis using PIA, the cost associated with also requiring this product to be provided on a strictly equivalent basis is hard to justify, and we have not sought to do so.

Nevertheless, PIA is very important as it allows other providers to compete to invest in superfast broadband in areas where VULA is not available.

In areas where BT has no commercial plans to invest, PIA will allow other providers to bid for government funding in competition with one another and BT.

These remedies have the potential to improve the contestability of the government's procurement process.

BT published a draft offer for access to its ducts and poles in January, followed by a period of industry trials. This led BT to publishing a revised offer in October, reflecting a number of product developments, and featuring substantially lower prices to those originally proposed.

These new lower prices are now available. Depending on your precise calculation method, the new prices are between 30% and 50% lower than the original prices. They are among the lowest, if not the lowest, for comparable products in Europe.

So a big step forward has been taken from a regulatory and product perspective. Of course there is more to do. But the basis for a competitive procurement seems to be in place and we very much hope that this is the next step in this important process.

The evolution of regulation

We have focused on VULA for the commercial areas of deployment and on PIA for the subsidised areas (in combination with our open access and system interoperability criteria).

We have done this on the basis of industry input to our consultations on NGA. We have responded to what industry told us mattered most and where they believed investment could be financed and where competition could be sustained.

In line with my theme of a degree of uncertainty in the new fibre era, these assumptions may change.

We may see deployment plans change and players turn their interest to sub-loop unbundling, or SLU. It's worth noting that this product has been available in the UK for over 10 years, and remains available today. But demand for SLU, both in theory and in practice, has remained extremely marginal throughout the period in which we have been considering these questions.

Similarly, it is possible that PIA could become relevant within the commercial areas of NGA deployment in due course. Our analysis suggests that this is unlikely at present. But once again, the uncertainty means that we will retain an open mind to the approach we have taken and to what might develop in the future.

We will want to assess such questions alongside alternative approaches which might offer even more radical possibilities in the longer term. These might include the idea of a full 'fibre migration' alongside anchor pricing to protect consumers, for example, which we have floated in previous consultations but which has not yet attracted significant support. But its time may yet come.

Competition amid converging technology and services

Finally, we need to make sure that we create a wider environment that supports investment and effective competition. To do this means understanding that competition is changing fast, driven by consolidation, convergence and complexity.

We have seen consolidation in the fixed and mobile sector in recent years. Most players have established substantial customer bases. Most are intensely concerned about churn and would like consumers to be as sticky as possible.

But for consumers, it's very important that enough consumers do churn; that enough consumers exercise their choice to extract the benefits for all consumers from the disciplines of competition.

We start with the uncontroversial idea that for competition to be effective consumers must be able to switch fairly easily between providers.

With this at the forefront of our minds, we embarked upon a major project examining the switching processes between communications services. We wanted to examine not only how well they deliver for consumers and competition today but how they are able to cope with the increasing trend towards bundles, the development of new services and the evolving technologies of the future.

Our current long term view is that the 'gaining provider' leading the process of switching will result in greater consumer and competition benefits than the alternatives.

Our focus, at this stage of the review, is on fixed voice and broadband services where we know that there are a number of significant concerns, including the mix of different processes, concerns about the reliability of switching and loss of services, concerns about competitive neutrality and in relation to the level of consumer protection in place.

We intend to publish proposals on this early next year.

In a converged world competition is about the bundle of services offered, not just bandwidth but also content.

This is one of the reasons why we believe Sky should provide wholesale access to its premium sports content and why we have fought hard for this to happen over the last four years.

The next step in this debate rests with the Competition Appeals Tribunal and we look forward to their conclusions, but I have little doubt that this will not be the end of the story.

In addition, we welcome the Competition Commission's views on our reference to the CC in relation to premium movie content, where again, in our view there are genuine concerns that stand in the way of establishing fair and effective competition and by extension are contrary to the consumer interest.

We are delighted that the Competition Commission’s provisional analysis appears to be consistent with our central concerns and we look forward to the next steps which focus on identifying the appropriate action to address these concerns.

The third set of measures are those designed to help the consumer battle the complexity often found in modern communications services.

The most obvious example is our now well established work on broadband speeds. This ground-breaking work has enabled consumers to understand more clearly exactly what they are purchasing and to make more informed choices between different providers and different services.

It remains the most popular of any Ofcom report or analysis in our history.

The longer term

For now, outside of the network competition between Virgin and BT and the alternative delivery systems of mobile and satellite, for regulatory purposes VULA is the deepest point in the network where effective and sustainable competition can be established.

If, however, technology and economics conspire to offer alternatives, then we will revisit this conclusion.

One particular area that we are keeping a close eye on is wavelength unbundling, which we see as a potential long-term option.

Currently, FTTH Passive Optical Networks use a single frequency, or wavelength of light, to carry data.

However, multiple wavelengths can be carried through a single optical fibre simultaneously, boosting the capacity of that fibre massively to multiple tens of gigabits per second.

This technology, known as wavelength division multiplexing, is already used in the core network but as the economics improve, and with the right equipment at the head-end and in customer'’ premises, it is expected to move into the access network also.

With these elements in place the potential may exist for a competition remedy - unbundling at the wavelength level - in which different communications providers address individual customers each using a different wavelength.

The result could be competition among providers offering differentiated services on the same single piece of fibre, each able to offer far greater speeds than on offer at present.

We are tracking and assessing the progress of developments and we will continue to engage with operators and standards communities.

 However, we envisage that we may need to take the initiative, with others, to help drive standards in order to promote potentially attractive forms of competition in the longer term.

We need to think about these questions on an EU-wide basis and beyond. Ofcom is working closely with the European Commission and with BEREC (the Body of European Regulators for Electronic Communications), of which we are a member. We aim to ensure that regulation is as adaptable, flexible and innovative as the broadband technologies that are being deployed.

In fact, many in the industry regard FTTH and GPON as solutions that feel more durable than FTTC. But that doesn’t mean FTTC isn’t right for now - nor that there is any easy fix to the regulatory and competitive challenges that go with it.

I would add, in this context, that while I have emphasised the challenges of investment and of new technologies, we should be clear that there is no future in which we will not support vibrant and effective competition.

It is not for us to determine the particular outcome of that competition - only to ensure that there is such competition. We are clear about this task. It is a task we regard as a long term commitment - one born of our duties and one which permeates the attitude and objectives of everyone within our organisation.


In summary, we know the direction in which we are headed and what kind of outcomes we are seeking to achieve.

The demand for more and more capacity to carry more and more data, both on mobile and fixed platforms seems inexorable.

But the precise route to widespread, innovative and competitively provided superfast broadband services cannot be precisely prescribed. The variables are too many, the market, technologies and consumer demand too uncertain.

Ofcom's strategy needs to reflect this dynamism and this uncertainty.

We will aim to provide clarity and focus wherever we can, and flexibility and an ability to respond to changing circumstances where we cannot.

It is a careful and delicate balancing act, but one that reflects the major shift from competition in the era of 'late copper' to investment and competition in a new 'age of fibre' in the access network.

In concrete terms, this adds up to four critical elements, all now either in place or well underway. It means:

  • promoting competition at the deepest possible level of the infrastructure at which it is effective and sustainable;
  • selectively identifying the key interventions that will support both investment and competition in the immediate period; VULA and PIA
  • being flexible and ready to respond to developments which will enable longer term innovation, investment and alternative forms of competition; and
  • taking a wider view through selective interventions that promote competition in a world characterised by consolidation, convergence and complexity.

We will remain open to new ideas and fresh thinking at all times. If the facts change, then we will change our minds but for now, this is our approach.