Home & Office

Finding a funder for fibre to the home

The Communications Management Association's David Harrington argues there's a case for using government funds to roll out fibre to the home across the UK
Written by Richard Thurston, Contributor

As bandwidth demands continue to increase, the requirement for fibre-optic internet connections has risen up the regulatory agenda. Ofcom is now making obvious signals that it wants the issue investigated and the government is tackling the subject through the Caio Review.

At the centre of the debate is the Communications Management Association (CMA), which is representing businesses' IT interests through its regulatory forum leader, David Harrington.

A veteran of the telecoms industry, Harrington has seen vast changes in the regulatory scene, from the dismantling of BT's monopoly to the transformation of Oftel into Ofcom.

Speaking exclusively to ZDNet.co.uk, Harrington explained why providing a fibre-optic connection to every UK home and business is essential; how Ofcom is finally coming round to supporting the idea, having suffered a "lack of thought leadership"; and why European commissioner for information society and media Viviane Reding is simply "the greatest".

The need for fibre-based broadband access is one of the most hotly debated regulatory issues at the moment. Why do we need fibre to the home (FTTH)? Copper is pretty fast these days.
Copper provides at most 24Mbps — if you live in the exchange — and it's not symmetrical. Although services that require over 10Mbps are not yet here, they are coming. There is an ever-growing demand for higher and higher bandwidth. Viviane Reding said it's the most important topic in the telecoms world today. The challenge is where the investment comes from and how it's funded. How do we regulate when we can't apply yesterday's thinking?

Equipping the UK with nationwide fibre to the home would cost around £15bn. That's a lot of money, which few companies would be prepared to invest. Who would do it?
That's the issue. Openreach has the Ebbsfleet [fibre-to-the-home] contract. The most likely company is BT. No-one else has the reach. The thing is, BT won't [invest] at the moment because it has enough problems with 21CN [its £10bn backbone network upgrade]. The last thing they need is another £15bn [investment].

What about Virgin? It already has a network covering half of all homes.
Yes, but they are half way not there.

Where is the business model? It's dangerous to invest £15bn with uncertain returns.
Ed Richards [Ofcom's chief executive] said a couple of weeks ago that it could be that telecoms is one of those areas where "build it and they will come" [as a model] might work. That is a sea change in Ofcom's thinking.

It's interesting that you mention Ofcom. The regulator issued a consultation document on fibre access late last year. The CMA responded quite critically at the time.
It was a terrible document. It completely lacked thought leadership. From an economist's point of view, it was probably just, but that's all I could say of it. But Ed Richards' speech was far more upbeat: "We are going to do something about this."

But £15bn is a lot of money to invest.
Well, £15bn is the cost of Crossrail and, arguably, the cost of the 2012 Olympics. If you put a similar amount into fibre, think of the longer-lasting impact, compared to the other two projects.

The other two projects are backed by taxpayers' money. Should the government help pay for FTTH?
It has national benefits, so it needs a national investment. Why should we entrust it to a private company and go hands-off at the government level? There is a case for using government money.

Do you think taxpayers will find it acceptable?

So, if it is too expensive and the returns are too uncertain for a private company to invest, and it's unpalatable for the taxpayer, how could it happen?
I would be a rich man if I had an answer. That's the whole point of what Francesco Caio is doing: how to find an answer to this conundrum [Caio, former chief executive of Cable & Wireless, is investigating the barriers to next-generation networks on behalf of the government]. My best guess is that we will see islands of fibre. Look at Amsterdam, where the council has entered into a joint venture with private money [to roll out fibre in the city]. If we had the same here, we could have islands of fibre in cities, then run it out into some of the communities. It seems a logical explanation of what is likely to happen. Some could be operated by different operators. It could be allowed [by regulators] as long as there is open access [for competitors to sell services using the infrastructure].

Do you have a preference for the type of architecture used? Openreach's Ebbsfleet deployment uses a PON-based approach (passive optical networking), but it could be done using point-to-point.
Because point-to-point has an electrical component at the switch, it is more amenable to [sharing the network with] other providers. There is some concern, based on present knowledge of Ebbsfleet, that other service providers might not be able to...

...deploy services to their customers because of the technology used. Point-to-point does have virtually unlimited bandwidth, although it's 30 percent more expensive than PON.

There has been much debate over the universal service obligation (USO), which obliges BT to provide telephony services to any UK resident who wants them. Some people argue that the USO should be extended to include broadband. Should it include broadband, in your opinion?
At some point, yes. But it comes to the issue of: "Who?". If it goes to BT, you're enforcing a monopoly. My personal view is there should be a universal fund for anyone that wants to deploy the infrastructure.

How would a universal fund work?
All players in the industry would contribute to it. That fund could be allocated by the government according to their rules. It could be auctioned off by geographic severage.

It might be difficult to split the USO between multiple carriers though.
Don't forget that we already have the case of Hull, where the USO is carried by Kingston, not BT.

When would you want this extension to the USO to be in place?
The sooner we start, the better. If we wanted a broadband infrastructure by 2012, we should have started last year.

You hinted earlier that there are problems with BT's own network, 21CN. Are you satisfied with progress?
No. Nobody is. It's delayed by 18 months to two years. I understand there are technical difficulties. I understand [BT's chief executive] Ben Verwaayen was told that it wouldn't be done before 2014. But he said it must be done by 2011, in time for the Olympics. So huge pressure was put on the technical staff, against their advice.

So you think BT's [already delayed] completion deadline of 2011 will slip?
Almost definitely. It's a hugely expensive programme that's almost two years out-of-date. There is bound to be some impact on the end date. Whether BT can recover lost time — I wouldn't put my shirt on it.

The CMA's chief executive, Glenn Powell, told me a couple of years ago that BT was poor at communicating its plans for 21CN, meaning businesses were being kept in the dark on how they would be affected. Has that situation changed?
That was so 18 to 24 months ago, but not now. They need to talk to us so enterprises can plan ahead and put money in their own budgets.

Do you think 21CN will deliver the benefits BT promised?
I think so. They [will] have an end-to-end IP network. It's bound to encourage the development of IP-based applications. And it will have carrier-grade quality of service (QoS), which is important.

One of the more important regulatory events this decade was the functional separation of BT, which created Openreach. While the theory behind Openreach was largely welcomed, carriers argued vehemently for a long time afterwards. Are you satisfied with Openreach?
It's a darn sight better than anything we had before, but it's not perfect. The imposition of SLAs and SLGs [service-level agreements and service-level guarantees] by Ofcom will help. That has been too long coming. That is just an example of how they're improving. From our point of view, functional separation has been a huge success. The EU has now embraced the concept. One of the arguments against it is: it inhibits investments. But it hasn't inhibited investment here.

Do you think we finally have equivalence of treatment between service providers?
Equivalence now is 100 percent or 1,000 percent better than anything we had before [Ofcom's] strategic review. If you'd talked to BT's competition, they would have given you a yard-long list of things not done. If you were to ask now, the list would only have a dozen things on it.

Ofcom celebrates its fifth birthday this year. Would you say it has been a success story?
Undoubtedly. For sure.

What about its latest chief executive, Ed Richards?
Ed Richards has got his own agenda. He comes from the entertainment sector, rather than the telecoms sector. One sometimes wonders if there are enough senior posts in Ofcom filled by people from telecoms, rather than the media.

Isn't that just a result of Ofcom covering five disciplines? Telecoms is only one.
The question is: is the balance right? Stephen Carter was the last visible telecoms guy. That's not a criticism, just a comment on the way things are going. Ofcom has been a damn sight better than Oftel ever was.

There have been proposals made by the European Commission for one single super-regulator for Europe. Would you be supportive of a super-regulator?
I think it's absolutely essential. A new regulator at the centre will have more teeth. We cannot go on as we are. Effectively there is no single market in telecoms and that means that European enterprises are disadvantaged in the global market. ABN Amro told the Gartner summit in April that they would save huge amounts of money if there was a single market. One Italian car manufacturer has said it has 872 contracts to support its pan-EU telecoms needs. That's ridiculous.

Do you think the regulatory set-up is equal across European countries?
Some countries are good; some countries are not. Some regulators are not independent of the state, like Poland, for example. Some countries' governments have control of the incumbent [telco]. Some have inadequate competition rules or don't impose equal-access obligations.

Is everyone supportive of the need for a European super-regulator?
Incumbents have said: "You can't do this". Most national regulatory authorities, including Ofcom, see this as a threat to their autonomy and oppose any extension of centralised authority.

How would such an organisation be created?
There is a huge debate over what form it would take. The solution is very likely to be a development of the existing European Regulators Group, although it's by no means clear just how it will be funded and how independent it will be from political interference.

One of the other big European issues is roaming. Call costs have fallen dramatically, following the demands of Viviane Reding. As a representative body for business telecoms users, you must be pleased with this.
It's great; absolutely terrific. It took us nine years of badgering.

Presumably you are supportive of Reding?
She is the greatest. This really chimes with what the CMA are doing.

Do you think data roaming prices need to fall?
Unless prices come down by July, [Reding] will legislate, and she means it. These stories of £10,000 bills are crazy. Even national downloads are crazy at the moment.

Editorial standards