The NBN competition conundrum

Communications Minister Malcolm Turnbull has been banging the infrastructure-based competition drum for a while, and Professor Fred Hilmer (Vice-Chancellor of the University of New South Wales) recently joined forces with the minister in pressing the case for greater fixed-line telecommunications infrastructure competition.

Theoretically that’s a noble aim but the reality is rather different.

First of all, there is already a good level of competition in this market, between fixed and mobile telecommunications infrastructure based providers.

• most of today’s telephone traffic takes place over mobile networks rather than a fixed network.
• a large chunk of broadband traffic is also currently taking place over these mobile networks.
• add WiFi networks to that, as they are often operated by others in the market, and this widens the area of infrastructure-based competition.
• over-the-top (OTT) companies such as Google, Facebook, Skype, Amazon and many others now offer, over their sophisticated ICT infrastructure, all of the traditional telecoms services in competition with the traditional players; and they also offer cloud computing, data centres, data analytics and M2M service.

So it’s important to remember that there’s already healthy infrastructure-based competition between telecoms companies.

Forty years too late

In general terms pure infrastructure competition only works if the competing infrastructure is adding value to what another set of infrastructure cannot deliver. Obviously this is very clear in the case of mobile versus fixed infrastructure competition. However, Minister Turnbull and Professor Hilmer seem to be arguing for competition between fixed networks and, as we saw in the 1990s, this is rather more complicated.

Australia missed the boat in the early 1980s when Kerry Packer successfully lobbied the then Coalition government – during the Davidson Inquiry – to place a moratorium on the building of new HFC infrastructure. This policy was only reversed in the mid-1990s – far too late for this model to deliver economically-viable fixed infrastructure competition.

When, after the end of the moratorium, Optus announced its plans to roll out HFC in competition with Telstra’s DSL network that strategy clearly fitted the competition arguments above – this infrastructure offered better broadband than Telstra’s copper-based network.

But what happened was that Telstra overbuilt 90 per cent of Optus’s network with a similar HFC infrastructure (not adding any additional value). As a consequence we ended up with only 25 per cent penetration and an investment write-off of around $US4 billion ($4.24 billion).

As mentioned, Australia had by that time already missed the boat on fixed infrastructure-based competition. During the 1970s and 1980s Europe and North America built their HFC networks, which indeed delivered excellent infrastructure-based competition to the incumbent telcos. Today we are 40 years down the track – or 40 years behind – and trying to fix this now is simply impossible.

At the same time, both Telstra and Optus ultimately favoured DSL technologies for their broadband services and the HFC networks were not further developed to any significant degree.

So much for the noble theory of pure infrastructure-based competition.

Who is going to invest in competing fixed infrastructure?

Infrastructure investment is an expensive business, with long-term returns only. For that reason competitive infrastructure – if it does indeed add value over and above other infrastructure – will only be deployed in commercially viable areas. So, under the right regulatory circumstances (avoiding events like the HFC disaster), fixed-based infrastructure competition could possibly work in some of those areas.

But in all reality there is very little evidence of new national, or at least large-scale, infrastructure-based competition activities anywhere in the world. We are 40 years too late for the HFC versus copper model. Nearly all of today’s new fixed telecoms infrastructure investments are now based on fibre networks.

In 2006, Telstra indicated that it could overbuild some of its own old infrastructure with new infrastructure in order to add value. It could commercially deploy Fibre to the Node (FttN) to approximately 50 per cent of the population, based on its vertically-integrated monopoly model that it was able to operate at that time. There was no room for any effective competition in their model – as a matter of fact the company wanted a 13-year regulatory holiday in exchange for rolling out this FttN infrastructure.

The engineers in the industry all agree that in relation to fixed telecoms infrastructure both HFC and copper-based networks will eventually evolve into Fibre to the Premises (FttP) networks. Players in this market are extremely wary of new forms of infrastructure-based competition, since at some point they will no longer be able to compete on a value-added basis, as everybody will eventually operate FttP.

So, like so many other infrastructures, it becomes a regulated monopoly, whether politicians or academics like it or not. Investors are very much aware that it does not make economic sense to run two or more national FttP networks to even the wealthiest customers in the market.

That’s why nobody in the world is offering fixed infrastructure on a national scale in competition with the incumbent fixed network operators – not in North America, Europe or Asia.

What you see is either national operators building these networks (in which case it remains in effect a monopoly) or you see cherry-picking, as is the case in the USA. There are no instances of any large-scale FttP infrastructure-based competition models.

Where are the competition models?

It’s easy to shout ‘competition’ from a great political or academic height, but it’s much more difficult to give it a serious commercial foundation. Perhaps this is why those advocating this level of competition do not come up with ideas or suggestions to show how it could work.

If we follow an open infrastructure competition policy we will have to find another solution for the other 50 per cent of the population – assuming that both Professor Hilmer and Minister Turnbull agree that broadband access has to be a universal service.

In the future more essential services such as healthcare, education, government services, smart energy services and business services will be delivered over that infrastructure, with the assistance of cloud computing, M2M and Big Data, to everybody in the country – this cannot just be given to the 50 per cent who will be serviced by the commercial models.

Furthermore they will have to come up with good reasons for investors to invest – beyond cherry-picking – in competing fixed line infrastructure, as current HFC and copper infrastructure will all eventually end up in vanilla fibre networks, with no economically-viable business model for two or more of those competing FttP networks.

So far there are no investors chasing investment models based on a level of infrastructure competition such as this.

In all reality the final outcome of fixed line telecoms will be FttP, and that will be a utility – be it in five, 10 or 20 years. Investors base their fixed infrastructure plans on 20- to 30-year investment cycles and will therefore be wary of these developments, as they can very easily ruin their investment.

Competition policy should be focused on the services, not the infrastructure

But to those advocating competition, the real value – rather than the political or dogmatic value – is that for those using it the infrastructure itself is not important. They are interested in the services that are delivered over it, and this is where competition policies should be focused.

Now, it can be argued that infrastructure development can be done differently – not through cross-subsidies. I have no issue with that but if community leaders argue for infrastructure-based competition they will, at the same time, have to indicate:

• how to service the non-commercial parts of the country; and
• how to safeguard very expensive competitive infrastructure investments from becoming write-offs – as happened in the case of the Australian HFC debacle.

So far – as with all other infrastructure in this country, whether electricity, water, roads, hospitals or schools – we use cross-subsidy models to provide a universal service.

It's easy to argue for infrastructure-based competition, but it is not as easy to turn that into a social economic policy. Ignoring the other half of the country where this level of infrastructure competition will not work without addressing the perilous investment issues that are involved in such a competition model.

This is an edited version of a post originally published on October 8. Paul Budde is the managing director of BuddeComm, an independent telecommunications research and consultancy company, which includes 45 national and international researchers in 15 countries.