Speech to Broadband World Forum Asia - Kuala Lumpur

May 15, 2012
Communications & Broadband

How We Got Here

Australia’s National Broadband Network is unique in the world.

The current Labor government has established a new fixed line monopoly which is intended to provide a fibre to the premises customer access network to 93% of Australia’s households, the balance in the more remote areas being served by fixed wireless and satellite.

The project is unique in at least three respects.

First, it involves the creation of a new Government telecommunications utility which is designed to be the only fixed line customer access network to Australian households. To that end, the owners of existing HFC networks, Telstra and Optus, are being paid billions not to use their networks for either broadband or voice after the NBN is deployed in their footprint. Furthermore various regulatory and legislative measures have been enacted to make it practically impossible for any new entrants to challenge the fixed line monopoly.

Second, the amount of money proposed to be spent on the NBN far exceeds the amount committed by any other Government. The Government claims this project will cost over $40 billion to complete – but it is early days and many experienced people believe the cost is likely to be much more than that. One estimate puts it between $60 and $80 billion. I note that Australia’s population is 22 million with around 12 million household and business premises eligible for connection.

Third, the project involves an extension of fibre to the premises to a much larger percentage of the population in any comparable country.

Not only is the current Australian Government’s approach to broadband at odds with broadband policies everywhere else in the world, it is also at odds with the broad arc of economic reform over the past thirty years or so.

Let’s rewind thirty years. Mrs Thatcher was in Number 10 and Ronald Reagan was settling into the White House, but most governments still saw telecommunications as both a natural monopoly and a service appropriately delivered by the Government..

Even in countries where networks were privately owned, they were heavily regulated. In 1980 the US government was six years into its anti-trust investigation of AT&T but still two years away from the consent decree that would break it into eight smaller companies.

Of course, over the next two decades everything changed: first with the rapid growth in the US of cable and satellite networks, then exponential growth for wireless, and finally the emergence of the Internet as a mass medium Governments and carriers embraced the economic liberalism of the times, particularly that era’s ‘triple play’ – private ownership, market deregulation and the enabling and indeed promotion of competition.

According to the World Bank-Stanford telecommunication database, in 1980 only 2 per cent of telecommunications carriers in 167 countries were privately owned. By 1998 that number had increased to 42 per cent[1]. This period of reform wasn’t just an ideological fad. It rapidly generated large increases in the volume and variety of services produced by the industry, in economic productivity, and in consumer welfare.[2]

A multi-country study by the OECD found that actual and even prospective competition generated lower prices, higher productivity and better quality in telecommunications.[3]

Of course, industry restructuring and a reduced public sector role were not always managed as well as they could have been, and missteps of some sort can be identified in many jurisdictions. In Australia at least three errors were made, and both sides of politics contributed to them:

  • First by allowing the Government domestic PTT, then called Telecom, to merge with the Government owned international carrier, OTC, to form a single dominant telco, Telstra. ;
  • Then, when cable television arrived, by permitting Telstra to deploy a pay TV cable network in direct competition with a similar rollout by its sole licensed competitor, Optus Vision.
  • And finally by selling Telstra as a vertically integrated operation without effectively separating its customer access network from its retail business.

The third error, selling Telstra to the public as a vertically integrated telco, is widely cited as the worst mistake, but with the benefit of hindsight I think the second mistake was the worst. Wherever HFC networks have been owned other than by the traditional telephone companies we are seeing real facilities based competition which has resulted in the telcos moving to upgrade their copper networks to enable the high speed broadband capable of competing with the cable companies in providing a triple play of video, broadband data and voice.

Nonetheless, until a few years ago the Australian telecoms scene was not dramatically different to that in many other countries – a highly competitive wireless market and a fixed line market dominated by the incumbent PTT with constant battles about the appropriate terms of access to its legacy infrastructure.

Following a failure to implement its original, and by comparison modest (at $4.7 billion) broadband policy, the Labor Government in 2009 embarked on the extraordinary NBN policy I have described.

It has justified this by reference to three goals:

  • The structural separation of Telstra’s customer access network from its retail business – however it achieves this not requiring a separation of that network but rather by overbuilding it at public expense and then paying its owner, Telstra, not to use it any more.
  • The upgrade of Australia’s fixed line broadband services – however instead of using a mix of technologies adapted to local circumstances, it has gone for the most expensive possible solution – near universal FTTP.
  • Finally, the provision of broadband services by fixed wireless and satellite to the more remote 7% of the population – however it is doing this by building an entirely separate fixed wireless network and acquiring itself two Ka band satellites. A more rational approach would have been to contract with existing wireless and satellite providers to deliver the services.

The Situation in Asia

During the past two years as Shadow Broadband and Communications Minister, I have visited and carefully studied broadband policy and deployments in Singapore, New Zealand, Hong Kong, Korea, China today in Malaysia as well as the United States, the UK and Germany.

The key themes of Malaysia’s policies are:

  • Promotion of facilities based competition. Unlike in Australia the Malaysian Government is not seeking to establish a monopoly service provider in the fixed-broadband market. Although still the largest player in the market Telekom Malaysia faces competition from other fixed-broadband network operators including Time dotCom – and more network operators could still arrive in the market.
  • Fixed-Wireless players are promoting competition The government here has also used fixed-wireless access to extend coverage and provide competition for fixed-broadband operators with WiMAX operators like Packet One and YTL already moving towards nationwide networks.
  • Technological agnosticism. Among technologists, there is an evangelism when it comes to fibre as the end-point solution for any broadband infrastructure. However, telecos everywhere struggle to persuade consumers to pay a premium for superfast broadband and so appropriately Malaysia is promoting a range of technologies including FTTP, DSL, and wireless where it is economically rational to do so. .
  • Reliance and partnership with the private sector where possible. Malaysia’s HSBB initiative is a 30 per cent public/ 70 per cent private partnership between the Government and Telekom Malaysia, which will be rolled out to 1.3 million homes in priority economic regions (primarily the Klang valley). Although the network covers only 20% of total homes there is now talk about extending it to other parts of the country, such has been the success of the initial deployment.
  • And finally, a push to lower the price of retail broadband, recognising that price sensitivity is one of the key hindrances of increasing broadband take-up[4].

The Situation in Australia

While boosters of the National Broadband Network in Australia claim it is an elegant solution solving all of the policy mistakes of the past 30 years and delivering a first-class infrastructure, it is important to note how many trade-offs the Labor Government has made in getting there – not least because we don’t want to be here in 30 years’ time trying to figure out how to overcome an even bigger series of challenges.

These trade offs include:

  • The scrapping of infrastructure based competition. Under highly controversial anti-cherry picking laws which the Coalition opposed, the Labor Government outlawed the building of superfast networks, unless it is as a wholesale-only, layer-2 bitstream service built to NBN specifications. This policy decision has been particularly egregious in greenfields areas, where there was a thriving and competitive industry already delivering superfast infrastructure. We agreed with the Government’s policy goal that fibre should be deployed to newly built premises – but argued that the competitive industry structure should be retained to the greatest extent possible. And yet, the NBN is paying subcontractor Fujitsu to act as a ‘provider of last resort’, offering developers to connect premises for no upfront costs, other than that of the pit and pipe, making it in reality the provider of first resort. That caused a huge backlog of orders as developers naturally sought to take advantage of the Government’s offer – and yet the NBN has delivered just over 100 connections from over 100,000 applications received in the past year[5].
  • Upward Pressure on Prices. Instead of a competitive market the model the Government has chosen for the NBN is a regulated monopoly. Although the regulatory framework for the NBN is still being set, it is clear from access-seekers currently negotiating over the regime that the NBN has been given wide ranging powers to set prices[6]. The NBN has designed a complex pricing model which will require retail service providers to pay for both Access Virtual Circuits (per connection) and as well require them to purchase blocks of bandwidth in Connectivity Virtual Circuit charges. As Simon Hackett, CEO of ISP Interode recently wrote, “The NBN Co backhaul (CVC) costs are far too high, as currently framed: In a network built with abundant bandwidth, the current ‘CVC’ pricing will create (and charge for) artificial scarcity”[7].

If the NBN is going to recoup its massive investment it will have to increase monthly wholesale ARPU from $37 in 2012 to $64 in 2023, a rise of more than 6 per cent annually. That compares with a real fall of 7 per cent per year in the cost of services over Telstra’s copper network since 2000.

And of course there has to be a reality check here. Just as you cannot suspend the laws of physics, so too can you not suspend the “laws” of economics. An overcapitalised government, or indeed private, monopoly is not going to deliver lower prices. Only competition can ensure that.

The case for FTTP often ends up as “we are building a network for the applications twenty years in the future.” There at least two problems with that. First forecasting twenty years in technology is the acme of speculation. Second, if a network’s cost bears no relation to its current utility, the consequent higher prices will actually inhibit demand and takeup. In South Korea, for instance, customers have proved to be very price sensitive and operators like KT have seen a substantial churn down the speed ladder as consumers look to save small amounts of money per month[8].

  • And finally, the slow lead time in getting to Market. On current forecasts, the NBN is scheduled to be completed by December 2020[9] although we know the project is well behind schedule. By the end of next month, the NBN was forecast to have passed 152,000 houses with its fibre[10] but as of the end of March, only 18,200 premises had been passed with fibre[11]. It doesn’t compare well to Malaysia’s HSBB rollout, which was initiated around the same time as the NBN and is on track to have 1.3 million premises in its footprint by the end of the year.[12]

Over the last year there has developed a narrow but extremely aggressive campaign by supporters of the NBN to frame the debate as being a contest between copper versus fibre. One Internet site recently posted photos of some egregiously run-down components of the old copper network and then asked me “is this infrastructure worth upgrading?”

Let me make two points about this. Firstly, politicians are relieved when the damaging photos being sent to their office are of naked copper connections, as opposed to naked other things. But secondly, and more seriously, I would make this very important point: The longevity of the copper network should be decided by the market’s desire and willingness to still purchase internet services over that network – and not by arbitrary decisions made by politicians in Canberra or elsewhere on when to shut it down.

In so doing, we shouldn’t think of this debate as a purely technical one – that is, which technology is best at this very moment. Rather, we should think of this about two alternative upgrade paths where the endpoint will be very superfast broadband that exceeds the capacity of current networks.

In a question of technologies, the answers tend to be simplistic and absolute. But in a question of alternative upgrade paths, the trade-offs are often complex and you need to take a hard headed approach. As the European Commission’s Vice President Responsible for the Digital Economy Neelie Kroes, said in March:

“Cutting off competition is not the way to promote sustainable investment up to 2020. Because we have to realise there is no magic solution to achieve our ambitious targets overnight. If there were, we would have picked an earlier date than 2020.

Rather, we will meet those targets with a gradual approach based on a mix of technologies. Whether it’s fibre to the home, fibre to the cabinet, next generation mobile solutions, or of course upgraded cable: they all have their part to play. We need a complementary combination of solutions, introduced incrementally, and tailored to local needs.”[13]

Where to From Here?

So assuming the Liberal National Party Coalition wins Government in 2013 what will we do with the NBN?

Part of the problem we have in Australia is that too many people have become obsessed about pursuing the best theoretical technical broadband solution rather than searching for the best practical solution to fit the unique needs of a market like Australia.

On paper one can see the benefits of the NBN rollout, but the network is not being deployed on paper: It is being deployed in real streets and real front-yards across the least suitable market in the world for an FTTP- network deployment in which most people live in sprawling suburbs in detached homes – which as we all know are hugely expensive to connect to FTTP.

One need only consider Hong Kong – the most highly penetrated FTTP market in the world – to see how operators in an intensely competitive fixed-broadband market handle the issue of connecting standalone properties to FTTP.

In Hong Kong, neither PCCW or Hong Kong Broadband Network have deployed FTTP to the detached properties owned by wealthy residents of The Peak and other high-income areas with non-MDU dwellings because the business case for doing so does not exist – unless the residents themselves agree to at least part-pay for the connection to be laid.

The plain fact of the matter is that if one looks across the Asia Pacific region – where the vast majority of global FTTP connections exist – you can quite easily see that the overwhelming majority of FTTP connections in Japan, South Korea, Taiwan, Hong Kong and Singapore are in Multi-Dwelling Units (MDUs). In Australia, however, only 34 per cent of premises are MDUs[14].

As any network operator can tell you, even deploying FTTP to an MDU or apartment building is a huge challenge in itself in terms of gaining legal access to buildings and then replacing existing in-building wiring is a huge challenge which operators in supposedly “FTTP friendly” markets like Hong Kong and Singapore could expand upon at length.

But let me just lay out some important principles which will guide our approach to the NBN.

Firstly, we will seek to encourage facilities-based competition wherever possible. The Government has signed deals with Telstra and Singtel subsidiary Optus to shut down their HFC networks for the purpose of delivering broadband. We will seek to negotiate a reversal of those agreements. Likewise, we will seek to reverse the cherry-picking laws preventing competitive investment in superfast networks and try to level the uncompetitive playing field in the greenfields industry.

Secondly, where because of geography it is not economic to provide broadband services at prices comparable to those available in the big cities, rather than an opaque cross-subsidisation bolstered by a monopoly, we will provide open and transparent subsidies to ensure regional and remote Australia gets equitable access to the digital economy. I note that if the Labor Government hadn’t cancelled the Howard Government’s OPEL scheme those remote and regional areas would have had upgraded broadband services in operation for several years now.

Thirdly, in completing the upgrade of broadband services across Australia, we will be technologically agnostic – there will continue to be FTTP, but also FTTN and, if it can be satisfactorily negotiated, wholesale access to existing HFC networks. Our focus will be on a service quality outcome – not a particular technology

With this approach we are satisfied that we can complete the construction of a national broadband network faster, because a mix of technologies will upgrade services sooner than near universal FTTP, at less cost to the taxpayer and more affordably for end users, because the combination of a less expensive network and the return of competition will put downward pressure on prices.

The Government has sought to present the debate about the NBN in Australia as being broadband vs dial up; the future vs the past; progress vs reaction. But the truth is that there is nobody in Australia more committed to, and more excited by, the promise of the digital economy than me.

Of course the vast bulk of broadband applications, be they e-commerce, e-health, e-learning, do not require 100 mbps connectivity. Indeed as telcos everywhere are discovering the absence of applications requiring high speeds makes it very difficult to extract a premium for super fast broadband services.

All too often the advocates of FTTP argue that every conceivable application, even down to smart metering whose bandwidth requirements are extremely modest, requires super fast services.

FTTN services are comfortably offering speeds well over 25 mbps, as high as 80 mbps in the UK for example, and that is more than adequate for the vast majority of residential customers. And as BT is demonstrating in the UK, it is also possible in a VDSL area nonetheless to run FTTP to particular premises who need super fast services.

And as for those who argue that super fast services will trigger a tsunami of innovation, we can only wonder then how it is most of the truly innovative online businesses of our time have not come from the superfast nirvanas of Korea and Japan but rather from the apparently broadband deficient United States – think Facebook, Google, Amazon, Twitter, Netflix, Youtube, Skype – it’s a very long list indeed.

In my judgement the scarcest resource is not bandwidth, or even technology, but rather technological imagination. And it is no accident that innovation is at its greatest in the markets with the most competition and the most freedom.

Our opponents have accused us of wanting to destroy the NBN. Nothing could be further from the truth. Rather we are determined to set it free.

[1] Li, W., & Xu, L.C., (2002), “The Impact of Privatization and Competition in the Telecommunications Sector Around the Word”, Social Science Research Network Electronic Paper Collection, Working Paper No 02-13.
[2] “58% of the increases in log output in the telecommunications sector around the world between 1990 and 1998 can be attributable to the joint impact of privatization and competition on the sector’s total factor productivity Ibid, p.24

[3]“The pooled estimates broadly suggest that countries having stronger actual and prospective competition tend to have higher productivity levels, lower prices and better quality levels in telecommunications. The role of the “time to liberalisation” variable is particularly interesting because it would suggest that the mere prospect of competition generates pressures that lead to gains in efficiency and consumer welfare” Boylaud, O., & Nicoletti, G., (2000), “Regulation, Market Structure and Performance in Telecommunications”, OECD Economics Department Working Papers No.237, p.21, available online here.

[4] ICT, (2010), “Malaysian Government to Examine Retail Broadband Pricing”, available online here.

[5] The NBN reported that in the six months from January 2011, received 1,804 development applications representing more than and 146,608 premises and forecast that it would pass approximately 65,000 lots and connect approximately 40,000 premises in greenfields areas (NBN, 2011, “submission to the Joint NBN Committee, available online here). The NBN later confirmed that: “As at 31 December 2011, NBN had received 2,956 applications from developers, with 1,988 active applications covering 109,988 lots. As at 31 December 2011, there were 110 premises receiving active services” (NBN, 2012, Third Progress Report, p.11)

[6] See ACCC, (2012), “Submission in Consultation on NBN Co SAU”, available online here.

[7] Hackett, S., (2011), “Further Corrections to the Record on NBN Pricing Models”, available online here.

[8] See my website, here and here. An Analysys Mason study concluded: “If customers continue to perceive no advantage in 100Mbps speeds compared with, for example, 10–60Mbps speeds, then they are unlikely to be prepared to pay a price premium for them. However, if customers do recognise value in 100Mbps services and are willing to pay for them, NGA operators are likely to maximise revenue by continuing to charge a price premium for such services. Therefore, even if the basic subscription cost of an NGA connection becomes comparable to that of a basic broadband connection, and operators are able to migrate standard broadband users over to an NGA technology, there is still likely to be a price premium for the fastest services. This could represent a significant roadblock to reaching the goal of 50% household take-up of 100Mbps services in the European Union.” Analysys Mason, 2012, “Will Price Premiums be a Roadblock to the Adoption of Next Generation Broadband in Europe?” available online here.

[9] Corporate Plan, p.23

[10] Corporate Plan, p.19

[11] NBN, (2012), “NBN Fibre Rollout”, available online here.

[12] Telegeography, (2012), available online here.

[13] Kroes, N., (2012), “Unblocking the Broadband Bottleneck”, available online here.

[14] McKinsey, (2009), “Implementation Study”, p.78

Recommended Posts