What now for the NBN? Answers to questions from iTWire.

I have been asked by iTWire to provide my thoughts on where the NBN should go from here as the rollout of the network is announced as completed.

The article appeared on 20 July 2020 and can be read here.

My full answers to the questions submitted are as follows :

I would like your views on the future direction that the NBN should take.

The debate over the NBN has always been framed as one about which technology should be deployed. While I was Chief Technology Officer at NBN Co the debate saw politicians (ie. Conroy v Turnbull) continually argue the merits of fibre (or more specifically Gigabit Passive Optical Network or GPON) versus copper (Very High Speed Digital Subscriber Line or VDSL). No other political debate, perhaps with the possible exception of climate change, has seen politicians get into such a detailed debate about something they have little or no expertise in. On many days I felt as though the politicians were vying for my job rather than who should set the policy for improving Australia’s lagging broadband performance.

Other countries have not had this type of intense technology debate by politicians on how taxpayers’ dollars should be spent for telecommunications infrastructure. Why is this case? Is this a significant part of why Australia is a broadband laggard? To get to the bottom of this dilemma it is necessary to look at the broader  history and context of Australia’s fixed broadband problems.

In my view the reason for this peculiar state of affairs is that our political parties, Liberal, National and Labor, are actually unified in an erroneous view that fixed broadband is always a natural monopoly. Most other markets (Western, Asian and others) have been able to move from the entrenched monopolies of the fixed telephony era to structures where competition between technologies (ie. VDSL, HFC, GPON and mobile) has driven investment in higher speed broadband networks in urban areas. Some markets (such as New Zealand and Singapore) have reverted to monopoly structures for deployment of national fibre networks – but none have done this for the less capable VDSL and HFC technologies. Australia has been unique in keeping a monopoly in place through the long transition from the 20th century copper networks to the 21st century fibre network future. Competition between the different fixed broadband technologies has not been part of the policy framework. As a result politicians have been asked to adjudicate on complex engineering and economic decisions.

The reason for Australia’s tri-partite political alignment on a telecommunications monopoly has always intrigued me. The best I can come up with is that each party’s traditional support base has a vested interest in extracting the inherent extra profits (or rents) of the telecom monopoly. The Liberals, as the party of business, have an incentive to orchestrate the privatisation of the monopoly as we have seen with the tortured privatisation of Telstra. The Nationals, as the party of rural and remote interests, have an incentive to see a (hidden) cross subsidy within the monopoly be directed to more infrastructure for the bush. Labor, as the party of trade unions, has seen fit to have a state-owned monopoly that supports a unionised workforce that would have an advantage in directing the rents towards workers in the industry.

Since World War II we can see this play out with the various stages of the development of Australia’s telecommunications. From the 1950s to the 1970s, the then Country Party were able to use their position to direct large investment in the then government owned Postmaster General’s Department’s rollout of telephony infrastructure across most of regional Australia. During the 1980s and early 1990s we saw Labor, at the behest of its unions, consolidate OTC and Telecom Australia into what became Telstra in a way that made it difficult for any new competitors to seriously threaten its dominance. In the late 1990s and 2000s we saw the privatisation of Telstra by the Liberals, finally overcoming one lone National senator’s (ie. Barnaby Joyce’s) objections. The Liberal’s business base had successfully lobbied (remember ATUG) both Labor and Liberal to open up the business telecommunications sector to competition but in so doing, left the consumer sector subject to the monopoly. This left the Liberals a clear path to satisfy the growing financial and investment community with a Telstra monopoly in the consumer sector that would allow it to recover some of its lost lustre after the tech boom crashed its share price.

We have seen the urban consumer benefit hugely from competition in the mobile network sector since the early 1990s. Technology has advanced steadily from 2G to 5G while pricing has decreased. Efficient mobile technology deployments and rollouts have been driven by competition. Technology competition between GSM, CDMA, WiMax, LTE and others has played out in the market rather than in political contests. This contrasts with fixed broadband in Australia where monopoly structures have restricted investment for the urban consumer and forced Australians to react through the political process rather than the market. Meanwhile, large business customers have benefitted from fibre network competition in CBDs and most dedicated business parks.

The NBN era has seen both Labor and Liberal retain this monopoly structure for residential broadband infrastructure deployment. Labor’s FTTP policy had at least some support from an economic perspective of justifying the future network as a return to a natural monopoly. Due to its virtual unlimited bandwidth capacity, an optical fibre can achieve a natural monopoly position which economists define as ever decreasing average unit costs of supply as demand (for bandwidth) increases. Singapore and New Zealand, with government funding supporting private companies, have taken this path. Labor’s deals with Telstra and Optus, along with legislation to disincentivise alternative fibre network rollouts, were aimed at cementing this natural monopoly position. Whilst in Opposition, Malcolm Turnbull (see here) and Paul Fletcher (here), were highly critical of Labor’s monopoly policy. However, rather than take up a pro-competition policy, the preferred area for public debate before the crucial 2013 election was on technology, build cost and time to build the network. This led to the alternative policy for a VDSL FTTN network and subsequent transition to the Multi Technology Model (MTM) when in government.

Once in power, the Liberal’s criticism of Labor’s monopoly policy quickly faded. Undoubtedly a factor in this was TPG Telecom’s immediately announced plans to rollout a competing VDSL FTTB network to about 500,000 inner city apartments. This move by TPG Telecom highlighted the susceptibility of the Coalition’s NBN plan to increased competition, in contrast to Labor’s FTTP more robust natural monopoly characteristics.

Rather than embrace competition as they had done in Opposition, the Liberal policy was quickly redirected towards protecting the monopoly as espoused by the new chairman, Ziggy Switkowski. A range of regulations and a new levy were quickly announced along with new regulations imposed by the ACCC (normally pro-competition but during the NBN era a compliant part of both Labor and Liberal monopoly structures) requiring incumbent style open access regulations on all new broadband networks. The plan to protect NBN Co’s policy has been successful to date, but its ongoing susceptibility to competition continues to be exposed by Telstra, Optus and Vodafone’s plans to bypass the NBN with their Fixed Wireless Broadband services on both 4G and 5G networks.

As is to be expected, the debate at the end of the rollout period, with the Liberals in power, has returned to how best to orchestrate the privatisation of NBN Co. Telstra, in a pre-emptive move, has restructured its business, creating a new business unit,  InfraCo, in anticipation of taking advantage of the privatisation. Such a privatisation process will benefit the financial and investor community, but is unlikely to be positive for the urban and rural consumer, as incentives to upgrade the network will be very limited. There has been no discussion about reform of the sector to increase competition prior to NBN Co’s privatisation.

It would seem the political and industry consensus is that Australia’s fixed consumer and small business market should remain an NBN Co (either privately or public owned) dominated monopoly going forward. The debate has been narrowed to, firstly, whether it is a private or public sector monopoly and, secondly, how to regulate the monopoly to make it invest in upgrading the network as broadband demand inevitably increases. While regulating monopolies is not something new, this is usually about the prices they charge, not about forcing them to invest in upgrading technology. Regardless of whether the monopoly is private or public, it is likely the investment will only be made if the government provides the funds or allows higher prices to be charged. The vested interests involved (investors, unions, bureaucrats) will push for more funding or higher prices rather than looking to innovations and efficiencies. As a result technology, and its costs, becomes the frontline of the debate rather than the most efficient way to incentivise the investment.

How do you think the network should be managed in its next stage?

Many commentators in Australia point to the New Zealand FTTP experience as the way Australia should go. In New Zealand there are four wholesale only operators who have been subsidised by the government to build FTTP networks in separate defined geographies. In order to participate in the subsidy scheme, Telecom New Zealand was required to separate its network business from its retail business and create Chorus. These subsidised operators are not protected monopolies – they still face competition from new entrants, Chorus’ older FTTN network and potentially from each other through expanded network builds.

Would this work in Australia? It may have at the beginning of the NBN era if some non-Telstra companies had been available to create FTTP network build momentum and threaten Telstra enough so that it voluntarily separated like Telecom New Zealand. However, with NBN Co now dominating the market with regulatory protections against competition, the subsidies necessary to force upgrades to FTTP would be very high. The only competitive threat would be from mobile operators offering Fixed Wireless Broadband services. This is unlikely to be sufficient to stimulate widespread upgrades of the NBN with deeper fibre networks.

A more informative market to look at for a way forward is the UK. Before becoming Communications Minister, Malcolm Turnbull, consistently used the UK as the prime example of a market that was taking the ‘sensible’ FTTN VDSL approach. British Telecom were in the midst of a large upgrade of ADSL to VDSL very similar to what was proposed by the Liberal NBN policy platform in 2013. However, seven years on, as a result of competition from HFC (Virgin Media) and a range of FTTP companies, British Telecom has now switched to an FTTP upgrade plan. The UK regulator, Ofcom, has encouraged competition and in particular the opening up of British Telecom’s ducts and poles to allow the building of competitive fibre networks.

The option for government sponsored funding of a revised buildout of an FTTP network appears remote given the $51+ billion of funding already committed. The Coalition has ruled out more funding as it prepares the stage for privatisation of NBN Co.  A change of government to Labor could see increased taxpayer funding as privatisation is scrapped, but would this funding be sufficient to achieve the natural monopoly advantage of a deep fibre network?  Will the Australian public buy the proposition that more taxpayer funding can be efficiently spent when there is likely to be a loss rather than a return from the funding committed to date?

Again the debate would be about technology and costs, not about creating a market structure that would sustainably invest in future upgrades.

Absent large amounts of additional government sponsored funding for NBN Co to pursue a revised buildout of an FTTP network, the only option is for pro-competition reforms along the following lines :

  1. Promote policies that will encourage competition in the building of alternative fibre networks
  2. Remove current restrictions aimed at protecting NBN Co from competition
  3. Create an alternative subsidy scheme for rural and regional broadband investment that does not penalise competitors to NBN Co.

How do you think the network should be managed in its next stage?

The most significant way that NBN Co can be restructured to promote more competition is for the company to be split into technology based operating units. This was suggested by the Vertigan inquiry that was put in place by the Coalition in 2014. However, the Coalition government ruled out disaggregation at the time and instead doubled down on the NBN Co monopoly as described above.

In my view a minimum of four separate units, with different management, should be established. The first should cover the FTTP, FTTC and FTTN architectures as they all have a common network technology evolution end-point, namely Passive Optical Network FTTP. Let’s call this entity NBN_FTTx. The second should cover the HFC network architecture, let’s call this NBN_HFC. The third would cover Fixed Wireless and Satellite, let’s call that NBN_Wireless. A fourth unit, NBN_Core, would cover the core network elements that mainly rely on Telstra’s underlying assets – exchanges, ducts and transit network dark fibre. NBN_Core would provide a common interface for RSPs at Points of Interconnect and associated IT system interfaces.

A diagram may help explain this more effectively :

Each of the access companies (NBN_FTTx, NBN_Wireless, NBN_HFC) would be able to extend and upgrade their networks. Significant overlaps exist between the NBN_FTTx and both of the NBN_HFC and NBN_Wireless networks enabling increasing competition to drive investment.

Progressive stages of privatisation should be used to raise capital for investment in expansion and upgrading of the access networks. Government’s share of each unit would progressively decline, but maybe at different rates. NBN_Wireless  will require a subsidy (discussed further below) which may necessitate Government retaining more control over the operations and direction of this entity.

NBN_Core could be merged with InfraCo if a suitable arrangement could be negotiated given the large commercial payments tied up with the lease of ducts, dark fibre and exchange space from Telstra.

This disaggregation of NBN Co needs to be planned and managed to ensure existing relationship with RSPs and customers can be transitioned to the new arrangements. Opportunities for efficient use of new fibre infrastructure investment to service both fixed and mobile (ie. 5G) requirements should be encouraged by allowing the mobile operators to participate in the progressive privatisation of the NBN access companies.

However, privatisation of the NBN Co entities needs to be managed in a way to ensure none of the RSPs are able to re-create a monopoly and dominate the NBN NBN eco-system through the application of normal competition law.

How much money (approx) will be needed in your estimate?

The FTTN and HFC networks will require the most investment. A total of 7.2 million premises will require upgrades with average costs likely to be in the order of $1,000 to $1,500 per premise in order to reach a likely final destination of fibre to within the last 100m of all households. This makes the total upgrade cost between approximately $7 and $11 billion.

The technology evolution of FTTN network will be towards FTTC and FTTP using PON cable architectures. HFC will develop along the clear evolution of the DOCSIS ecosystem with progressive node splits to bring fibre closer to households and business overtime. Depending on the involvement and requirements of 5G (and further mobile network evolution) we are likely to see new fibre architectures develop to achieve efficient fibre deployments that combine both fixed and mobile network use cases.

Further investment will also be necessary for the 1 million premises covered by Fixed Wireless and Satellite in order that speeds of greater than 25Mbps can be delivered on a sustainable basis. Ongoing upgrades will involve more cell sites and new/replacement of satellites. Costs will depend on technology developments but could involve up to $0.5 billion per year in capital expenditure.

How should the remainder of the build (or upgrade if that’s a better word) be handled?

Upgrades in most urban areas would be driven by competition combined with private investment in the progressive privatisation of the NBN Co access companies. Progressive privatisations will allow for the decreasing of risk from an investment perspective leading to higher valuations for the NBN access companies as they achieve a financial track record. Government equity will progressively be of higher value over time as a more sustainable industry eventuates on the back of private sector investment.

Upgrades and ongoing operations in regional areas, where competition is not sustainable due to lower household densities, will require ongoing funding through an appropriate subsidy scheme. The best way to do this would be to impose a Medicare style levy on all telecommunications services at the retail level. With a market size of approximately $40 billion per annum, a 2.5% levy could raise $1 billion per year. This would be sufficient to cover the higher operational and capital costs per households for regional and remote services for both fixed and mobile services (ie. cover more mobile blackspots on an ongoing basis).

To avoid political pork-barrelling, an apolitical agency should be established to determine the best use of these funds to achieve goals associated with uniform pricing and service levels for broadband in areas where competition is failing to achieve minimum benchmarks. A possible model for such an arrangements could be the Pharmaceutical Benefits Scheme which follows the advice of an expert committee (the Pharmaceutical Benefits Advisory Committee) on how best to subsidise different pharmaceuticals needed by all Australians.

Do you think we should end up with all-fibre wherever possible?

While fibre is technically ideal, it is not likely to be economically practical for all Australians. After all the copper network didn’t reach all Australians either. With the technical advances in wireless communications (terrestrial and satellite) we can expect ongoing improvements that can be brought to market by companies with the right incentives.

I do expect the FTTN and HFC networks to be progressively upgraded so that fibre is within at least 100m of most premises in their footprints.

It is most important that we have an efficient investment regime that ensures just in-time upgrades to meet the increasing data demands of our society. Building a one-size fits all network may in fact end up wasting a lot of resources which could be better used elsewhere.

Do you see that kind of outcome eventuating if the NBN is bought by some private interest?

This outcome is possible if private interests are subject to competition as described above. As I mentioned earlier most of the markets that are higher up the scales on international broadband rankings are there because they have created pro-competition policies encouraging the private sector to make fibre investments.

If NBN Co is sold as a monopoly then we can expect what we saw when Telstra was privatised. After much to-ing and fro-ing the rent-seeking monopolists would only invest in upgrades if the government provided the funding (remember Sol Trujillo’s regular trips to Canberra). It would be the worst of all possible outcomes. If the definition of insanity is to do the same thing over and over again and expect different results then privatising NBN Co as a monopoly would surely qualify.

And if it is sold, which private company is best suited to take charge and remedy the faults that exist?

It doesn’t matter which companies end up participating as long as competition is forthcoming. It is most important to ensure that no one company ends up dominating the market and as a result refuses to invest without guaranteed exorbitant profits. We have been in that situation before and need to avoid it at all costs.

But if NBN Co is privatised as a monopoly then we can expect little, if any, upgrades towards a deep fibre network without significant government subsidies thrown it. The Australian private sector is very good at extracting money from government to enhance its profits. If this path is followed then Australian taxpayers and consumers will have again suffered a case of losses being socialised while the profits are privatised.

So if Australia cannot find its way out of its fixed broadband monopoly dilemma then it would be much better to keep NBN Co as a state-owned company.  Better to have the government accountable for the cost and profits rather than let Australian private sector establish yet another private monopoly.