It was a cold Saturday morning in Hobart in late July 2009. I was standing next to Stephen Conroy, the then minister for broadband, communications and the digital economy. We were there to hear Prime Minister Kevin Rudd announce that services of 100 megabits per second were going to be available from the National Broadband Network in three locations in Tasmania by July the following year.

For many of the attendees on that chilly morning this may not have seemed like a particularly challenging commitment, but to me it seemed to be exactly that. Only a few days earlier I had become the CEO and first employee of the company created to build Australia’s broadband network.

NBN Co had no premises and no employees. At that time it existed only on paper, with four Commonwealth officers as interim directors. So I turned to Senator Conroy and asked how the government intended to provide those services within a year. In his usual forthright manner he replied, with a smile, “Well, now you know why I hired you.”

As it turned out, those 100-megabits-per-second services to the three Tasmanian locations were made available by NBN Co in the time frame set on that July morning. Given the extensive product design, planning and architecture work required for the main NBN deployment, providing those services to three Tasmanian locations was a bit of a diversion. Such things were an accepted reality in managing a government business enterprise – just as regular attendance at Senate select committees, Senate Estimates hearings and parliamentary joint committees was an accepted part of the job for senior managers of NBN Co.

The diversion that was serious – in fact, disastrous – for the medium- and long-term future of fixed broadband services in Australia was the one introduced by the incoming Coalition government in late 2013.

Tony Abbott’s government replaced the largely fibre-to-the-premises (FTTP) model initiated by the Labor government with a Multi Technology Mix (MTM) model, which is still being deployed today.

The MTM, which uses a range of fixed-line architectures and technologies including fibre-to-the-node, fibre-to-the-curb and Hybrid Fibre Coaxial (HFC), will end up costing Australian taxpayers billions of dollars more than if the original deep-fibre NBN had been allowed to continue. It will also result in a network that is considerably less capable of meeting the nation’s future broadband needs.

The policy decisions made in September 2013, which mandated the use of existing copper and pay-TV cable technologies, have also created significant financial problems for NBN Co. In its most recent corporate plan, released in August 2018, NBN Co reported that the peak funding projection (that is, of the total equity plus debt required to complete the project) had increased to $50.9 billion. It is surprising that there has not been more of a public outcry regarding this figure, given that the Coalition estimate made in April 2013 was $29.5 billion. This represents a 70 per cent increase, largely due to the inherent costs of using the old copper and pay-TV cables.

And that may not be the end of these cost increases. There are some worrying trends that can be seen in NBN Co’s weekly progress report. At the end of June 2018 NBN Co reported that 1.44 million premises were not yet ready to connect even though they were in areas declared “ready for service”. This was up from 77,000 in June 2016. NBN Co is pushing a bow wave in front of them, particularly on the HFC network, which is going to have to be dealt with before the end of the project.

By comparison, the 5 per cent increase in the peak funding of Labor’s original FTTP-based plan, from $43 billion in April 2009 to $45 billion in September 2013, had both the Coalition and elements of the media declaring that Labor’s NBN was a disaster due to cost “blowouts”.

By way of example, in mid August 2010 a young reporter from The Australian contacted NBN Co’s communications department. He was writing a story after hearing somewhere that every home connected to the NBN would need to be rewired, at a cost of up to $3000. I spent 45 minutes on the phone with him, explaining that there was absolutely no need for any household to upgrade their internal wiring when connecting to the NBN. By the end of the telephone conversation he agreed that they had it wrong and he was going to write his story based on the facts I had provided.

I rather naively assumed that the 45 minutes on the phone was time well spent in averting the creation of an erroneous article. However, when published a day or two later, the article repeated the same claim of $3000 wiring costs and went on to say:

The NBN, if built, will pass more than 10 million premises. If every premise spent an average of $500 on bandwidth distribution equipment, the extra cost to the nation beyond the $43bn price tag for the network would be $5bn.

The Coalition has vowed if elected to scrap the NBN and substitute a modest $6.3bn network using existing copper, wireless and HFC cable technologies.

---

Even today, speculation continues as to why the press was so opposed to Labor’s NBN. The former prime minister Kevin Rudd recently wrote that the Murdoch media, in attacking Labor’s NBN, was trying to protect its interests in the Foxtel pay-TV asset. During my four years as CEO of NBN Co, I never saw any evidence that it was worried about NBN’s impact on Foxtel. In fact, it could be argued – and this was a position that I put to Kim Williams when he was Foxtel’s CEO – that a good outcome for Foxtel would be to transition off the HFC cable, deliver their broadcast content via satellite, and use the NBN for streamed services.

So if protecting the Foxtel pay-TV asset wasn’t the motivation for the relentless attacks on Labor’s NBN, what was driving the constant negative coverage?

The answer became clear to me in a meeting set up by Williams after he became CEO of News Limited (now News Corp Australia), with nine or ten of his editorial staff in attendance. This was a positive initiative aimed, I believe, at trying to improve the relationship between News Limited and NBN Co. Addressing the meeting, my NBN Co colleagues and I summarised the motivation and rationale behind the FTTP network, and reminded everyone of the view of the Australian Competition and Consumer Commission’s chair, Graeme Samuel, detailed in a May 2009 speech. The NBN, he said, could offer “serious advantages in a range of critical areas” such as health, education and teleworking, and an opportunity to correct longstanding structural problems in the telecommunications sector. He was referring to the problems created by successive governments privatising Telstra and not putting in place a competitive market structure.

No doubt the ACCC also recognised that private enterprise was never going to build a new fixed broadband network across Australia. The investment horizon is too long and the capital investment too high.

Certainly private enterprise would have built parts of such a network in areas of Australia where a good financial return could have been generated. But, as we told the meeting, it was unrealistic to expect private enterprise to provide every Australian premises with high-speed broadband access, and provide it at the same price regardless of where it is located.

We pointed out that of the three major fixed-line telecommunications deployments in Australia prior to the NBN, only two could be regarded as being successful – both built by governments. The first was the overland telegraph line built by the South Australian government in the 1870s. The second was the existing telephone copper network progressively built by the federal Postmaster-General’s Department, starting in the early part of the 20^th century. Both turned out to be of great value to Australia.

The third major fixed-line investment – in HFC cable technology for pay-TV purposes – was attempted by two private sector telcos, Telstra and Optus. The HFC infrastructure was duplicated as the two companies chased each other down Australian streets until they both called a halt to their rollouts, after only covering the more wealthy and profitable parts of the cities. In terms of the utilisation of capital, this multi-billion dollar duplication of assets was hardly an ideal outcome.

News Limited editors did not argue any of these points during our meeting, but by the end of the discussion it was clear that none of these facts was convincing for them. News Limited’s position regarding the NBN seemed to be one of principle. Never mind that the private sector would never build an NBN, or that the current market structure was flawed, or that government had a successful history of fixed-line infrastructure building. The view from News Limited’s side of the table was that the government just shouldn’t be building this sort of public infrastructure.

News Limited was not alone in its position. One of the most frequent comments I heard during the dozens of public speeches and presentations that I gave on the NBN was “Why are you building this? No commercial company would undertake this project. The returns are too low and the risks are too high.” That was precisely the reason the government was doing it, I would reply – because no purely commercial entity would undertake a project like the NBN. The government was also doing it because of the massive social and economic benefits to the nation in having a ubiquitous, wholesale, uniformly priced, high-speed broadband network. And an FTTP-based network could provide an acceptable long-term financial return for the government while providing ever-greater speeds and utility for end users.

Another question I regularly had to answer during presentations was “Why are you building a new fixed-line network when the whole world is going mobile?” This was also one that Malcolm Turnbull put to me when he first became the shadow communications minister after the 2010 election. The reality is that both fixed and mobile networks will be needed long into the future, a view held by almost every telecommunications professional. This is supported by Australian Bureau of Statistics information on the amount of data that Australians are downloading, measured in gigabytes per month. In a three-month period ending June 2018, almost 94 per cent of data downloaded was on fixed networks. This means only 6 per cent of data was downloaded on mobile networks. This ratio has not changed significantly over the past decade. Not what you would expect to see if fixed networks really were going to become redundant.

It is true that 5G mobile will be more efficient and allow greater speeds than 4G, but not to the extent that one often hears hyped in the media. The shared spectrum on which mobile networks depend is a scarce resource, so there are always going to be performance limitations. These can be mitigated by building many more small mobile cell sites (and each will need to be connected to a fixed-line fibre network) or by using much higher frequencies where possible (at higher frequencies, however, the interference from buildings, foliage and heavy rain becomes much more significant).

We need mobile networks that allow us to roam freely, and that connect the many remote devices that will be part of the Internet of Things. But we also need fixed-line networks and, in particular, high-capacity fibre infrastructure to do the “heavy lifting” for requirements such as streaming video to large screens, remote health and education services, and data science applications that require large file transfers. Those in the industry who take a more realistic view of emerging technologies know that the same traffic congestion issues will still need to be tackled with 5G, and the improvement is going to be incremental.

For those remote premises (approximately 4 per cent) that couldn’t be reached by the deep-fibre or fixed wireless network, NBN Co had planned to launch two high-capacity satellites to provide broadband access. Turnbull spent many hours in parliamentary joint committee meetings trying to prove that NBN Co management was providing a “Rolls Royce” solution well beyond what was required. He repeatedly made the same argument to the media. “There is enough capacity on private satellites already in orbit or scheduled for launch for the NBN to deliver broadband to the 200,000 or so premises in remote Australia without building its own.” Even as late as August 2015, in an interview with Fairfax Media, he said, “There’s no question I was very critical of the proposal for government to build two satellites for the national broadband network.”

One of the companies Turnbull encouraged NBN Co to consider was NewSat. NBN Co had already concluded that NewSat’s business plan was dubious at best, and even if it did launch its Jabiru-1 satellite it was planned to be located over the Bay of Bengal – hardly ideal for providing broadband services across the Australian continent. But it seems NewSat’s management had Turnbull convinced that its planned satellite should be an integral part of the NBN.

Unfortunately for Turnbull’s credibility, NewSat collapsed amid concerns over mismanagement and questionable financial dealings. The Sydney Morning Herald in September 2018 reported that NewSat’s founder and CEO had been committed to stand trial following allegations that he funnelled significant amounts of money from NewSat into his private yacht business.

But one had to admire Turnbull’s flexibility. Not long after these statements were made he was spruiking the NBN satellites as a “game changer” for broadband services in rural and remote Australia, and his Strategic Review posited the launch of three NBN satellites rather than two!

Unfortunately, Turnbull’s flexibility did not extend to his MTM, even when his original April 2013 predictions were proven to be wildly optimistic. It may have been better if Turnbull, as the incoming shadow communications minister in September 2013, had followed the earlier exhortations of his leader, Tony Abbott, to use the money being spent on Labor’s NBN to instead duplicate the Pacific Highway or assist with flood recovery. The result may have been a delayed NBN, but it would more likely have been built using a future-proof technology.

The Coalition’s move to an MTM model has created ongoing financial problems for NBN Co and its shareholder, the government. In 2018, NBN Co reported that the estimated time to achieve positive cash flow had slipped out by a year, to 2022, and the peak funding had increased to $50.9 billion.

On current projections, the initial network build will be completed by 2020. From that time, the ongoing cash flow will be largely a function of revenue and operating costs, with interest costs playing a part. And, of course, another round of capital spending will be required before too long, because many of the poorly served fibre-to-the-node areas with long copper lines will require upgrades to cope with an ever increasing demand for speed and bandwidth. Unfortunately, the potential to generate revenue and reduce operating costs has been significantly degraded by the move to the MTM, so it will be a serious challenge for NBN Co to achieve a positive cash flow and long-term financial viability.

Predictable speed performance and higher reliability would have made the revenue generation capability of FTTP considerably better than that of the MTM model – both in terms of take-up rates and average revenue per user. End users who are prepared to pay for a higher capability service don’t have that option with the various technologies of the Coalition’s MTM.

And then there are the operating costs.

It is usually a telco’s ongoing operating costs, not its capital costs, that determine its cashflow performance. Telcos constantly strive to reduce operating costs by simplifying their product set and associated processes. With the change to the MTM we have gone in the other direction. Using multiple technologies increases the complexity of the required IT systems, the maintenance costs, the holding costs for spare parts, the training costs and the overhead costs.

This point was admitted in April 2018 by the then CEO of NBN Co, Bill Morrow, in his paper published by NBN Co. He stated that the MTM model meant lower speeds, more faults and higher maintenance costs than Labor’s FTTP-based network. (He was, of course, also obliged to say that the original FTTP network would have cost more to build than the MTM. But this claim was, as we will soon see, not supported by the facts.)

It is also an unfortunate reality that the renegotiated Telstra deal put in place by Turnbull and Morrow has increased NBN Co’s costs. In the original deal, responsibility to make Telstra’s pit and duct infrastructure “fit for purpose” was squarely with Telstra. But under the deal renegotiated for the MTM, that responsibility, including any asbestos problems, was transferred to NBN Co – and by extension to the Australian taxpayer.

From every angle, the decision to use the old copper and HFC (pay-TV) cables has turned out much worse than the Coalition’s plan anticipated. There were plenty of warnings. A senior Telstra manager suggested in 2003 that the copper phone network would need to be replaced within 15 years, saying the ageing lines are now at “five minutes to midnight”. He also indicated the use of ADSL was the “last sweat” of revenue that could be wrung out of the old copper network.

Given that NBN Co had purchased more than 27,000 kilometres of new copper cable as of mid November 2018, in some cases to provide a barely acceptable service, the Telstra manager has been proven correct.

A similar story has unfolded with the attempts to use the HFC cable networks built by Optus and Telstra. In 2010, the team at NBN Co evaluated the possible use of these networks as part of the NBN. They rapidly came to the conclusion that this was not a viable option: at that time, they believed the best option was to use neither of the HFC networks, and instead to financially compensate Telstra and Optus for progressively transferring their HFC broadband customers onto the FTTP-based NBN. In fact, the NBN Co team took only a few weeks to come to the conclusion that Optus’s HFC network was unusable for hi-speed broadband without a major investment in remediation and upgrades.

This conclusion was heavily criticised by Turnbull, and he subsequently directed NBN Co to make use of both HFC networks as part of the MTM when the Coalition formed government in 2013; an interesting directive, given the claims of adopting a “technology neutral” approach. By 2015 it was evident that the Optus HFC network was indeed unusable and the 450,000 NBN services that Turnbull had projected would be delivered using the Optus HFC network were reduced to the 25,000 that had already been deployed.

The picture was somewhat better with the Telstra HFC network; nevertheless the number of NBN services to be provided using this network has decreased from initial estimates, while the implementation time and costs have increased dramatically. As recently as October 2018, the CEO of NBN Co advised Senate Estimates that close to a billion dollars of the peak funding increase are due to the “HFC pause” to fix problems and to “optimise” the HFC network.

There are now no easy solutions to these revenue and operations challenges, and to the impact they are going to have on NBN Co’s finances. The cash already spent on implementing the MTM cannot be recovered.

These woes, especially the problem with future cash flow, are exacerbated by NBN Co’s inability to participate significantly in 5G rollouts across Australia. Telcos understand that a deep-fibre deployment is vital to underpin 5G coverage, and Labor’s original FTTP plan would have been able to play a complementary role in the deployment of 5G. This was anticipated by NBN Co’s management when they allocated the number of fibres to each area as part of the original FTTP network planning.

If the deep-fibre FTTP network had been completed, NBN Co could now be developing another important wholesale revenue stream by offering wholesale fibre connections to the large number of 5G mobile cell sites needed. With the MTM, using the old copper infrastructure, this revenue opportunity has largely evaporated. This has turned 5G from a potential revenue opportunity for NBN Co into a competitive threat.

---

Despite the MTM strategy’s evident failures, Turnbull and the Coalition were very successful at convincing the media and the general public that the original NBN would have been much worse. The misinformation campaign waged almost from the start was very effective. One early experience I had as CEO illustrates this.

I was visiting a very senior public servant in Canberra to provide him with an update on the NBN. He was an intelligent and informed individual, but one who had no direct day-to-day involvement in the NBN. After we had exchanged pleasantries he said, “Mike, can you tell me why the NBN is so over budget?” My response was “Why do you think it’s over budget?” I then showed him the data as to how the project was actually progressing.

It was running about nine months behind the original schedule because of the time it took to complete the highly complex, but advantageous, deal with Telstra. However, the capital costs, the operating costs, the peak funding and the internal rate of return were virtually unchanged from the original estimates.

He reflected for a minute and said, almost to himself, “Why did I think it was over budget?”

It was because on an almost daily basis he had read or heard that the NBN was way behind schedule and way over budget – a good example of that famous adage variously attributed to Vladimir Lenin or Joseph Goebbels: “A lie told often enough becomes the truth.”

This technique is still being used with similar effect today by the current minister for communications, Mitch Fifield, and other members of the Coalition. When asked about the delays and cost overruns of the MTM, their response is to repeat ad nauseam that the original FTTP-based NBN would cost $30 billion more in peak funding and take six to eight years longer than the MTM. This is demonstrably false.

The reason that ministers have continued to repeat this claim is fairly obvious. The increase in the MTM peak funding and deployment times created quite a problem for the Coalition.

Their “solution” was to claim that the original NBN was always going to be more expensive and take much longer than the estimates made in September 2013 by NBN Co. (Those estimates showed a peak funding of $45 billion and completion by the end of 2021.)

The range of peak funding costs for the FTTP NBN claimed by Turnbull and then Mitch Fifield went from $94 billion in April 2013, to $64–73 billion in December 2013, to the $74–84 billion estimate made by NBN Co in August 2015.

The $94 billion figure included in the April 2013 Coalition policy document was a complete work of fiction. In fact, the Strategic Review initiated by Turnbull very soon after the 2013 election could not support the $94 billion number, even though its authors did their best to inflate the peak funding requirement of Labor’s FTTP NBN.

One would assume NBN Co could be trusted to make an accurate estimate. In its corporate plan for 2016, NBN Co says, “Management estimates that an all-FTTP fixed-line rollout could be completed by 2026 but possibly as late as 2028, with a peak funding range of $74–84 billion (vs. $46–56 billion for MTM)”. This statement became the basis for the answer given by Fifield to every question asked about the network: that Labor’s original NBN would have taken six to eight years longer and been $30 billion more expensive than the current MTM.

But as Bill Morrow, NBN Co’s then CEO, admitted in a Senate Estimates hearing in October 2015, the analysis performed by NBN Co did not attempt to estimate the peak funding and timing for a continuation of the FTTP-based NBN. Instead, they were asked by the government to predict the costs and timing of winding down the MTM and restarting the FTTP NBN – after two years of cost blowouts, commitments and delays with the MTM.

It is clear the $74–$84 billion peak funding and the calculation that it would take six to eight years longer was a nonsense concocted to allow the spurious argument that the original NBN plan would have been even more expensive and taken much longer to complete than the delayed and very costly Coalition alternative.

Not only did NBN Co make its estimate based on a restart rather than a continuation of the original NBN, they also made an even more egregious assumption: that there would be no significant reductions in the cost-per-premises FTTP fibre rollout over the life of the project. The original management team planned for a reduction of approximately 25 to 30 per cent over the eight-year life of the project. This is what was built into the corporate plan, which resulted in a peak funding cost of $45 billion. The new management team appointed by Turnbull, and the board he installed, assumed there would be no reduction in rollout costs over the life of the project. In fact NBN Co’s 2018 corporate plan continues to show the same $4400 for the FTTP cost-per-premises that it quoted back in 2013.

The difference in total costs, from disregarding this reduction, is enormous. One has only to remember that the cost-per-premises figure is multiplied by 10 million, the number of services to be completed by the end of the NBN build. This sleight of hand added billions of dollars in capital, interest and peak funding costs to the projected cost of the original FTTP NBN.

This of course raises the question: Could the 25 to 30 per cent cost reduction assumed by NBN Co, while I was CEO, have been achieved? Was this a reasonable assumption?

The answer is a resounding “yes”. Telcos all around the world have been reporting major reductions in fibre rollout costs over the past several years, due to advances in technology and practices. A comparison close to home is that of the largest FTTP builder in New Zealand, Chorus, which reported a 44 per cent reduction in per-premises costs over the last few years to 2017. This makes a mockery of NBN Co’s continued quoting of the same $4400 per-premises figure.

---

Given the complexity of all of this information, it is little wonder that the average person has problems sorting out fact from fiction concerning the original NBN. This is also why Turnbull as communications minister was so effective in “recasting” the numbers to support his assertions, and to seem, as usual, so erudite and plausible as he did so.

It is surprising, however, that Turnbull’s financial and technical credibility were not severely damaged when his assertions were blown out of the water by his own team only 12 weeks after he became minister. Remember: in April 2013, he said that his MTM would cost $29.5 billion, and all premises would be connected to the NBN by 2016 – he was $21.4 billion and four years off target.

It is also surprising that Turnbull persisted with his wildly optimistic estimates, when he must have been getting advice from many directions that the time frames and costs were not realistic. I advised him, in September 2013 in a meeting he requested at his Edgecliff electorate office, that it would take a long time and much effort to renegotiate the Telstra deal, as he had very little leverage. His response was one of quiet mirth. He was very confident it would all be straightforward.

It did turn out to be straightforward – for Telstra. NBN Co ended up with much of the responsibility for making the infrastructure Telstra was leasing fit for purpose. This was why Telstra’s CEO could say that the new deal was “unquestionably better for shareholders”. In other words, the deal negotiated by Turnbull was unquestionably worse for taxpayers.

While the average person may not follow all the financial and technical details of the NBN, most people do suspect that something has gone badly wrong with what was a popular and greatly needed national infrastructure project. They can also spot when an “independent report” is plainly wrong, such as when the 2014 Vertigan Review, authored by a panel hand-picked by Turnbull, made the somewhat odd statement that a broadband connection with a peak rate of 15 megabits per second would be sufficient for 50 per cent of households in 2023. Even NBN Co CEO Bill Morrow thought this was a little “curious”.

Moreover, people know that it is incredibly short-sighted to spend billions of dollars to build a major piece of national infrastructure that just about meets today’s demand, but doesn’t allow for any significant growth in that demand over the next 10 or 20 years without very large upgrade costs.

It has taken several years for a clearer picture to emerge, but we now know the decision to change to the MTM was thoroughly flawed – and the network performance and NBN Co’s financials demonstrate this. The MTM network costs more and does less.

The nation will be bearing the consequences of that decision for years to come in an area that is critical to its long-term future.

Betting tens of billions of taxpayer dollars on a myopic, expensive and backward-looking network based on copper, as the Coalition has done – while the world was moving away from copper and embracing optical fibre – was a huge miscalculation. It was not driven by a sophisticated analysis of the best technology choices for Australia’s NBN, but by ideology and politics. As Paddy Manning observed in his 2015 biography of Turnbull, Tony Abbott was intent on killing off the NBN if the Coalition won government in 2013, and Turnbull believed the Labor plan for the NBN was flying in the face of 30 years of governments exiting from operating businesses.

The fact that such a huge amount of money has been invested in performance-limited MTM technologies means that a writedown of these investments is almost a certainty.

Unfortunately, and contrary to what is sometimes suggested, writing down the value of these assets will not improve NBN Co’s cash flow. And if NBN Co were to accede to the calls for significant reductions in NBN wholesale prices its cash flow would be further affected.

These calls are being made in spite of the fact that telecommunications costs, including mobile and fixed broadband, have decreased by 6 per cent since 1990 whereas electricity costs have increased by more than 200 per cent, water by more than 150 per cent and the consumer price index by more than 60 per cent.

All of this highlights the challenges for any incoming government to find a viable financial path forward for NBN Co. With the amount of taxpayer money sunk into deficient network assets, the increases in operating costs, and the now limited opportunities for revenue increases, the options for the NBN are quite constrained.

Even so, an incoming government will need to find a way to improve the NBN over time, given the importance of ubiquitous high-speed and high-capacity broadband. We can only hope that sensible options can be identified to allow the customer performance and financial viability of the NBN to be restored, after the disastrous infrastructure changes and misleading politicking that have dogged such an essential national project.

---

Looking back on the experience and challenges of starting up a huge government enterprise in a technically complex area, and which, from the very start, was operating in a politically charged environment, there is one lesson that stands out: the need to maintain an undistracted focus on the objectives set for the organisation undertaking a project such as the NBN. This requires a discipline on the part of the government in setting and maintaining the objectives, an operating company staffed by experienced people who understand the business they are in, and a governance structure that allows fast and efficient decision-making.

In the early days of the NBN, the employees displayed a passion for and commitment to the project because they knew they were part of something that would be of lasting benefit to the nation. They did not need to be motivated by financial incentives and bonuses because they were sufficiently energised by the project itself. I believe that if a government could ever set a new path for the NBN, then that same spirit could again be leveraged to deliver a broadband network that will truly meet the long-term needs of the Australian public.