Warning Signs at NBN - Malcolm Turnbull

I want to provide an update on the Labor Government’s National Broadband Network and the Coalition’s alternative approach to communications, which of course is a critical enabler of the digital economy and business input.

At the outset I’ll stress the Coalition is committed to ensuring all Australians have access to fast, reliable and affordable broadband, regardless of where they live or work. But we don’t support Labor’s NBN, and we don’t believe it can achieve that objective. By now I suspect our key criticisms are familiar:

• The NBN is expensive. Completing the network and moving customers onto it will cost $50 billion and very likely more. More capital inevitably means higher prices for consumers and businesses. Most users have no near-term use for the high speeds a fibre to the premises (FTTP) network can deliver, and international experience tells us the few who do will not pay much of a premium. So the increased cost of the network won’t be shouldered mainly by those who take advantage of its capabilities; they will be spread across all users.

• The NBN is anti-competitive. As it is rolled out, three existing networks – Telstra’s copper, the Optus HFC cable network, and Telstra’s HFC network – will no longer be permitted to carry voice or broadband, and the first two will be decommissioned. This is economically wasteful and harmful to competition. But it no doubt will prove helpful to an over-capitalized government monopoly keen to recover its costs.

• The NBN will take years to rectify every area with substandard broadband. Much of Australia has good communications infrastructure but not all; at least 2 million premises cannot access broadband or are constrained by limited speeds. Given the scale and logistical complexity of the NBN rollout, it will take a decade to reach all of them even if it rolls out on schedule.

• And the NBN puts the Government back in a conflicted position as the owner of a large player (that at some stage will likely be privatized) in a market it regulates.

Broadly these were our objections when the current version of the NBN and its startling price tag were unveiled in April 2009. In the two and a half years since, none has been adequately answered by Senator Conroy or any other devotee of Labor’s current policy.

Today, however, I want to focus on some other aspects of the NBN that have been canvassed in the public domain recently, and prompted red flags to be raised in some quarters:

• NBN Co’s accountability and transparency.
• NBN Co’s obsolete financial projections.
• Rollout priorities, black spots and take-up.


From the outset, the Labor Government has done its best to minimize scrutiny of the NBN: refusing to have the project evaluated by either Infrastructure Australia or the Productivity Commission; largely exempting it from FOI laws; declining to release un-redacted versions of NBN Co corporate documents and forecasts; and limiting Parliamentary committee hearings.

Many details of this $50 billion gamble of taxpayer funds are not visible to members of Parliament – even those on multiple committees overseeing the NBN, such as myself – or to the taxpayers who are funding construction of the network and the generous salaries paid to executives.

For example we don’t know whether (having received $1.7 billion in capital from taxpayers since 2009) NBN Co actually earned any revenue from selling communications services in the September quarter. Or if it did, how much. This is because Senator Conroy rejected a proposal for quarterly reporting from the Government’s advisors.

We don’t know the timeframe for the NBN in areas where broadband is inadequate or unavailable, unless they are on a recently-released 12 month schedule. Nor is this likely to change; Labor would presumably prefer to use the NBN schedule as part of its marginal seats strategy at the next election than reveal to long-suffering citizens in under-served areas when their broadband will finally be upgraded.

We don’t even know how many customers NBN Co has at present, although NBN Co CEO Michael Quigley told a Senate estimates hearing in October there were 870 satellite users and 1500 active services over fibre at that time. (Such numbers are about 4 per cent of the 2011-2013 Corporate Plan’s implied Q3 2011 forecast.)

There does not appear to be a commercial reason for such information being withheld – rather it reflects an apparent preference for secrecy and reluctance to be accountable at NBN Co.

Likewise, an aversion to oversight or having to justify its decisions is evident in NBN Co’s year-long attempt to ‘lock down’ key variables in its Special Access Undertaking to the ACCC so they cannot be changed – a push confirmed in yesterday’s public release of the final SAU submission.

NBN Co is a monopoly building an expensive network that access seekers will have to pay to use, but as the SAU underscores, it opposes both any ACCC review of its rollout investment, and more generally, any second-guessing by the regulator of other management decisions it takes.

Unsurprisingly, access seekers are not impressed. As the CEO of Optus, Mr Paul O’Sullivan recently responded: “Well, I’m sorry, but this is an infrastructure monopoly, someone has to make sure its spending is efficient…if they spend too much, you and I will end up paying for it.”

Mr Quigley has shown a similar reluctance to provide Parliament with even the most anodyne details of commercial matters. While commercial in confidence is a legitimate protection for some data in such circumstances, it should not become a shield that allows any question, however legitimate, to be loftily brushed aside.

Not only has NBN Co preferred opacity and avoided accountability, it has also several times shown a cavalier disregard for facts – especially reporting on its own progress. For instance the second quarterly report of the Parliament’s Joint Committee on the NBN relies in part on the following NBN Co statement to shareholder ministers: “For Greenfields [new housing developments] NBN Co has exceeded its expected target with 75,000 lots passed”.

The 75,000 target reflected a time when competitive private infrastructure builders were expected to handle greenfield sites. NBN Co instead chose to manage and build its greenfield rollout directly and appoint an exclusive contractor, Fujitsu, to replace private infrastructure builders. In addition, Telstra was later added as network infrastructure provider of last resort for greenfield sites with less than 100 lots (the majority will receive copper.)

Given all this, it is difficult to determine how many (if any) of the 75,000 households NBN Co claims to have passed in 2010-11 were in fact passed by it; or were passed by fibre at all. But if we count NBN Co greenfield estates under contract – neither passed or connected but scheduled for work in coming months – the number in November 2011 was about 8000.

And however we cut the numbers it does not appear to be correct that NBN passed 75,000 greenfield lots prior to June 2011, which is what was apparently stated to shareholder ministers.

Perhaps the most glaring recent example of excessive secrecy over NBN Co (which is, after all, an agency every Australian taxpayer is supporting that will have no direct competitors as they have all been paid off or prohibited by the Government) was last week’s release of a Greenhill Caliburn review of the NBN Co 2011-2013 Corporate Plan. Over its last 25 pages only 15 full paragraphs were entirely free of redaction.


Ironically one controversial item in the Greenhill Caliburn document not blacked out was its estimate that NBN Co capex between 2011 and 2028 would total $50.6 billion in nominal dollars. Since this was substantially more than the Labor Government’s claim of a $37 billion rollout, it raised questions about NBN Co’s publicly available financial projections (those published in its 2011-2013 Corporate Plan.)

“It is a fallacy to suggest replacement costs and other capex is a part of the cost of the NBN,” declared Senator Conroy after the Greenhill Caliburn document was published. I have bad news for the Senator: Capital spending on maintenance and network upgrades is part of the cost of the NBN, as it is part of the cost of any enduring piece of infrastructure.

And what Greenhill Caliburn and NBN Co’s own figures reveal is that ‘replacement capex’ after the initial network is connected amounts to over $100 per connected premise per year – roughly the same as maintenance costs on the current copper network. So much for ‘future-proof’ in terms of reducing upkeep costs.

What else can we learn from NBN Co’s financial projections?

In truth a cursory glance reveals the financial forecasts provided to the public in December 2010 must have already been obsolete when published. They assumed most early NBN users would arrive via fibre rollouts in greenfield areas – but as mentioned earlier, greenfield activity stalled throughout 2010-11 as Senator Conroy struggled with his policy, and so NBN Co utterly flunked the 2010-11 targets it had set for itself:

• The Corporate Plan forecast 58,000 premises passed (mostly at greenfield sites and not counting 4,000 premises passed at the initial trial sites in Tasmania) and 35,000 premises with active service by June 2011.

• In reality only 18,243 premises were passed and only 622 had active service.

It is early days for the NBN. But rollout, coverage and take-up delays matter greatly.

They are one of two main drivers (along with ARPU or revenue per user) of NBN Co’s revenue estimates, which many analysts see as optimistic. The Corporate Plan estimates $23.6 billion in revenue will be generated between 2010 and 2021. Over the same period ARPU in nominal dollars per month is projected to rise from $38 to $58.

Earning that much revenue may prove difficult, and any shortfall means more debt for NBN Co.

But even more unrealistic are NBN Co’s current forecasts for operating expenses, which according to the 2011-2013 Corporate Plan will total $23.1 billion between 2010 and 2021, including $15 billion for payments to Telstra.

On 23 June 2011, NBN Co agreed to make decommissioning and customer migration payments of $4 billion over 10 years and infrastructure leasing payments of $5 billion over 35 years to Telstra. Crucially, that agreed value is measured in after-tax 2010 dollars using a discount rate of 10 per cent.

NBN Co also agreed to make decommissioning payments to Optus with an after-tax present value of $800 million (calculated using an undisclosed discount rate which we can probably assume is not dissimilar to the one used in the Telstra deal).

Clearly, to meet these obligations, NBN Co must make a series of much larger nominal payments, which will be taxed in the hands of Telstra or Optus and then discounted back to 2010 dollars.

The exact nominal operating expenses for NBN Co that result from these deals depend on the precise timing and profile of payments and pace of the rollout.

But broadly they mean between 2010 and 2021 NBN Co will incur operating expenses of about $19 billion to pay Telstra and about $2 billion to pay Optus (most investment analysts covering the telecom sector project roughly these nominal costs to deliver on the two deals)). Note that any delay in the rollout will increase the nominal value of the payments due to Telstra and Optus.

It follows that if NBN Co still needs to budget a further $8 billion to cover its ‘real’ direct and indirect operating expenses (as in the existing 2011-2013 Corporate Plan) and the rollout proceeds on time, then NBN Co’s total nominal operating expenses between 2010 and 2021 will not be $23 billion. On the face of it would appear they will be more like$29 billion – but in the absence of timely and accurate financial information who would know for sure?

If a listed public company signed deals such as the NBN Co/Telstra and NBN Co/Optus deals, it would be required to immediately disclose the impact on its existing financial guidance and forecasts to the market.

Not so NBN Co, whose level of disclosure has been so inadequate that it falls far below the standards required of a publicly listed company.

There is an additional question about NBN Co’s financials. Where exactly are any inside-the-house costs associated with shifting customers and their equipment from copper to fibre and ensuring service continuity? So far the Government and NBN Co have been quite slippery about the liability for any such costs – it is unclear whether these will lie with households, retailers or the NBN itself.

The high costs and potential complexity of switching premises to fibre seems to be verified by real world data. Well-sourced industry talk is that at Telstra’s fibre rollout in South Brisbane it is taking two technicians a joint half-day per premise to get services up and running. This is in addition to the costs of missed or rescheduled appointments, letter boxing and other ads to alert the community, and other co-ordination activities.

Given NBN Co’s financial data is so opaque, it is hard to tell which costs are counted where – but the organization should respond to these questions with a sense of urgency and release more accurate, realistic and clear financial projections. To fail to do so would be to treat taxpayers with a contempt that no listed company would ever dare show towards its shareholders.

So, Senator Conroy, one simple question: what exactly is the latest estimate of the peak capital requirement for the NBN?

As it is the NBN Co projects a defensive and non-transparent approach which seems like a cross between the Kremlin and the Church of Scientology.

The irony is that notwithstanding it belongs to the public, it is less accountable and less transparent to the taxpayers who ultimately own it than any of the publicly listed telcos to which it is currently dictating terms, with all of the inevitable arrogance of a massive government owned monopoly.


The only rationale for an NBN in the first place was to ensure regions with substandard broadband where market forces had failed were properly served. Yet this priority seems to have been abandoned on the way to NBN Co.

As Mr Quigley explained to a Senate hearing on October 18, NBN Co is choosing sites for its early rollouts based on access to Telstra’s dark fibre and exchanges or the deals it has agreed with its contractors – not on how urgently households and communities need access to good quality broadband.

Mr Quigley stated, and I quote: “Overwhelmingly in these early stages the availability of Telstra infrastructure dictates where we can go in the roll out. That is one factor. The second factor is what deals we have done with which construction contractors and how they can load, level and mobilise their workforces.”

So six weeks ago, when NBN Co announced a three-year plan to run fibre past 485,000 premises, only a fraction of the areas chosen were broadband blackspots.

It is simply unbelievable that after bungling its first tilt at an NBN and wasting its first term, the Government now cannot get its wholly-owned taxpayer-funded monopoly to prioritize the neediest areas. Ed Husic, the ALP Member for Chifley in Western Sydney and a former president of the telecommunications union, was sufficiently frustrated about this to risk retribution from Senator Conroy and other factional bosses and speak out in October.

Allow me to quote him: “NBN Co is making a big deal out of rolling out broadband in new estates where people haven’t even moved in, while down the road people are tearing their hair out to get ADSL or decent wireless access. I don’t begrudge new areas getting access – I’m happy for them. But how do you explain NBN Co’s priorities to residents in Woodcroft and Doonside, who are struggling to get decent internet access? It seems to me NBN Co is just reaching out for easy targets, hugging geographic areas within close proximity of exchanges.”

One of my concerns is that as the rollout confirms the NBN to be logistically daunting and financially untenable, NBN Co and the Government will try to obscure this by focusing on the easiest areas, not those most in need. That would be a travesty of social justice from a party that so loudly claims to believe in it.

Meanwhile, take-up is also proving an issue across the rollout. Internal NBN data from mid-October obtained by The Australian showed only one in nine of the premises passed by the NBN’s fibre had connected to a service at that date. Take-up rates were as low as 2 per cent in Armidale, 5 per cent in Townsville, 6 per cent in Brunswick (where NBN Co was overbuilding an area with HFC pay TV cables, providing comparable broadband), 9 per cent in Kiama and a slightly healthier 19 per cent in Willunga in South Australia.

The figures show higher take-up rates ranging from 12 to 25 per cent in longer-established Tasmanian trial sites where broadband options were far less plentiful – underscoring the value of focusing the rollout on poorly-served areas.

In the long run, if it continues being rolled out, NBN has little to fear from sluggish take-up given existing or prospective rival fixed-line infrastructure will be decommissioned or legislatively deterred. If consumers or businesses want a fixed line, they will have to use the only network available. That is why the monopoly has been imposed.


The end of the year is always a good time to take stock and review the decisions of the past twelve months. The Gillard Government’s NBN has been criticized on two main grounds: it is spending far more money than is needed to achieve the objective and it is massive, anti-competitive monopoly in an industry where competition has delivered better services and lower prices and, as a consequence, consumers want more competition not less.

The evidence from the NBN, even at this early stage, is that those criticisms were on the mark.

The rest of the industry is in uproar as Michael Quigley, in a manner reminiscent of another Michael, tells them the NBN Co’s wholesale broadband agreement is an offer they can’t refuse.

There are still only a handful of Australians connected to the NBN. The approach taken by NBN Co and Senator Conroy is utterly at odds with the best practices of carriers and governments in almost every other comparable market.

It’s not too late to take stock, conduct the cost benefit analysis and get this project right.