The Rise and fall of the Australian NBN
Green Nylon jacketed 576 fibre optic ribbon cable at an installation site used for the backbone of the NBN in Australia
Benoît Felten takes a closer look at what went wrong with Australia’s ambitious National Broadband Network
In 2019, the Australian National Broadband Network (NBN) will celebrate its 10th anniversary. It would be an understatement to describe that (quasi) first decade as rocky. When the announcement for the NBN was made in April 2009, the world looked on in awe. This was the most progressive technology policy anywhere, promising fibre to the premises (FTTP) to more than 90 per cent of the population.
The project (of course) was also widely criticised, both in Australia, where opposition politicians argued that the demand did not exist for such high-speed services, and that therefore the public investment was not justified. More generally, there was a criticism of what some saw as a reversal of the telecom liberalisation of the 90s, bringing the national network back into the public fold.
- There are three crucial aspects of the Australian NBN that must be detailed to understand how its story unfolded: The first is that it is fully publicly financed, not a public-private partnership (where a private entity is financed in part to build something for the public good), but a fully-owned government operation. The Australian government was accused by detractors of creating a new ministry and, in many ways, it was.
- The second is that the existing network infrastructure of incumbent Telstra was not to be structurally separated (as happened at the same time in New Zealand) in order for this new network to be built, but rather leased to Telstra in perpetuity.
- The third is that, in order to meet its business model targets, the NBN could not face any competition. It was to be an absolute monopoly with no presence in the retail market (it would resell access to internet service providers). This meant decommissioning Telstra’s copper network as FTTP overbuilt it, plus purchasing existing cable broadband operations.
The first phase of the NBN’s history was slow moving. A deal had to be negotiated with Telstra for access to its infrastructure and decommissioning of its copper network. A deal had to be negotiated with Optus for similar access to its cable infrastructure. The entire operation had to be put in place. To cut a long story short, by mid-2013 NBN had only connected around 30,000 customers to FTTP. In the national elections that year, the Rudd government lost its majority and the Abbott government took over. It immediately announced changes to the NBN.
The key aspect of these changes was embracing a ‘multi-technology mix’ to achieve the same coverage as the initial plan, albeit at slower speeds. The fundamental premise of a publicly-owned wholesale broadband network was not called into question, probably because by that stage it was too far advanced to roll back without a massive financial write-off (by that stage billions were owed to Telstra for use of its existing ducts and poles). By reducing NBN’s ambitions, large amounts of money were supposed to be saved.
Grand plan derailed
Four years on, it’s safe to say the promised savings have not materialised. If anything, multiple reports and analysis suggest that costs may have further exploded. Amidst a sharp increase in customer complaints that the NBN was not delivering the promised speeds and that customer service was unsatisfactory, Prime Minister Malcolm Turnbull stated that the NBN was a ‘calamitous train wreck’. NBN CEO Bill Morrow, for his part, declared that the NBN might never make a profit. They (unsurprisingly) blamed the state of affairs on the previous government.
On the (slightly more) positive side, NBN Co currently connects roughly 2.5 million homes across its different technologies. Still a far cry from the 13 million of the initial plan, but a significant achievement nonetheless. This article’s aim is not to decide who is responsible for the train wreck, but to explore what went wrong, so that those looking at policy intervention in broadband deployment can learn from these mistakes. If we examine the life of the project, there are a number of issues that stand out.
From inception, the project was extremely political, something which plagued it over its decade of existence. Building national infrastructure is a policy decision, but it is also a long-term endeavour, one that inevitably will go beyond a single government term, and most likely over several. Therefore, if political stability could not be guaranteed, at least some form of bipartisan policy agreement should have been sought from the get go. This is how Stockholm’s Stokab (on a smaller scale, obviously) managed to survive and thrive despite political changes.
The initial accusation that the government was creating a ministry should have been seen as a warning to avoid doing that at all costs. Instead, NBN Co was staffed virtually from day one with more than a thousand employees.
Notwithstanding the ambition of the initial plan, it’s hard to imagine why so many people would be needed, especially in light of the very slow start of the network deployment and adoption. As a comparison basis, we can (again) use Stokab which operates a full fibre network of roughly 500,000 access points with never more than 90 full-time employees in its history. Since, in both cases, the build-out is subcontracted, the comparison isn’t fully accurate but does give a sense of the potential profligacy of NBN.
Looking back, it’s clear structural separation, New Zealand style, would have been a smarter route to achieve more with less public funding. The success of the New Zealand NBN is a slap in the face for Australia. The question of whether it would have been possible for the government to force a separation on Telstra is up in the air, though. Telstra in 2009 was a very powerful force with a lot of political capital. It was also a very acrimonious firm, suing the regulator and government on a regular basis. It’s not very likely that Telstra would have embraced a structural separation the way Telecom NZ did, despite the carrot (public funding) and stick (risk of a competitor overbuilding).
An often-overlooked issue revolved around the structure of the wholesale access points of the network. Early on, NBN decided it would only have 121 access points nationally (compared with Orange’s 19,534 access points in France, for example). NBN Co’s reasoning was to eliminate what was perceived as an unfair advantage for Telstra, which was already connected to all the existing access points of the legacy copper network in the country. The consequence, however, was to push the wholesale offer towards a high opex model, as economies of scale were, by definition, limited and transport of traffic from end users to access points was significant. This also forced NBN Co to create a dual price model for ISPs leasing lines, with both an access charge and a traffic charge. The former was at a fixed price, but the latter was left to the ISPs choice. The consequences of these structural choices was that prices for broadband over the NBN was high, with little to no prospect of lowering these costs over time (since service provider’s capital investment in connecting closer to end users is not an option).
In addition, ISPs have been regularly under-provisioning their traffic in order to lower costs, which has resulted in end user dissatisfaction targeted at NBN Co, rather than the service providers themselves. Both factors have had a negative impact on demand, which in turn hurts the NBN Co business model.
Fibre optic mass closure with dome, closing ring on installation table, NBN, in Australia
Obviously, the changes to the NBN Co’s technology mix mid-stream after the 2013 elections – just when NBN Co’s FTTP operation was starting to hit its stride – didn’t help. We cannot know where the NBN might be had this change not been forced upon it, but one thing is for sure: network design isn’t a process that can be changed halfway through without massive disruption, and it’s likely that NBN Co is still to this day paying the price for that disruption. Additionally, the political angle has led the Abbott and Turnbull governments to fully embrace technologies whose impacts on Telstra’s existing copper network were not well understood. This, in turn, created more cost and quality issues. It was clear to observing analysts from the start that the promise of savings on the overall cost would never be achieved, which of course now feeds a further round of blame and acrimony around the NBN.
So where does the NBN go from here? The most likely scenario is that it will continue on its rocky path for the foreseeable future, while being continually criticised on all sides. After the admission of its commercial failure and the ‘train wreck’ comments of 2017, the national (and political) conversation shifted to the idea of the government writing off most of the public investment. This would allow NBN Co to drastically lower prices to service providers and keep a viable business model. The government has clearly expressed it had no intention to do so.
Either way it’s likely that, despite the initial enthusiasm for its broadband ambitions, the Australian National Broadband Network will for a long time be seen as a failure. However, history has a way of forgiving past infrastructure investments if they survive to bolster economic growth in forthcoming eras (US railways or transatlantic fibre cables spring to mind). Let’s hope – provided NBN Co doesn’t stay too focused on short-term technological solutions – Australia a decade from now benefits from this. l
Benoît Felten is founder and CEO of strategy and consultancy firm Diffraction Analysis