Rip-off NBN enters financial ‘death spiral’

Competition from mobile broadband is a growing threat to Australia’s National Broadband Network (NBN).

Telstra, Optus and TPG have each introduced fast, cost effective Fixed Wireless Access 5G broadband, which is capable of delivering high data transfer speeds and low latency.

5G has already stolen significant market share from New Zealand’s far superior NBN-equivalent infrastructure “Spark”. The same is happening in Australia with Telcos ramping up sales of their own 5G products, which they claim are faster and cheaper than the NBN:

More than a quarter of Australians consider their phone and internet costs to be unaffordable… and retailers are now increasingly pushing wireless 5G plans which in many cases are superior – and less expensive – than those offered by NBN…

In May TPG’s subsidiary brands launched a new $60 per month 5G home internet service that the telco said offers superior value compared to NBN’s 50 megabit per second plans…

Since launching its 4G and 5G home broadband services last year, it has signed up more than 110,000 customers nationally…

Two out of three customers switching across to TPG’s fixed wireless services are former NBN customers, and the telco expects over the long term that about 20 per cent of its customer base will take up wireless NBN alternatives.

The mass exodus from the NBN will likely worsen given the NBN Co’s submission to the ACCC, whereby it wants to lock in price rises until 2040, and double the price of entry-tier plans over the next decade. NBN Co argues that it needs to lift prices to recoup the costs of building and upgrading the network.

The NBN is now expected to cost taxpayers and private financiers a total of $57 billion from 2024, which is nearly double the original cost estimate of $29.5 billion proposed by the former Coalition government.

Given NBN Co’s massive fixed costs, it needs almost every Australian household to sign up if it is to remain solvent.

So if a significant share of households instead leave the NBN for 5G wireless, then these fixed costs will be spread over a smaller subscriber base, resulting in further price escalation and then more households ditching the NBN for cheaper alternatives.

If this switching process takes hold, then NBN Co faces the prospect of a ‘death spiral’ of falling subscriber numbers amid rising costs.

Unconventional Economist
Latest posts by Unconventional Economist (see all)


  1. Stephen Morris

    I know I’ll be jumped on by the True Believers but I could never understand the logic of NBN.

    Why not provide high capacity where it is required and use wireless for the rest, with progressively reduced cell sizes?

    Rudd’s original scheme struck me as analogous to the attempt to force standard gauge railways on Australia in the 19th century when there was clearly no need for that level of technology. By the time demand caught up with the cost of standard gauge, railways had been replaced by the new technology of road transport anyway.

    Here’s the original comment from 25th September 2013:

    What has always interested me is the parallel between NBN and the development of railways in the 19th century.

    Both were the state-of-the-art infrastructure of their day. In both cases there was enormous pressure from the technical boffins and political megalomaniacs to build a system to the highest technological standard.

    In the case of railways this manifested itself as the gauge battle. The engineering megalomaniacs wanted Australian railways to be built to the highest possible gauge. New South Wales adopted the British (4′ 8 1/2″) standard gauge while Victoria went one further and adopted broad (5′ 3″) gauge.

    Fortunately for the interests of allocative efficiency, it was not possible to impose these wasteful policies on the continent as a whole.

    Queensland, Western Australia, most of South Australia, Tasmania, New Zealand and the Northern Territory were free to adopt the more appropriate 3′ 6″ gauge, and eventually there were more route miles built to this standard than to any other. (This gauge was also used for the vast majority of track in geographically comparable countries such as Brazil and South Africa, while the similar metre gauge was also retro-fitted into the Indian rail network to augment the broad gauge where appropriate.)

    • drsmithyMEMBER

      I know I’ll be jumped on by the True Believers but I could never understand the logic of NBN.

      Universal accessibility and public ownership of critical (and often outside of major population centres, monopolised) infrastructure.

      Also, the NBN as it exists today is a very different beast from the NBN as originally envisaged.

    • Fart Mechanic

      It’s a nice theory…

      But to use the railway example, we are still using the
      same railways we rolled out in 1854!

      ..sure some small changes here and there, but its still
      the backbone, no other means of transport has been able
      to replace it.

      Many different types of alternative transports have come
      and gone. And we are very soon gone change from ICE cars
      to EV.

      But the railways are still gone be here..

      And thats it.. simply 150+ years of infrastructure service
      and still relevant. 5G/6G/7G are temporary and will have
      to be replaced every 3-5 years to keep up.. just like
      2G/3G/4G did.. with fiber, no..

      I dont think we need calculator to work this one out..

      • Stephen Morris

        The vast majority of Australia’s railway route-mileage is gone. It was technologically obsolete in the face of road transport and was closed. The same is true of rail infrastructure around the world in all countries except those with high population densities.

        Even in Europe it’s hard for rail to compete with road for freight. I recall being in Oslo in January 2013 trying to get to Copenhagen. There were no trains available. I asked the pleasant young woman at the station, “How would YOU go to Copenhagen?” She replied, “We fly.” (And as for the new high speed trains, they’re built to a completely different running gauge from the original system. They are a different technology which couldn’t run on the old alignments.)

        Most Australian rail which remains is special-purpose rail for coal and mineral transportation (in Queensland and Western Australia) and was build since the late 1960s . . . . in response to actual demand.

        As noted in the link:

        As anyone studying rail development in Australia will be aware, the primary demand was from the hinterlands to the ports. Even within states – where gauge change was not an issue – coastal lines were amongst the last to be built.

        Had the needlessly expensive standard gauge – or broad gauge!! – technology been imposed across the continent, it is plausible that much of the 19th century transport network would never have been built. It would have been too expensive. Politically powerful centres would have got their railways, and everyone else would have been left to their own devices.

        In the 20th century, as demand for inter-capital connections rose, these were retro-fitted with standard gauge on those lines were it was economically justified.

        It is worth noting that at a discount rate of, say 7% real, every extra dollar spent in 1870 in the name of uniformity, would need to have saved $867 (in constant prices) in 1970 when uniformity was actually required.

        Of course, by that time the vast majority of the rail network in Australia – whatever the gauge – had become technologically obsolete in the face of road transport and was being abandoned!

  2. People will flock to 5g – right up to the point where too many people are on it and service quality drops, or they realise they can’t live within the data cap.

    Yes, Optus has unlimited plans. In practice, they’re going to have congestion issues when peak streaming time happens. We’ll see if they roll out new towers before users leave.

    4k streaming is particularly burdensome on wireless. 3.5-7gb per hour. With Telstra’s current 400gb 5g plan, and let’s assume it’s 5gb per hour, that’s only 80 hours of tv a month before you’re out. A lot, but I dare say some people manage it.

    • Stephen Morris

      Why new towers? Why not put wireless nodes on the power poles . . . as and when demand justifies it?

      I can imagine a day when every 20th pole has a node. Then every 10th. Then every 5th.

      Eventually – 20 years time perhaps – it might even be worth installing fibre, where the demand exists. In the meantime the cost of capital has been saved.

      That’s assuming Elon Musk hasn’t devised a technology to deliver from satellite.

      The world is full of unknown unknowns – including the unknown unknown of new technology. That’s one of the reasons that capital has a cost.

      • Will the wireless nodes on the power poles interfere with the nano-bots in my Covid vaccine? I’ve never had better 5G reception since I’ve had the vaccine!!

      • Towers was used as a catchall term. Yes, they come in many sizes and formats.

        Satellite has all the problems of wireless spread over a much larger area (ie even more customers). Seriously. Get one and make a voice call. It’ll be odd. Some of it can be masked, but most of it is just inherent.

      • macrofishMEMBER

        Why dont we put femtos cells in to each house or on each pole, we can then run a bunch of fibre to connect them to, it will very national and a very braodband like network.

        we can call it the national broadband network.


        • Stephen Morris


          a) time;

          b) risk;

          c) cost of capital.

          A technology that allows progressive roll-out in response to observed demand is a technology that preserves “real options”: the options to speed up, slow down, modify, or stop altogether.

          Demand might never rise to the level required to justify further investment. It might rise but more slowly than expected. There might be new technologies.

          The world is full of unknown unknowns. That’s one of the main reasons capital has a cost.

  3. This article is from the Australian, ruperts glove puppets who destroyed the proper fibre optic nbn to save foxtel. This gave us a hobbled overpriced service, running on dying copper. 5g is no competitor, as it quickly gets clogged and slow and puts big data limits up. If ruperts monkey had let us have proper fibre we wouldn’t have to spend billions propping up dead copper.

    • This blog post today prompted me to sign up to Telstra’s 5G home plan.

      We are in a small regional town in NNSW with 5G coverage which provides far greater performance than the current NBN service to our address. We only get a paltry 27 Mbps download due to a copper lead-in a long way from the node.

      5G performs well on the mobile at home so should perform well.

      Look forward to receiving the modem. It’s only an extra $5 per month too.

  4. I know it’s poor form to say “Told you so” but some of us did tell you that 5G would kill the NBN
    I lost count of how many times I got told to learn some RF basics before commenting again. Optical fiber is great stuff no argument on that point but modern RF systems are capable of pushing right up to the Shannon channel limits (and maybe even going beyond)

          • Shannon’s capacity limits only strictly apply to Point -to-Point data communications.
            The channel refers to the medium connecting the points, in the case of RF comms it’s air.
            Now Shannon’s theorem is still correct (which is amazing considering that he deduced this in 1948) however RF doesn’t need to be Point to point, one can configure a single channel as Multi Point Transmit to Multi Point receive, at which point it is possible to exceed Shannon’s Capacity limits for two reasons, firstly you have multiple possible parallel channels (each taking a different path ) and the second is that some of the “noise” on the channels is correlated. If most of the noise on channel 1 and channel 2 is correlated than it’s no longer random as seen by the receiver so it is possible to correct for these errors.
            Modern RF systems are playing games with FEC (forward error correction) along with dynamic carrier loading (Higher order QAM) multiple diverse Tx/Rx channels. to achieve some pretty amazing results.

            Sorry for the Technical stuff, I know most MB’ers don’t like to read tech facts.

          • call me ArtieMEMBER

            Reply to Dodgy:
            Great stuff, well explained. I understood what you said. I enjoy your comments.

          • @dodgyas thanks man! That made sense. Makes me wish I took more than just one semester of Data Transmission in uni, many aeons ago.

          • Jumping jack flash

            “some of the “noise” on the channels is correlated. If most of the noise on channel 1 and channel 2 is correlated than it’s no longer random as seen by the receiver so it is possible to correct for these errors.”

            Ah yes.. takes me back to introductory signal processing

  5. Quantitative FleecingMEMBER

    Isn’t it good to still have the NBN infrastructure in place as a failsafe? 5G is great and all but if it ever comes to war again, space infrastructure & satellites would be an easy target. A few missiles could bring down an entire global 5G network in a couple of minutes, whereas it’s not so easy to go and destroy every undersea cable at once?

    My knowledge on this stuff is limited but I think that’s a factor. Just my 2 cents.

  6. 5G will never kill the NBN. Due to the law of physics.

    They can go to 5G but it will get over saturated.

    The NBN got us through the the pandemic by the skin of our teeth. Imagine 5G LOL.

      • So to keep critical infrastructure humming along we should drop the need of a internal rate of return.

        Musk’s Starklink could help the NBN ditch it’s skymuster costs because that’s the big POS dragging the NBN down.

        Oh the the people who complain that the NBN is terrible is because their ISP sucks. There’s only a single good ISP in Australia’s market. Aussie BB.

        • call me ArtieMEMBER

          Thanks Luke. You done good there. I bought Aussie Broad Band shares a couple of years ago because I read their prospectus and looked at their business model (i.e. create happy customers). The shares have been doing great for me and I expect it to do better as time goes by. Going thru a bit of a soft spot right now like everything, but I’m HODLing

      • I’m half way through the month as a single person. 134GB down 40GB uploaded. There’s no 5G plan that caters for this. I don’t even game. I work from home and watch TV at night.

      • Fart Mechanic

        ..yea, this is correct!

        Telstra/Optus are selling olive-oil extracted with cancer causing
        chemicals as “extra virgin oil” and they are getting away with it.

        Selling “unlimited 5G” which are physically limited but no
        regulation in place to stop them from disclosing it.. that could
        break the NBN’s back!

        They simply sold more on false promises and dominated the market..

        5G has its place for sure, its designed for the mobile user!

        If this occurs then the outcome of this is really the same
        as what NBN tried to fix:

        The country with the slowest internet in the world!

        ..and voila.. you got tricked.. again..
        Paying premium for inferior, thinking you got a the best deal!

    • I wish people would google the technical side to all of this. Guys, 5G is not going to replace the NBN. Not here, not anywhere, now or ever. We will have to upgrade to universal FTTP at some point, everyone will.

    • I agree with the average revenue per user, but when the article goes on to say that 5G will only get better when it footprint gets bigger is false.

  7. kiwikarynMEMBER

    As a smug Kiwi on an unlimited 300 Mbps plan as standard, I don’t even know how much data I use as nobody bothers to track it anymore because there are no data limits. Which is great, because things like P2P are now fully unhobbled so torrent speeds are fantastic. I don’t watch broadcast TV, everything is streamed on demand. Low latency and no drop outs is also very important, as I run software programs all day that require 100% uptime connectivity to overseas server farms. You don’t know what you are missing.

  8. I’m a bit late to this party, but can anyone comment on the strategic risk of wireless vs fixed line given the role of electronic warfare in conflict?

    • UpperWestsideMEMBER

      In a Carrington Event ( be it natural or man made EMP) you will fry all your unshielded electronics starting with the most exposed first. So …. anything with an antenna gets hit first. Fibre is better than copper ( as the network of copper acts as a big antenna) but you are still going to lose a big chunk of electronics.
      But … in electronic warfare this isn’t really the scenario. All the adversary needs to do is systematically brick all your routers and you are screwed. Internet offline, electricity offline, banks offline, transport offline ( including any “over the air” capable cars, trucks, tractors). Live in a high-rise ?. No elevators, no water ( no power for the water pumps ). Live with that for a couple of weeks and it’s hard to bounce back from.

Leave a reply

You must be logged in to post a comment. Log in now