Ever since I was asked to present to the joint Australian, South Korean and New Zealand KANZ Broadband Governmental summit a month ago I’ve been ruminating  about the justification for Governments spending huge sums on rolling out ultrafast broadband in a country. A post I was alerted to by my friend Chris Auld spurred me into jotting down some thoughts.

First some background – down in my neck of the woods, the Australian and New Zealand Governments are spending up large on ultrafast broadband, spending $5B (it seems I used an incorrect reference there – Wikipedia says the NBN will cost AUD35.7B) and $1.5B respectively – massive sums, especially for New Zealand where it equates to a significant proportion of GDP, and where we’re also faced with a $20B or so repair bill from the recent Christchurch earthquakes. One would want to be pretty goddam sure that an investment of those sorts of levels would make an appreciable difference to society, and specifically the economy (since much of the justification for UFB is in fact a productivity increase leading to heightened GDP).

Now I have to say I’m not a luddite – I’m part of the knowledge economy that works remotely and brings back USD into New Zealand – 90% or so of my earning are from outside of New Zealand and they’re enabled almost entirely by the internet – as such one would expect that I’d be right behind ultrafast for all? Well not so fast – I live a long way from a city and hence don’t have fantastic connectivity – I average around 1Mbps – not ultrafast by any stretch of the imagination, but still sufficient to enable me to do whatever I need to – conference calls, video chat etc. I’m not sure what a few more Mbps would offer me, apart from quicker uploads to YouTube, something that would be nice of course, but not really transformational.

At the same time as our Government is investing heavily in uber internet for all, a group of private investors is contemplating building a new undersea cable between New Zealand and the US. Their rationale for doing so is the monopoly that currently exists due to having only one provider, Southern Cross. While I have concerns about the chances for Pacific Fibre (after all, as soon as it looks close to happening won’t Southern Cross just open up under-utilized capacity and kill Pacific Fibre’s financial business case?) the fact is that it is a private enterprise with private funding – so long as our Government doesn’t get a rush of blood to the head and decide to invest in what is essentially a private business (I’m talking to you Minister Joyce) then all luck to them. (For reference it is worth checking out Pacific Fibre founder Sam Morgan’s rationale for the business here).

The UFB project on the other hand is different, it’s a case of our Government taking a huge bet with our money on a perceived future benefit. So does it stack up? Well I need to point to the headline on the Crown Fibre website which says – “Crown Fibre Holdings – Increasing our Productivity”. So does it? Well according to a working paper from independent and well respected economic think tank Motu, there is no empirical evidence for this. Paul Walker goes into detail about the report over on his blog but the crux of the argument can be summed up in Motu’s conclusion which states;

…we conclude that a shift to broadband connectivity (from dial-up) appears to raise firm productivity but a move from slow to fast broadband may have no effect on productivity at all.

I absolutely understand the kind of “build it and they will come” arguments as espoused by Morgan in his interview (link to video above). As he rightly contents, we cannot predict the future and it is impossible to intuit where the game changing developments might come from and, by extension, what requirement they will have for connectivity. But in the same way it is impossible to know what technological advances will provide alternatives to traditional approaches towards broadband – maybe the demand profile will change, maybe we’ll build data centrers in New Zealand and hence avoid large amounts of international traffic, maybe we’ll discover ways of gaining amazing connectivity without laying cables under the sea.

With that many unknowns, is it right for Governments to make a bet on UFB? As one commenter points out, it seems the Government has a strange cause and effect mindset going on here;

The wonderful causality effect is alive and well and flying in the face of excellent research, common logic and market forces. The government have their cause and effect backwards (ie: high GDP countries have high speed broadband, therefore if we have high speed broadband we will have high GDP).

I’ve stated before that I’m not a luddite – I would love New Zealand to be in a position to build the sort of businesses that have a use-case and a need for ultrafast broadband (although I would say that many of those businesses would also be happy to be located where there is already great broadband anyway so the transformational aspects of UFB are further diluted). But until I see proven empirical evidence for the spend, I’m sitting firmly on the fence on this one.

As always, keen to hear the thoughts of others about this…


Ben Kepes

Ben Kepes is a technology evangelist, an investor, a commentator and a business adviser. Ben covers the convergence of technology, mobile, ubiquity and agility, all enabled by the Cloud. His areas of interest extend to enterprise software, software integration, financial/accounting software, platforms and infrastructure as well as articulating technology simply for everyday users.

  • I am in Brisbane,Australia and I am currently developing line of business software for a small-medium business.

    This business has small branches scattered around regional Australia, and the lack of high speed internet is a real hindrance.

    Ideally the regional staff should be able to connect in and use the company’s systems just like a head office employee, but for a lot of things it is just too slow. Instead they have to phone in and ask people to do things, people at head office have to specially prepare spreadsheets and email them out to the regions, and of course that data quickly becomes out of date….

    Until recently the internet was slow even at the head office. The company had to spend ~$20,000 to have a special fibre cable laid and we are only about 2km from the CBD.

    In summary, if high speed bandwidth was ubiquitous, this company’s staff could focus more on clients/sales, they would spend less money on me developing software that deals with the bandwidth constraint, and they would not have needed to outlay $20,000.

    • Fair comment – although I’d be interested to know what sort of connection speed you have and what sort of systems won’t work with a moderate speed broadband connection…

      • Ben,

        There’s a huge difference between 1 person working from home on a “slow” broadband connection and a small (or worse – medium) office sharing the same connection.

        1MB/s per person is probably ok for the vast majority of things.

        1MB/s / 5 = 200KB/s for a small office – things can get very tight then.

        One person watching a 10min video at 480p on youtube can (and does) impact the performance of all of the “remote” line of business applications that office uses.


  • Here is a quick attempt to explain my point. I don’t have a great deal of time to do a better job, sorry.

    The regional offices have adsl1, which I suppose at best is 1024/256 but my understanding is that they get half that at best. Also, the head office has to serve all these users which consumes the upload bandwidth which is invariably less/slower than the download speed.

    A while before my time they settled on using a terminal server solution for the regional users. Which has pros and cons, for which I have my personal opinion 🙂

    The main positive is that it is much cheaper than a vpn which would be conceptually equivelant but would consume more bandwidth, and much cheaper than building a web/http solution which opens a can of worms for infrastructure, security and cost of development.

    The main applications are ‘forms over data’, ‘decision support’, ‘business intelligence’ and also integrates and manages data from 40 remote unmaned facilities.

    If it was a bigger business, there is no doubt it could afford to employ lets say a 4 person software team, who could architect and build a secure and available web solution that would work well over low bandwidth. And I am sure all the software development purists would be screaming for this solution.

    However, business men aren’t software purists, they want to reduce costs, and software developers, system admins and security experts are costly.

    I guess my point is that with unlimited bandwidth the business could build crude but cheap systems, and with lower bandwidth the business is forced to build more efficient and therefore expensive systems.

    • “I guess my point is that with unlimited bandwidth the business could build crude but cheap systems, and with lower bandwidth the business is forced to build more efficient and therefore expensive systems.”

      Which isn’t actually creating any value. All it is doing is shifting costs. (from the telco to the crude and software developers)
      True payback in a national term means creating more value… not shifting it

      • I don’t think it is shifting costs at all. The economies of scale mean that all businesses will have cheap fast bandwidth, and therefore in the example of my client they can spend more resources creating business value (extra r&d, extra salesmen, extra branches, extra mines) rather than blowing money on either expensive bespoke fibre or expensive software developers who are working around the limitations.

        • The State (ie the taxpayer) is funding the network, not economies of scale. True you will get some EoS, but its being paid for by us.

          So the true argument is – do you think you will get more GDP growth from fast internet, or the tax break (which you could have invested in plant, branches, holes in the ground)…

          • I am in no way qualified to debate the economic case. I’m in way too deep here.

            All I really wanted to say was that I am consulting to a real business that has had tangible and significant costs arising from slow bandwidth in Australia.

            If this is the only business in Austalia with this problem, there is clearly no economic case for the NBN. If every business in Australia had the same problems there would be an overwhelming economic case (even at the tax payers expense I would suggest). Who knows where the truth lies?

          • All good points Ross and I’m absolutely sure some companies can benefit. All I was trying to say is that a true/ full tco is needed to understand If this will give a return at a national level.

  • http://www.nzherald.co.nz/opinion/news/article.cfm?c_id=466&objectid=10730596
    links nicely to your post

    Our slide down the oecd ladders is because we went from a frugal nation who was a farm for the UK to a flargrant consumerist who is a farm but the prices for the farm goods are a lot less (compeititve effects)
    So untill you get new ways of making money going you should slow spending …
    but we don’t, there isn’t any business creation thought to UFB. only consumers and IPTV….

  • BTW you’ll get diss’d for evena asking this question but the same mob who role out ‘roads’ as a similar type investment…
    But we had things (trucks) to use the roads… we don’t have businesses to use the UFB to create enough incremental GDP to fund this…
    2nd world nation trying to keep up with the rich coz..

    • yup – I’ve already been called a troll this morning… ah well, par for the course I guess…

      • A well travelled troll… 🙂

      • Not a troll at all.

        This is what TelstraClear CEO has been saying all along for years, and people say his just unhappy TCL is not in the main side of receiving money from the project.

        As it stands now, what you hear is a lot of people at home, downloading and hoarding more digital content they could ever consume trying to get all for $19.95/month and paid by the govt.

        I work from home, and having fast Internet access is great. But nothing that would dramatically change someone’s lives today.

        Sure, it makes a difference for offices, companies trying to produce digital goods or even connect to distributors, but what big deal will it be for the average household? Nothing. Just more YouTube, Facebook, and copyright breach…

  • I just saw an article in today’s Australian Financial Review that speculates the NBN could end up costing australia $1Bn / year in the long run in interest alone.

    Certainly worth questioning, but due to the huge contracts already signed it is nigh unstoppable over here.

  • What’s the cloud computing angle on this?

    My brother is ceo of a museum, with about 10-20 staff. He recently asked my advice about upgrading their servers. I suggested it might be a good opportunity for him to consider migrating to the cloud (mail, accounting, resource management…), rather than upgrading.

    Anyway, we ended up concluding that whilst that was a great idea in principle, there poor dsl connection (256kb up) made it too risky a proposition.

    • I wonder what the decision would have been though if it had been a case of 1 – 2 Mbps both up and down – not ultrafast, just fast enough….

      • Ben,

        I’m as cynical about this as you – but this is a bad argument.

        If you’ve ever had to manage a shared internet connection – then you’ll know that 200KB/s per user isn’t enough for “cloud apps”.

        If you don’t remember what this feels like – then try turning 3G off on your android and tethering your laptop. Then visit youtube and try to make a skype call.

        Imagine having to access Gmail/Calendar/General Web/Youtube/Some specialist LOB apps (Force.com/Oracle/SAP) + maybe VPIP over a connection like that.

        We tried it – and our phone systems were hugely unreliable until we installed a dedicated fibre link.

        Just because residential broadband is mostly enough for 1 user (or a family) doesn’t prove anything about bandwidth+latency needs of small and medium businesses.


        • Correct, which is why most of them buy business grade network products today…

          So, they have a solution, UFB is a substitute (and probably not that radically cheaper), so where again is the benefit?

          • The crux of the problem in Australia at least, is that the ‘business grade network product’ is not available at all, or is extremely expensive if the business is located on a street that the private ISP has decided not to lay fibre.

            Outside the metropolitan areas, the ‘business grade network product’ is not even an option regardless of how much you want to pay.

          • In many locations – business grade is _no_ better than retail broadband.

            The only difference is how long it takes someone to pick up the phone when I call to complain.

  • The case for the UFB.
    When the Auckland Harbour Bridge was being designed there was the discussion about should it have more than 4 lanes? What would we ever do with 4 lanes across the harbour!

    Why do we commute? Because we like to see people – many people. What would happen if we could feed our eyes the data that gives us the impression we are “really” meeting someone? What is the bandwidth that would make such meetings real? Real enough that you wouldn’t want to commute. Real enough that you wouldn’t want (need??) to fly to the other side of the world for a tech conference? ~8-)


  • There is a bunch of comment above about businesses and the impracticality of sharing a 2MB connection among say 10 users.

    I think this is a good point, but, to be it almost reinforces the primary point made by Ben. Do we really need to invest in UFB to every node in the network. Or is ‘fast enough’ OK for most home users and as such we only need UFB for a much much smaller subset of nodes that make up the businesses of which you speak.

    The thing with rolling out to every node is that some are going to cost a disproportionate amount to service. We’ve already accepted that we won’t be rolling fibre into rural homes. I think we need to take a serious look at whether the cost/benefit makes sense for urban homes as well.

    Sure let’s do a decent job of hotting up Internet for businesses, but, I don’t think it makes sense for individual customers.

  • Interesting article, but I think it would be short sighted to assume that if we can ‘get by’ on what we have now there’s no need for ultra fast broadband.

    5 years from now our demand will have increased significantly, and if our internet can’t cope we won’t even be able to interact with international businesses let alone compete.

    NZ has already been excluded from the release of Chromebook with inadequate internet being cited as one of the reasons.

    What will be excluded from 5 years from now if we don’t adopt ultra fast broadband?

    Quite possibly something far more critical leaving NZ businesses hopelessly behind.

  • Imagine a bell curve. At one end is your ‘Grandma’, still on dial-up and perhaps just migrating to broadband now so that she can maintain a low resolution Skype chat with your children. Also on that end are families that cannot afford broadband connections, and so their kids are unable to properly join the online world. (Perhaps those kids are lucky, and are at somewhere like Pt England school where they give every kid a laptop for $15 per month and connect to the KAREN network to provide decent connectivity.)

    At the other end of the bell curve are people that consume vast amounts of data at high speeds. They might need it for media consumption (of increasingly high definition video), for work (I’m using Dropbox as I write this to transfer large files) or for Skyping their grandkids on high definition. Or perhaps they are creating a business that requires those services. Or perhaps it’s like the majority of Christchurch schools who want to get HD video connections so that students from one school can attend classes over video conferencing with teachers and students from another school, saving us money and increasing the knowledge of our kids. That’s happening right now, thanks to some tireless workers and the support of Enable’s fibre to the schools project.

    But regardless of where we are on that curve, that entire bell curve moves each day, as the carriers deliver and we require higher and higher speeds, and as we increasingly accept always-on high speed internet as a requirement. Sadly NZ and Australia are well behind the rest of the developed world on this, while the definition of developed world itself expands as countries like South Korea invest in fibre (and other areas) and leapfrog their economy over ours.

    In the future it will be normal to have some of your kid’s classes conducted remotely at school or even to your home. Teaching material such as Khan academy will become the norm and for the video resolution of everything to be such that it’s ‘just like being there’.

    It’s the same for work. Skype may work for start-ups, but for dealing with corporates we need high quality vide conferencing that is not only increasingly higher in resolution, but also never drops out and just works. That’s not Skype, and it’s incredibly expensive to deliver right now.

    It’s critical that we all understand that what’s sufficient today is insufficient tomorrow. Just as we all laugh now at Bill Gate’s assertion at 640 Kbytes should be enough for anybody, we in the technology and business communities also natively understand that in 5 or 10 years time 1 Mbit/s is going to be ludicrously slow. Imagine trying to load the homepage of Stuff (over 1Mbyte) on a 9.6 kpbs connection. Imagine trying to watch SHDTV emergency alerts about the latest ChCh earthquake over a 1 Mbit/s connection.

    Fibre, copper and wireless are all bound by physical limits. Proponents of each show that the capacity that can be delivered over each keeps rising, and that’s true. However it’s also consistently true that fibre is the technology that can deliver the most capacity, and by quite a margin. It’s expensive to deploy, but once in the ground it’s relatively cheap to upgrade. Meanwhile in New Zealand the copper network to the home is most often ductless – and thus exposed to be corroded and difficult to maintain. There’s a place for wireless, be it networks generated in the home or business, or cellular networks that cover the populated country or, via satellite, an entire hemisphere.

    The UFB and NBN programs are not abut 2011, nor even about 2015. They are long term focussed and are ensuring that the necessary infrastructure is in place. Copper just won’t cut it in 2020, but fibre will still be working and expanding in 2050 and beyond. Consider that the base case for Pacific Fibre is 2.56 Tbits/sec for a fibre pair on a 10000 km+ fibre leg, and we can see that the potential for the link from the exchange to the home is huge.

    Consider also that what I regard as pretty conservative externally commissioned projections show Australasia running out of international capacity before 2020. A project like Pacific Fibre takes a long time to set up, and a long time to build, but like the NBN and UFB we’ll all be glad it is there in 2020.

    • Lance – a well reasoned argument and, yes, I concede that there will be an increasing requirement for ever faster connectivity. But here’s an analogy… there’s an ever increasing availability of effective and staggeringly expensive drug treatments. In a resource constrained economy however the government needs to make priority decisions on what to fund. Herceptin vs UFB. More MRI scanners vs RBI. We can but assume that they were sure that UFB was worth forgoing expenditure in other areas….

      • Chris McHarg |

        Ben, your analogy is understandable but I think it’s invalid.

        UFB is a one time process to replace aged infrastructure that is already being used is ways that wasn’t designed for and can’t do well with something more suitable over a period of nearly ten years. Pacific Fibre will be the first to break a monopoly and the first the bring real redundancy to international connectivity to NZ. We’re just getting started and your analogy refers to long term and spiraling waste. Your argument, while does state a truth that we must always be conscious of, is just out of place in the context of UFB and Pacific Fibre.

      • If you are going to rank Government funding, I would rather have the UFB than the NACT holiday highways.

  • Ben,
    I can’t contribute much to the business case, but I do know that school kids are hanging out for it to help with their learning.
    Watch this ten yr old Samoan boy and his mates talk about what they are trying to do and at least think about supporting it for schools 🙂

    • Thanks Dorothy. Yeah, I saw that video. Comes down to priorities is all – high speed e-learning versus high cost drugs. Fibre to the home versus more cops or whatever…

  • Just had my say over at Lance Wiggs.

    I agree Ben, it’s not about speed. Sure, maybe in the future, but maybe we also need to think differently. We’re a small geographically constrained country. Mobile would makes more sense, in my opinion.

    But before any of that happens, data costs are the elephant in the room. No point having high speeds if you can’t afford to download anything.

    And then, where do our current productive sectors lie (tourism and agriculture), in rural areas primarily, currently outside the UFB deal (putting the RBI – Rural Broadband Initiative, aside for now)

    And I love the arguments that we need to be a knowledge economy etc etc so we need UFB, and then government does stupid things like consider removing deductions on unsuccessful software projects. Whose never had one (many) of those?

    • Chris McHarg |


      “Maybe in the future”? – Are we not talking about the future here? UFB is a long term infrastructure project taking 10 years to implement.

      Not about speed? – You go back to your internet speeds 10 years ago then.

      “We’re a small geographically constrained country” – Absolutely! It is full national and international internet infrastructure that can diminish that constraint.

      “Data costs are the elephant in the room” – Absolutely! It is Pacific Fibre that is trying to fix that.

      • Chris,

        Some points.
        1) if you were only interested in diversity, wouldn’t running a cable to the Sydney NOC with its 8? landed cables a better play?
        2) Data costs are made up of both the cost of the infrastructure, AND the cost of the content, something PF isn’t going to be able to radically change

  • Chris McHarg |

    Empirical evidence? Well, try this one. Would you be willing to go back to using the internet infrastructure as it was 10 to 30 years ago. That is what you are arguing people 10 to 30 years from now should do. I think you and a number of people you know would be out of jobs.

    I haven’t read the report you quote in your article, but their conclusion destroys any credibility as soon as you take ‘time’ into account. What does ‘dial-up’ even mean? It refers to a RELATIVELY slow speed for that time, probably 56Kbps. When I bough my first modem at 14Kbps, 56Kbps would have been ultrafast! That your 1Mbps will one day be the ‘dial-up’ equivalent of now is a forgone conclusion. The report itself has established the an increase of speed “appears to raise firm productivity”. It is just a matter of timing.

    Your argument also implies that investment in the creation of a modern and dependable communication infrastructure in NZ is mutually exclusive to investment in other social areas (that is, it has no positive flow on effects to those areas). Do you even really believe this yourself?

    No one would argue that investment in communication infrastructure doesn’t need to be balanced with other investments. But for someone in your position to say they need “evidence” of benefits first, when there already is an established history to look back on and when we are talking about such long time frames, I have to wonder what made you lose your context.

    To be honest, I’m just so disappointed that this blog post came from you, in the current CHCH context. Surely, now more than ever, it is a time for lighting candles, for creating possibilities.

  • Falafulu Fisi |

    Dorothy said…
    Watch this ten yr old Samoan boy

    Kids using the internet is a distraction and time-wasting. I’ve warned some parents (relatives of mines) to only allow their children perhaps 2 hours at most per week on the internet. Going on youtube and facebook doesn’t increase kids brain power and that’s an undeniable fact. The students’ level of achievements from 2 or 3 decades ago are almost the same as today, but those past students’ generations didn’t have internet, computers, calculators, etc,… Computing Technology has grown on an exponential level in the last 3 or 4 decades, however the level of student’s intelligence of today barely change much from those days.

    I challenge any teacher here to point me out to any study, etc,…, which established that more time kids spend on internet, the better their marks? I say it is nonsense. The less time kids spend on internet the better for their own education development. Their brain can be stimulated more by playing lego rather than wasting their time on internet (unless they’re using the internet to further their learning).

    Dorothy, here is an island kid that doesn’t spend more than 3 hours per week on the internet. He concentrates on learning rather than wasting time on the internet.

    Algebra/Calculus Kid

    PS : I’m an islander, so I have warned lots of island parents (who are naively thinking that more time on the internet for their kids will improve their learning) to limit their children access to the internet because it is a distraction.

  • Laurence Hill |

    Okay here is my gripe with the Government and ultrafast broadband in a time of massive economic upheaval and when the government is telling us to spend responsibly the said look at us a zero budget look at what we’ve saved isnt it incredible … oh and by the way we are blowing it all on Ultrafast broadband roll out granted it will be helpful in the long term but in the short term it makes no sense we needed a negative budget all we got was fiddling with the figures so the budget looked good but in reality we saved nothing!

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.