Some time ago I blogged about the Australian National Broadband Network (NBN), the centrepiece infrastructure policy of the current Labor government. I'd like to follow up on this issue from a different perspective. In the meantime, both the government and opposition have dedicated themselves to a national broadband policy. So I'd like to analyse the issue in this new context. In this post I will no longer ask the question "should the government build a national broadband network?", but rather "given that both sides of the House have committed themselves to a national broadband scheme, which is the superior model?". I'm firmly of the opinion that the Government's NBN policy is by far the superior model.
First let's compare the Government's and the Coalition's models. The Government's NBN policy will roll out optical fibre to almost every premise in the country (93% fibre coverage, with various other technologies, such as satellites or wireless, reaching the remainder that are remote and inaccessible). This is the so-called 'fibre to the home' (FTTH) approach. It will guarantee 100Mbps downstream bandwidths to all areas covered by fibre, and is easily upgradeable in the future to 1Gbps speeds (indeed optical fibre is capable of far more than this). The Coalition's scheme on the other hand relies on 'fibre to the node' (FTTN) technology, whereby fibre is rolled out to cabinets on the street corner, which are subsequently connected to individual premises using existing copper cables. The Coalition claims this will guarantee 25Mbps downstream speeds, but will be much cheaper than FTTH.
Let's begin by considering the cost issue. The Coalition criticises the NBN as being too costly, claiming that their FTTN approach is vastly less expensive (the Coalition claims $17b less). If we work off the assumption that the copper infrastructure 'comes for free' then this might be a reasonable claim. But it doesn't. The reality is that the Australian copper network is nearing the end of its lifetime and will be in need of complete replacement in the near future followed by ongoing maintenance. To my knowledge, this cost has not been factored into the Coalition's estimates, which significantly underestimates the total long-term cost of the network. Fibre has a very long lifespan - on the order of at least half a century. This is not the case for copper, which deteriorates very rapidly, requiring constant maintenance or downright replacement. I suspect that once this is factored into the pricing, the Coalition's plan will not be quite as cheap as touted. Telstra currently spends $1b per year maintaining their copper network. Accumulate that over the life expectancy of the NBN and you've got a hell of an expense on the order of $50b for maintenance alone. Then there's the energy consumption cost. Powering optical fibre is very cheap - light doesn't take much energy to produce and transmit. Copper on the other hand uses electrical signals, which, when deployed across the entire country, adds up to a very hefty electricity bill (according to one estimate I read, such a copper network would require the equivalent of at least a whole coal-fired power plant to drive). To my knowledge, this has also not been factored into the Coalition's estimates. In summary, it's highly debatable whether, all things considered, the Coalition's plan will actually be cheaper in the long term. But let's for a second give Abbott the benefit of the doubt and assume that he's spot on in estimating that his FTTN scheme is $17b cheaper than Labor's FTTH. With a population of roughly 22 million, and amortised over a life expectancy of around half a century, this amounts to $15 per person per year in net savings (admittedly not accounting for compound interest or return). This is a pretty small additional price to pay for an immensely better network, which almost certainly has economic multiplier effects worth well in excess of $15/person/year. The economic arguments being touted by Abbott and Turnbull seem like lunacy.
Next there's the bandwidth issue. The Coalition themselves admit that their FTTN scheme guarantees vastly inferior bandwidths compared to FTTH. In the case of FTTN they guarantee 25Mbps downstream (which can only be guaranteed if you're living right next to a node, and deteriorates exponentially with distance from the node). This is barely more than what lots of existing broadband customers can access with today's infrastructure. Furthermore, it's not upgradeable, as the 25Mbps figure all but saturates what's possible with copper technology. The Government's FTTH scheme on the other hand will guarantee 100Mbps downstream, which, as mentioned earlier, is easily upgradeable tenfold (and probably more) in the future. There are fundamental physical reasons why copper will never achieve these kinds of speeds (electrical channels are subject to capacitive coupling, interference and resistive loss - light isn't). Thus, if one of the objectives of a national broadband policy is to be future-proof then the Coalition's plan is dead in the water.
The Coalition's broadband policy seems incredibly shortsighted. We need to factor in Moore's Law - the exponential growth in demand for computing power and bandwidth. Tony Abbott and Malcolm Turnbull have stated that their alternative caters to today's demands (Turnbull: "The Coalition plan would meet current demand for broadband services"). While a couple of tens of megabits (at best) may satisfy today's needs, it most certainly doesn't satisfy tomorrow's, and it strikes me as myopic to base a major technological infrastructure project worth tens of billions of dollars purely on today's needs. The advent of ultra-high-def (4K) video will already saturate the bandwidths being promised by the Coalition, not to mention applications in 10 or 20 years time (e.g. future developments in cloud computing or higher-def, multi-channel video). By the time the Coalition's infrastructure is complete, it is likely to already be obsolete technology. If we're going to spend tens of billions of dollars on such an infrastructure project, then the most pressing requirement should be that it caters for tomorrow's needs, since this infrastructure, being as expensive as it is, should last us decades, not months.
If we're going to invest this kind of money in such infrastructure, then we should only pay for the infrastructure once. The coalition's plan will require paying for it over and over again as the copper network deteriorates, until, ultimately, people realise that it can't provide the bandwidths we need, at which point we're going to have to reinvest in the infrastructure from scratch and roll out FTTH anyway. So why not just do it right in the first place?
The final issue I'd like to touch upon is that of competition. The Coalition consistently criticises the NBN for being an uncompetitive monopoly. I heavily dispute this. Inevitably such infrastructure will be a natural monopoly. It makes zero sense to have half a dozen fibre lines running into each household, each owned by a different provider, to compete with one another. It would be hugely economically inefficient since the majority of it would be unused (of course, if secondary providers do decide they wish to run additional cables into people's households, they shouldn't be legislatively prevented from from doing so, but I can't see this happening). So the best we can hope to achieve is to maximise competition within the context of this natural monopoly. The way the Government intends to achieve this is by structurally separating the wholesale and retail divisions of the NBN, such that the infrastructure is owned by NBN Co., but they don't have the right to sell it to individual consumers. Rather, there is a level playing field in which third-party retailers can purchase bandwidth wholesale from NBN Co. and resell it to the consumer. This is exactly what's being proposed by the Government. Under the proposed scheme, there will be no barrier to market participants purchasing bandwidth wholesale, so that even small competitors will be able to enter the broadband market. This will create the closest to a competitive market that we can realistically hope to achieve with such a project. A broadband market with a level playing field in which even small competitors can compete is a pretty decent deal.
Laughably, Tony Abbott recently said it's a mistake to put all our eggs in the one basket (i.e. spend all our money on fibre as opposed to spreading the investment across a diverse range of technologies). This is an absolute joke. When it comes to traditional investment and portfolio management theory, certainly the ethos "don't put all your eggs in the one basket" is a very wise philosophy. But when it comes to technological infrastructure, this doesn't make any sense whatsoever. Surely it makes sense to choose the superior technology (fibre) and use it universally rather than investing in a mix of inferior technologies (copper) just in the name of "not putting all your eggs into one basket". Should we equip school computer laboratories with a mix of cutting edge PCs and 1990's 386 computers, just because we don't want to put all our eggs into the one basket? No, we should just choose the best technology and employ it universally. Should we equip half of our defence forces with Soviet MiG fighter jets for the sake of diversity, or should we just universally adopt the latest NATO fighters? It's a no brainer.
Given that both sides of politics have committed themselves to investing in such infrastructure using public money, it makes zero sense to choose the technology to cater only for today's needs, which has to be continually replaced and upgraded, and which doesn't cater for tomorrow's needs.
The NBN is inevitably going to be one of the big policy issues determining the upcoming federal election in September, and the fact of the matter is that the Coalition's policy is a joke - it's much (much) slower, it's almost certainly not cheaper, it's not at all upgradeable (unless we abandon the copper and switch to fibre), and it doesn't accommodate for tomorrow's needs. Why waste the money?