NBN – cost benefit analysis, cost of service, monopoly and other issues
Why should there be a cost-benefit analysis of the NBN?
The truth is that every major piece of infrastructure investment should be the subject of a rigorous cost benefit analysis. As Ken Henry famously said “government spending that does not pass an appropriately designed cost-benefit analysis necessarily detracts from Australia’s well being.”
Of course the larger the investment the more appropriate is that analysis and this is especially so when the investment is to create a business asset.
The failure to conduct the cost-benefit analysis is particularly egregious because it is this Labor Government which set up Infrastructure Australia for precisely the purpose of identifying and assessing infrastructure investments of national significance. It is difficult to think of any piece of infrastructure that would better qualify for the attention of Infrastructure Australia, but the Government has refused to allow it to consider the NBN.
So what would a cost-benefit analysis do? It would first make sure we were asking the right question. What is the problem we are trying to address? Put another way, if fibre to the home (FTTH) is the solution, what is the problem?
Unless the problem is that nobody in Australia currently has adequate broadband services to their home and nothing less than 100 mbps bandwidth will do, then the NBN FTTH network is an excessive investment.
So let us assume that the problem can be framed in this way: Many Australians have access to fast broadband services but many do not. In some parts of our big cities there are black spots where for historic network design reasons ADSL is not available. In regional and rural Australia there are many places where broadband access is very poor and certainly not comparable to that available in the cities. A lack of competition in the backhaul, trunk backbone of the network has prevented telcos from competing with Telstra in many areas.
So if that is a fair description of the problem then a cost benefit analysis would carefully consider the various options for addressing it and their cost. What would it cost to upgrade our existing network so that all residents of our cities and towns had access to ADSL2 for example?
What about 4G wireless? There seems to be a readiness by many to dismiss wireless as a broadband contender, but we must accept that people are voting with their feet (or credit cards) by taking wireless broadband at a rate that far outstrips fixed line broadband which is fairly static. The authors of the Alliance for Affordable Broadband paper are not fools (they include some of our most experienced telco CEOs) and their contention is “We believe that next generation 4G technologies are the best fit for purpose for the vast majority of consumers and SOHO clients currently without other broadband delivery options.” Read the full paper here.
The largest new national broadband initiative in the United States at present is the $7 billion LightSquared project to build a nationwide 4G wireless broadband network to provide up to 100mbps speed connectivity. See www.lightsquared.com I mention these matters not to suggest that wireless IS the answer or indeed the only answer, but simply to note that broadband wireless is a real option for many people, that its next evolution will involve much higher speeds and that to dismiss it as a non-contender (as so many do) is hardly credible. A rigorous cost-benefit analysis would give wireless very serious consideration, something the Government has failed to do.
Given neither FTTH nor ADSL will be viable in rural and remote areas, the technology solutions will be largely wireless and satellite. How best can that be delivered?
So a cost-benefit analysis would rigorously assess the most cost-effective means of meeting the perceived need. It would also ask whether this required the Government to establish a new, monopoly telecom to do so. Government support is doubtless required in rural and regional Australia and both sides of Australia recognise that, but how much support is needed in the cities? That should be rigorously assessed as well.
Another question that needs careful examination is the business case for these products. While many enthusiastic users of the Internet are salivating at the thought of higher speeds available to their home, Telstra has been selling 100 mbps broadband on its HFC network in Melbourne for over a year and so far has only a few thousand subscribers.
It is easy to see the use of fast broadband in schools, universities, hospitals, medical centres and businesses. But how many households would derive meaningful additional value from upgrading their broadband access from, say, 20 mbps to over 100 mbps?
And of course let us not forget that simply because you have a network capable of delivering 100mbps to your home or office does not mean that you will be able to download material at that speed. That will depend on the speed of every single connection between the server you are accessing and your own device.
Of course the value of such a network is not limited to the dollars and cents in the profit and loss account. We all recognise that there are spill over benefits to the wider economy from broadband connectivity. But again, while we can readily recognise the productivity benefits from being online with broadband access, what is the incremental benefit from going from 20 mbps (or less) up to 100 mbps for the vast majority of households?
What applications will take advantage of these higher speeds? It is easy to respond as Senator Conroy does by saying “Things we don’t even know about.” but it is hardly a credible basis for investment. Right now the applications which are consuming more and more bandwidth each year are video services, especially YouTube. Indeed only a few months ago I head David Thodey say that 80% of Telstra’s Internet bandwidth was taken up by video material.
FTTH offers the potential for hundreds of television channels, instant video downloads and high definition interactive gaming. That would seem to be the most likely use for the massive lift in bandwidth for most households. But if that is so, are people going to value that sufficiently to pay a higher rate per month to justify the massive investment in the NBN and whether they are or not, is making these applications more readily and abundantly available a worthy object of government subsidy? How does it rate against more investment in health, or schools let alone public transport services.
If the cost benefit analysis were to conclude that health, educational and business premises needed access to fibre then an assessment would be made of how to deliver that. As it happens all but one of our universities (Bond) is connected to a national and international high speed fibre network by AARNet. Many of our schools (not enough) are also on fibre as are many of our larger businesses. How should that fibre be expanded? And what is the most cost effective way of doing so.
And finally, given the structural separation of the network from the retail business of Telstra is cited as one of the great virtues of the NBN, let us not forget that structural separation of that network does not require its overbuild and replacement by an entirely new network. Structural separation is a consequence of the NBN agreement with Telstra, but the NBN is not a necessary companion of structural separation.
These are just some of the issues a cost-benefit analysis would consider. The Government’s failure to undertake that analysis has been rightly and widely condemned. If the NBN is the right decision then its proponents have nothing to fear.
Cost of Service
Right now we don’t know what the NBN will charge. The Implementation Study suggests an average wholesale price of $35 per month for voice and entry level basic broadband. That would suggest a retail price of at least $65 -$70 (so I am advised) which is higher than many plans currently available. It is not clear whether NBN would charge in addition to this the equivalent of a line rental fee either.
But rather than speculate about what NBN might charge, let’s look at a few facts. The first fact is that the NBN is going to involve a massive, new investment in telecommunications infrastructure. That investment is going to need to be serviced and common sense suggests that is going to put upward pressure on pricing.
Consider the recent massive hikes in electricity and water charges arising from the large investments in infrastructure made in both those sectors.
The Implementation Study acknowledges this and states at p. 254:
“The relative value of fibre is likely to increase over time as new services and uses emerge. As it does, NBN revenues should increase to provide a fair return on the network investment. An increase in revenues over time would likely correspond to an increase in user value.”
The Study assumes real growth in prices of between 0 and 2 percent, so around 1% above inflation – say 4% per annum. Now that may not seem much to some people, but for well over a decade we have been enjoying declining telecommunications costs. Now the NBN will see them increasing. Will the increase in cost be worth it? Well it may be for some people, but for many it won’t be. And as noted above, there isn’t much evidence that people will pay a premium for high speed services. Put another way, if 100 mbps is a much sought after nirvana why wouldn’t it make more sense to turn on the entire HFC network to that speed (for a very modest additional cost compared to the NBN) and see what the take up is?
In short, there is no reason to believe the NBN will see a continuation of declining Internet access charges. If the Implementation Study is to be believed and if the massive investment is to be serviced, the reverse will be the case.
Now we should not beat around the bush here. The object of the NBN is to create a monopoly. See the Implementation Study at p. 60
“Government policy in building the NBN is likely to create an industry structure with no significant competition between fixed networks, since the existing copper and HFC networks are likely to be deactivated over time.”
Too right there! The deal with Telstra will require it to decommission its copper network and not use its HFC network to compete with the NBN for voice and broadband business.
So how is that defensible? The study goes on to say:
“The trade-off is that NBN Co will offer wholesale-only equivalent access to service providers. The rationale for this is that – given the high cost of laying the fibre, the long asset life and the limited ability to differentiate – fibre has the characteristics of a natural, stable infrastructure monopoly analogous to an electricity grid or gas pipes.”
So the argument runs: yes it will be a monopoly, but that’s okay because it will be structurally separated from any retail provider and will provide access to all retail providers – as a common carrier. But that gets back to the question of structure. If the Government’s claim is that industry structure is the problem with the current arrangements, then that can be addressed without building a new network.
The establishment of this monopoly does carry with it real risks, as the Implementation Study notes at p. 446 and preceding. These include NBN expanding its activities beyond that of a wholesale carrier, operating inefficiently and lacking incentives to innovate. The Study acknowledges that the NBN approach is unique in the world.
“Competition between alternative networks has stimulated the roll-out of next generation broadband infrastructure in several countries around the world.” (p.433)
And nowhere more so than in America where the competition between the cable companies and the telcos has been especially intense.
“Cable and telecommunications companies found themselves competing to provide substitutable services and were compelled to match each other’s performance improvements.”
Remember the only reason for requiring Telstra to shut down its competitive infrastructure is to enhance the financial viability of the NBN.
In summary: If monopoly is bad, then the NBN could be built to operate as a competitor with the copper network and the HFC networks. If the Government’s claim is that vertical integration of Telstra is bad, then structural separation could occur without building a new network.