Conroy gets his NBN comeuppance

The cost-benefit analysis of the national broadband network tabled in Federal Parliament on Tuesday evening has produced precisely the conclusions that critics of Stephen Conroy’s decision not to commission a similar analysis would have predicted.

From the moment it was announced it was obvious that the costs of building a gold-plated fibre-to-the-premises network would far outweigh the benefits and that over-building perfectly useful existing HFC and ADSL2 networks would be a large-scale waste of taxpayer funds.

NBN Co’s own strategic review put the cost of Conroy’s NBN at $73 billion versus $41 billion for Malcolm Turnbull version, which uses a multi-technology approach that includes fibre-to-the-node, HFC, ADSL, satellite and fixed wireless.

The Vertigan Committee’s approach is different -- it incorporates benefits as well as costs -- but essentially buttresses the conclusions of NBN Co’s strategic review.

There’s never been any dispute that Australia needs higher and more ubiquitous broadband speeds. The core question is what approach delivers fast-enough speeds for the foreseeable future at a cost to taxpayers that can be justified.

Given that the biggest cost within Conroy’s fibre-to-the-premises network was the connection between exchanges and premises -- mainly households -- the nub of that question is what households might need and be prepared to pay for over the next decade or two.

The Vertigan Committee commissioned modelling of household demand for bandwidth and come up with some unsurprising results. The median household -- 50 per cent of households -- would need/use 15 Mbps by 2023. The top 5 per cent of households would require 43 Mbps. The top 1 per cent would need 45 Mbps. The top 0.01 per cent would need 48 Mbps.

That might suggest to the cynical that building a fibre-to-the-premises network that delivered 100 Mbps-plus speeds to households might be somewhat (read massively) over-the-top and wasteful.  

Modelling that confirms the obvious -- that consumers’ willingness to pay for higher speeds falls away dramatically as the bandwidth and its cost increases -- reinforces that conclusion.

Thus, had Conroy’s vision been realised (which it now won’t) we’d have ended up with an extremely expensive network capable of speeds that only a fraction of users would need or could use or would be prepared to pay for.

The committee also made the point -- which has been made consistently by critics of Conroy’s NBN -- that the vision he had of the NBN delivering e-health and e-education services doesn’t require 100 Mbps broadband. They are, in fact, low-bandwidth applications that are being delivered today without an all-fibre NBN.

The committee’s reference case for its cost-benefit analysis (done to enable comparisons with the alternatives rather than because there is any prospect of it occurring, given the billions already committed and the $12 billion of payments, in net present value terms, already contracted with Telstra and Optus) was an unsubsidised rollout.  

Given that simply halting the rollout is estimated to cause negative net benefits of $24 billion, the analysis does confirm that there is substantial net benefit in having a national broadband network, so the question is what type of network produces the best outcome in terms of costs and benefits.

In metropolitan areas, if there were no national broadband network at all -- if there were an unsubsidised rollout -- it would be reasonable to assume that the private sector would have continued to upgrade infrastructure and bandwidth in line with demand because it would have been profitable to do so. In rural and regional areas, however, there would have been no private investment in ubiquitous broadband because it would have been uneconomic.

The NBN, whatever technology were used, has to have an element of subsidy/cross-subsidy in it -- at a cost of around $5 billion -- to provide reasonable broadband speeds to rural and regional Australia.

In the multi-technology mix version of the NBN that NBN Co is now committed to there is a negative net benefit of $6.1 billion relative to the reference case and, if the NBN were fibre-to-the-premises, negative net benefits of $22.2 billion, with costs of $17.6 billion more than those of an unsubsidised rollout outweighing benefits that would be $4.7 billion less than the reference case.

That’s because a FTTP network would take significantly longer and cost substantially more than the more pragmatic Turnbull approach of using a mix of technologies and using existing infrastructure.

The committee also confirms what Turnbull has said in the past and NBN Co’s strategic review concluded. In almost every conceivable scenario it would be cheaper to build a multi-technology NBN and subsequently upgrade it to fibre-to-the-premises than to stick with a FTTP rollout (because of the extra time and cost it would take to deliver the higher speeds of an FTTP network).

Conroy and the FTTP zealots argued that there was no need for a cost-benefit analysis because the future demand for broadband and applications for which it would be used were unforeseeable -- the benefits would be incalculable -- and therefore any attempt to assess the costs and benefits would be meaningless.

The Vertigan Committee has demolished that argument, too, and provided a template approach for analysing future “build it and they will come” proposals (whoever ‘they’ might be and however many they might or might not be) to spend billions of taxpayer dollars in the process.