An economic analysis of the likely costs of the planned National Broadband Network suggests that end retail prices would have to exceed $200 per month for the project to return its costs and pay standard capital returns.

Communications Day asked Concept Economics to run the numbers on what is known to date about the planned network—estimated as costing up to $43 billion with fibre served to 90% of premises and wireless & satellite solutions covering the rest.

Chairman Henry Ergas told CommsDay that the end results are extremely sensitive to take-up rates. Assuming a passive optical network configuration, current retail margins and a committed information rate of 2 Mbit/s—which at a contention ratio of 50-1 provides for end user speeds of up to 100Mbit/s—the Concept calculations suggest that at an eventual 80% take up rate using a standard S-shaped diffusion process, national retail prices for the NBN would be around $215 monthly. If take-up was just 40%, the national price rises to $380.

If the prices were de-averaged, urban prices would fall by around 30%, but rural prices would have to rise by around 160%.

Concept’s estimated prices are also commensurate with international experience—Verizon’s FIOS service, which provides for 50Mbps speeds in predominantly urban areas of the north eastern US, generates a monthly ARPU of US$129 or A$185 at today’s exchange rates, albeit with only around quarter of the market taking the service up.

Asked by CommsDay if government plans to swap existing private assets for equity in the NBN venture could reduce its costs and end user prices, Ergas suggested not. He said Concept’s estimates already assume the NBN operator leases 80% of its metro backhaul and 20% of its rural backhaul, adding he was not sure why anyone would swap assets for less than their fair market value.

“Moreover, taking over existing assets might not be all that commercially attractive to the new entity. For example, the state government networks are very diverse, and they differ substantially from a local distribution network such as TransACT. Putting them and others under the NBNco umbrella would cause NBNco’s assets to resemble the Ansett fleet, a melange of highly diverse assets of greatly varying structure, vendor, age and quality,” said Ergas.

“This would increase operating costs, likely materially, though it is difficult to know by precisely how much. Ansett used to claim an operating cost penalty of 30% associated with its mixed fleet composition. Whether a cost penalty of this level would apply in these circumstances would need to be tested.”

SOCIALISE LOSSES? Asked by CommsDay if prices could be reduced if the government was able to keep prices low and move losses onto taxpayers, Ergas thought this was not possible.

“The Competition Principles Agreement [between state and federal governments-Ed] requires government owned entities to earn returns commensurate with those required by their privately owned competitors,” he said. “NBNco is assumed to have a 15% weighted average cost of capital which the European Commission has suggested is a reasonable weighted average cost of capital for NBNs.” The Concept calculations also estimate that by using a gigabit passive optical network configuration, the total capital cost is likely to be in the order of $40-50 billion.

“Adding operating costs over the construction period leads to around $60-70 billion. If start up losses are capitalised, then the costs are obviously greater,” Ergas said.

COULD CONTENT PAY? Asked if some of the revenues currently accruing to content that might be carried over the NBN—such as television and Internet advertising—could be employed to reduce prices, Ergas said: “Clearly, access to free to air content will not be a demand driver for the NBN, as that content is available for free on a close to ubiquitous basis. Foxtel and Austar are unlikely to accept lower revenues from the NBN then they could earn outside it, though they might use the NBN to widen their footprint.”

He continued: “The most attractive scenario for the NBN would be if it could split advertising revenues with content providers, say by giving subscription content providers a wider footprint in exchange for some share of their incremental advertising revenues. These third party charges paid by advertisers or other third parties would help reduce the NBN’s required retail prices.” But these sums would need to be large, he added, pointing out that the entire revenues of the existing free-to-air TV industry only equate to one-third of the likely operating costs of the NBN.

AUSTRALIA’S WORLD BEATING SUBSIDY: CommsDay also posited that the government might be able to raise its contribution to the NBN to reduce prices, but Ergas said that the current proposed government contribution was already the highest proposed in the world—twice that of the next highest, Greece, and considerably higher than other countries with stimulus policies such as the USA and South Korea. He also suggested that the NBN not only be viewed in terms of whether its benefits exceeded its costs, but whether alternatives might generate superior benefits.

“Even if the benefits of the NBN were greater than its costs, if the net benefit would be even greater with say an FTTN network instead, the NBN being an alternative to the FTTN, then evaluated properly, the NBN would fail a cost-benefit test, as the incremental benefits of the NBN compared to FTTN would be less than its incremental costs. In other words, costs and benefits need to be defined in opportunity cost terms, relative to the set of relevant alternatives.”

Concept’s calculations were performed by a team of its economists including Ergas, Dieter Schadt, Alex Robson, Garth Crawford and Geoff Petersen. Although Concept performs consultancy work for clients such as Telstra, Ergas says the results should not be imputed to them.

Ergas also pointed out that Australia’s proposed NBN subsidy is unusual in world terms. “Very few governments are spending large amounts of taxpayer money on deploying FTTP networks. In the EU, rules on subsidies to public or private commercial operations limit subsidies to situations where there is a clear market failure. This has meant that subsidies for broadband have largely been confined to service provision in otherwise uneconomic areas.”

Grahame Lynch