Australia’s NBN too expensive: EIU
Compared to what?
The headlines say all you need to know, surely? As reported all over the place, Australia’s NBN (National Broadband Network) has been rated as too expensive and relying too much on government support, according to the Economist Intelligence Unit.
Australia cherishes a characteristic called “cultural cringe”. A full explanation would occupy a PhD thesis, but one symptom is this: if someone overseas says “you’re hopeless”, our media will respond “yes, we are!”
Whether journalists tag the network’s price at $36 billion, $43 billion or $27.5bn (the government’s commitment), having the EIU slam the NBN is too tasty to resist.
And because Australians love an appeal to authority – we still sweat and fret over the deliberations of the same credit ratings agencies that gave junk mortgages the “AAA” stamp, thus sparking the GFC (global financial crisis) and beggaring local governments throughout the country – the report, which rates government broadband initiatives around the world, is accepted as holy writ.
Australia is rated poorly chiefly on the basis of the NBN’s cost to government. According to the Sydney Morning Herald's inaccurate headline, the network is “one tenth the speed” of South Korea’s network “at 24 times the price”.
South Korea is currently implementing an upgrade to its network to 1 Gbps speeds, whereas the NBN currently assumes a maximum end user plan speed of 100 Mbps (the network has already been tested at 1 Gbps).
Unable to contact the EIU by telephone due to time-zones, The Register put three questions to the organisation’s Iain Morris by email. We reproduce the questions and answers below.
El Reg: Reports have stated that the EIU calculates the Australian government NBN commitment at 7.58 per cent of total government revenues. If these reports are accurate, can you please outline the basis of this analysis?
Iain Morris: The 7.58 per cent figure is based on taking the public-sector commitment of the plan (A$27.1bn in Australia's case) as a percentage of annual government budget revenues for 2009 – the last year for which actual data was available when the report was being produced. It's simply a benchmark that allows comparison between countries. There is another benchmark in the report showing total public-sector spending as a percentage of annual fixed-line retail revenues in 2009, which also makes Australia stand out for the size of its commitment.
El Reg: Can you please identify which South Korean broadband programs are included in the analysis, and their value?
Iain Morris: It includes both the initial fibre network and the current initiative (upgrade to 1Gbps). The current plan has US$1bn of public funds versus about US$26bn of private funds; the earlier plan also had similar splits between public- and private-sector funds.
El Reg: As Senator Conroy pointed out (a calculation which I examined as an analyst in 2006) South Korea combines both high population density and small geography. Specifically, its 44 million (roughly) population resides in an area less than 1/8 that of New South Wales (population 7.24 million). How is such a disparity resolved in the EIU analysis?
Iain Morris: The Australian government is quoting population density figures to defend itself, but, as its NBN implementation report points out, 90 per cent of the population occupies just 0.2 per cent of the land mass. In fact, overall pop density and land area are red herrings – the cost is all in the last mile.
I should also point out that South Korea's costs are similar in total (around US$30bn plus) but the private sector has covered more than 95 per cent of the costs – both in the past and in the next phase of the plan – hence the 24 x factor, which is in relation to public spending, not total spending.
7.58 per cent of what, exactly?
The media’s treatment of the EIU’s “7.58 per cent of government revenue” measure is simply wrong; any journalist citing that figure without context is either sloppy, or an economic illiterate.
As Morris makes clear, the figure is only a benchmark. Long-term government commitments should not be compared to government revenue for a single year.
Australia’s government spend is meant to take place over eight years (at which point the government intends to sell the NBN). Australia’s spend on an annualised basis will be less than 1 per cent of total government revenues.
The EIU benchmark is useful only as a comparative measure – it offers no insight into the bottom-line budgetary impact of any country’s broadband program.
Australia vs South Korea
The EIU refers to “both the initial fibre network and the current initiative (upgrade to 1Gbps)” in accounting for South Korea’s government network spend.
This appears to overlook South Korea's very long history of government intervention in telecommunications (generally following a public / private model).
That history stretches back more than 15 years. To quote an article from 2006: “In 1995, the South Korean government made what must rank as one of the most shrewd and far-sighted investments in business history. It spent big on a nationwide high-capacity broadband network that any telecom operator could offer service on, and offered subsidies so that 45 million Koreans could buy cheap PCs”.
We do not know how much of that history – and how much South Korean government investment – has been included in the EIU's analysis.
The question of geography
Morris dismisses geographical considerations as “red herrings”, noting that 90 per cent of Australia’s population occupies 2 per cent of its land area.
Let’s instead look at population density solely in Australia’s urban areas: roughly 18 million people in 152,000 km2 (2 per cent of Australia’s 7.6 million km2 total).
Even Australia’s urban density of 118 people per square kilometre pales against South Korea’s total density – urban and rural – of 440 people per square kilometre.
On that basis, I do not consider geography to be a “red herring”.
Ignoring household size, a South Korean telco can expect any given service area to hold roughly four times as many people as a carrier can expect to find in urban Australia.
Politics and telecommunications
The Sydney Morning Herald described the EIU as “right leaning”.
The unit criticises not the total price tag of the NBN – as Morris noted, South Korea's upgrade cost is close to the Australian government's total NBN commitment.
The EIU marks Australia down because that investment is coming from government. The EIU is clear that it believes the NBN should be a private sector project.
This is not a neutral position: it follows a particular school of economic theory, and subscription to that theory is at least partly reflected in the left-right divide of Australian politics.
Such a simplistic view (public bad, private good) also ignores whether carriers are willing to co-operate in public/private projects, whether a nation's industry structure encourages that co-operation, and whether a government holds sufficient authority to enforce co-operation.
What do you get for your billions?
The most revealing response from the EIU still relates to the total price tag.
For a total budgeted cost (including private and public sector spend) of $43bn, Australia will update its entire network from copper to fibre, for 97 per cent of the population.
For $US27bn (public plus private spend), South Korea expects to upgrade an existing fibre network to gigabit speeds – excluding the initial build cost.
Last year, when NBN Co announced that its network tests had reached gigabit speeds (an announcement unfortunately made in close proximity to a Federal election), the network’s critics cited that speed as unnecessary, wasteful or even implausible.
South Korea’s private sector clearly believes those speeds justify a network upgrade that matches the total Australian government commitment to the NBN. That should, at least, put paid to the notion that “no network needs to go that fast.” ®