Australia has spent $7.3 billion on its highly touted National Broadband Network, but made fiber available to only about 260,000 premises, which works out to over $28,000 each. Nearly a decade ago, it had a chance to take a different path.
In 2005, as a consultant for Australian phone company Telstra, I helped CEO Sol Trujillo (who had previously run US West) and his top advisor Phil Burgess (0) develop the company’s proposal to trade regulatory liberalization for rapid deployment of a Fiber-to-the-Node (FTTN) infrastructure that would have brought next generation broadband to 98 percent of Australian households before the end of the decade. Under the plan, Telstra would cover 87 percent of households on its own nickel, with the government pitching in $2.6 billion to extend the build out into rural areas. The catch: Australian regulators would agree to follow the U.S. model of exempting next generation networks from network unbundling requirements, thus creating incentives for intermodal competition from cable and wireless.
The answer: No sale. Australian bureaucrats and politicians decided instead to heed the cries of access seekers and follow the European approach of “encouraging competition” through massive regulatory intervention. Unbundling requirements on Telstra actually increased, as regulators forced structural separation and refused to grant regulatory relief even in major urban areas where there was competition from cable. Wholesale rates – the prices at which Telstra was forced to lease its last-mile connections to competitors – were so low that Optus, its main competitor, often chose to serve customers using Telstra’s telephone network rather than its own cable infrastructure, even when the Optus cable was already running down the street in front of the customers’ premises.
Telstra’s response: Forget next generation fiber and instead focus on building the world’s most advanced HSPA+ wireless network, delivering 42 Mbps to nearly all Australians.
Then, in November 2007, Australia held an election in which broadband was a major issue (1). The incumbent Liberals (the more conservative of Australia’s major parties) promised to stay the course, with more regulation and a modest subsidy program to support rural deployment. Labour promised more: A government-owned National Broadband Network (NBN), with super-fast broadband for all but the most rural households. Labour won in a landslide (2), and for the next six years (until Labour was voted out, in September 2013) the NBN became world’s largest, purest experiment in government-owned broadband.
Things have not gone well. After more than a year of false starts and procedural delays, the government created the NBN Corporation (“NBN Co” for short) modeled explicitly on the Australian post office (no kidding (3)), with the mission of spending AUS$43 billion (That’s about $40 billion; the dollar figures quoted below are all in Australian dollars.) constructing a gigabit capable Fiber-to-the-Home (FTTH) network to 93 percent of Australian households. It took until December 2010 – more than 3 years from the time of Labour’s landslide election – to come up with a business plan. Since then, the NBN has been plagued with repeated delays and cost overruns, all of which played a significant role in Labour’s defeat by a resurgent Liberal Party in last fall’s elections. An NBN Co report issued after the election (4) concluded the project was years behind schedule and $30 billion over budget.
A new report (5) issued last month by NBN Co’s newly-installed management reveals more about just how badly things have gone. The company has spent $7.3 billion (6), but passed only 350,000 premises, or less than five percent of Australia’s 7.7 million households, and only about 260,000 premises are “serviceable” (meaning that you can actually order service). Uptake is running far below expectations. The NBN has only 78,000 fiber subscribers – which at $7.3 billion works out to $93,000 per subscriber.
But at least those 78,000 people have ultra-fast broadband, right? Gigabit connections where web pages download instantly and physicians can conduct brain surgery from the comfort of their living rooms? Not really. As it turns out, only 23 percent of NBN’s fiber subscribers choose the company’s 100Mbps/40 Mbps offering, with the remaining 77 percent choosing 25Mbps/5Mbps or slower services. A little algebra reveals that only about 25,000 customers subscribe to a service faster than what is commonly delivered over current generation DSL connections – at a cost of about $300,000 each thus far.
And the hole is getting deeper, because NBN Co. is losing money at an astonishing pace. To quote directly from the February report, “the NBN Co Group generated an operating loss after tax for the six months to 31 December 2013 of $715.8 million and telecommunication revenues of $21.8 million.” As my colleague Bronwyn Howell has explained (7), there are both engineering and economic reasons to believe it won’t get better anytime soon.
The new government is now wrestling with what to do with Australia’s digital pink elephant. The current plan seems to be to stay the course, but with a less ambitious plan (8) that relies more heavily on FTTN, thereby lowering the overall cost. In the meantime, however, and despite the government’s best efforts to kill it, competition appears to be breaking out again, with an upstart company threatening to start delivering a cheaper, better service (9). Telstra, the old incumbent whose wireline infrastructure was bought by NBN Co, is also knocking at the door. Thus, the government has a chance to return to the pro-competitive path it rejected nearly a decade ago.
All of this has to be seen in the context of the continuing advocacy for government-owned broadband networks in the U.S. Proponents of such policies have repeatedly held the NBN out as a role model for the rest of the world. Shortly after taking office in 2009, then White House broadband advisor Susan Crawford declared herself “fascinated” (10) with the Australian experiment and touted the adoption of a U.S. version (11) in the National Broadband Plan; she has continued to hold Australia out as a model (12) ever since. For some proponents, the notion of a government network stirs feelings of national greatness. In a 2010 blog, Public Knowledge’s Harold Feld lauded the NBN (13) and pointed to America’s decision not to emulate it as evidence of American decline. It is commonplace for certain American politicians (14) to refer to government investment in broadband infrastructure in similarly romantic terms.
As the Australians (and others (15)) have learned the hard way, however, there is nothing romantic about pouring billions of dollars down a broadband rathole – especially when, as the U.S. experience has amply demonstrated, the real path to better broadband lies in letting the market work. (16)