NBN Numbers Part Two: why it may never cover its costs

April 8, 2009 — 7 Comments

Cost per household of NBN (at 100% rollout): $5,063 per house.

We don’t know if this includes the cost of connecting each home. The policy says “connect each home” etc, but given this is a wholesale network, there may feasibly be an actual physical connection cost (+ equipment, unless the Government is handing out free modems/ network devices) for the average punter.

But that aside, when will it cover its costs?

At $50/ month with 100% takeup rate: 101.26 months. Nearly 8 1/2 years.

At $100/ month at 100% takeup rate, divide by two: 50.63 months. 4.2 years.

But then we have the problem that this is a wholesale network, not retail. Those reselling the network will want a margin, and it will need to be competitive with alternatives, some already being rolled out today.

At $25/ month wholesale with 100% takeup: 202.52 months, or just shy of 17 years.

At $50/ month with 100% takeup rate: 101.26 months. Nearly 8 1/2 years.

But then we need to consider that the network will most definitely not have 100% take up. In fact, 100% is an impossible figure. Knowing the rate of take up though in 2018, given we don’t know what the competitive landscape will be, is at best a guess. There are two considerations: competition and overall demand. We know the NBN will have competition from at least Telstra, possibly other companies by 2018 (and well before then.) Second, we know that not everyone takes up new technology (there are still people using Dial-Up today!). The demand builds over time, so you never get say 75% of the market demanding the option immediately; a quick look at broadband stats for Australia shows that demand for a new tech builds over time, often driven by increased competition and downward pricing.

Here’s some calculations based on wholesale rates

$25/ month at 25% takeup: 810.08 months, or 67.5 years
$50/ month at 25% takeup: 405.04 months or 33.75 years

$25/ month at 50% takeup: 405.04 months or 33.75 years
$50/ month at 50% takeup: 202.52 months, or just shy of 17 years.

Now we’ve made some presumptions on wholesale pricing vs retail. The wholesale price could be higher but the higher that price, the less appealing the network becomes, so you decrease demand. You’d also have to consider that Telstra with its competing network isn’t going to sit back and let the NBN undercut it by a huge margin either.


What the figures show is that the NBN may never cover its costs. The counter argument is that this is a nation building exercise and a huge investment by the Government, but that has a rather important flaw: the Government is building the network through a separate company they’ll majority own until sometime between 2020-2030. That company will have private investment as well, so it will have a duty to cover its costs somewhere along the line. Note also that the Government is talking about issuing infrastructure bonds; those bonds will offer repayment with interest. The money to repay these has to come from somewhere.

The problem is that the earliest the NBN might recover its costs, even under the most generous model, is 8-10 years. Realistically we’d be looking at 15 to 30 years, possibly longer. People may argue that the NBN will still be of great service in 2018, but does anyone really believe that this will still be state of the art tech in 2038 or beyond, and that nothing else will have come along that does it bigger, better and cheaper?

The numbers just don’t add up.

Update: should have added, these calculations also presume that the project will come in at cost. Government projects rarely do, so the starting cost of the network could be higher again, making it even harder to repay.