House debates

Wednesday, 19 June 2013

Matters of Public Importance

National Broadband Network

4:20 pm

Photo of Paul FletcherPaul Fletcher (Bradfield, Liberal Party) Share this | Hansard source

I am very pleased to rise to speak in this matter of public importance debate. The premise of this is that internet traffic is growing at a very sharp rate and therefore we urgently need the current deep fibre infrastructure build. That is the premise of the matter before the House this afternoon. In using the term 'deep fibre' we presume the member for Lyne is referring to a network architecture in which fibre goes all the way to the premises rather than to an intermediate point, such as the node. As the member for Lyne has made clear, the premise draws upon the forecasts that are provided regularly by Cisco. They most recently said that world internet traffic is expected to reach 1.4 zettabytes by the end of 2017.

I want to make three points in the time available to me. The first is that this matter for discussion draws too long a bow from what is contained in the Cisco forecast. The second is that, even if you accept the logic of this, it is predicated upon the assumption that the current build is going to be delivered on time so as to achieve this magnificent increase in capacity by 2017, when all the evidence demonstrates that that is a hopelessly unrealistic expectation. Thirdly, I want to make the point that the coalition's proposal to build out a national broadband network using fibre to the node and, to around 20 per cent of premises, fibre to the premises will deliver the necessary increases more quickly and more cost effectively.

Let us start with the proposition that the matter of public importance draws too long a bow from the Cisco forecast. I note that Mr Boal, of Cisco, a local executive, is quoted as saying quite specifically:

… we're agnostic on the access technology but clearly significant increase in capacity is required—

So any direct connection between the Cisco forecast and fibre to the premises or fibre to the node or any other technology, any other network architecture, is drawing too long a bow.

More interestingly, what Cisco says is that the average broadband residential speed in Australia today is nine megabits per second and by 2017 it will be 39 megabits per second globally and therefore, by implication, a very serious problem, we need to get our skates on et cetera. The point, when you dig into the data, is that the world speed is assumed to grow very, very sharply. Today the world speed, according to Cisco, on average is 11.3 megabits per second, not much higher than Australia's, and it is to reach that global target of 39 megabits per second by 2017. So the implication that Australia is in some way significantly behind the rest of the world does not follow from the Cisco analysis. In fact, what the Cisco analysis shows is that this is a set of forecasts which assume a very sharp growth in traffic volumes predicated in turn on an assumption that the average bandwidth and the average speed available, not just in Australia but also around the world, is going to rise very sharply.

Despite the sense of looming crisis that the member for Lyne was keen to generate, if you look carefully at what Cisco says, one of its points is that the rate of growth of internet traffic is slowing. It notes that global internet traffic increased fourfold in the last five years and in the next five years it will increase threefold. No dispute: these are still spectacular rates of growth, but the actual point is that you are seeing a slight reduction in the rates of growth.

But let us talk specifically about methodologies used in the Cisco forecast. What the Cisco paper says is:

… assumptions are tied to fundamental enablers such as broadband speed and computing speed.

Indeed, if you dig into the detail their methodology is to start with a number of users that they assume will be in place by 2017, then a number of minutes of video that the average user is assumed to watch, and then an assumed number of kilobits per second, which it is assumed that watching video will require. The point is that the video that you will watch will depend upon the infrastructure that is available. If you plug into a forecast a set of assumptions about bandwidth available not just in Australia but also around the world, then that will generate a set of forecasts about traffic volumes. But the traffic volumes here are the dependent variable, so I am afraid the member for Lyne has got his logic the wrong way around. This MPI is flawed in logic. It is saying we will have a problem because the volumes will be too big for the pipes. In fact, the volumes depend upon the pipes. That is quite clear from looking at the methodology that Cisco used.

One of the assumptions behind the MPI is that an increase in usage will follow when the speed that is available increases. Another Cisco paper notes that there are significant qualifications on that assumption. It says:

… there is often a delay between the increase in speed and the increased usage, which can range from a few months to several years.

Now the other thing that is significant in what the member for Lyne had to say was that he again gave us this article of faith from supporters of NBN Co. that one of the key reasons that we apparently must have two-way fibre is because upload speeds are jumping and jumping and jumping and there will be a need for symmetrical services. Nobody contests that the upload proportion of data is growing. But what is very interesting is that the Cisco report says:

With the exception of short-form video and video calling, most forms of Internet video do not have a large upstream component.

As a result, traffic is not becoming more symmetric as many expected when user-generated content first became popular.

That is a really critical point because that goes to one of the premises that we are constantly told underpins the fibre-to-the-premises model of the current government—that is, traffic is becoming symmetrical and we are hopelessly in the Stone Age if we do not immediately introduce fibre everywhere to respond to that. What Cisco is saying is that is not right; the assumption that data is increasingly symmetrical is actually not proved out by the data. But, unfortunately, so often in this debate about broadband we hear wafty assumptions that are not backed up by the data.

Let us be clear: the coalition are strong supporters of an improved broadband infrastructure. We are strong believers in the social and economic benefits that follow. Personally, I have worked in policy and in the private sector in this area since the mid-90s. I am a passionate believer in broadband and I am a passionate opponent of the wasteful, ill-conceived, poorly constructed broadband plan that the present government is pursuing, and I am deeply grateful that the coalition is pursuing a rational, cost-effective plan that will deliver broadband more quickly to most Australians.

One of the problems with the current government's plan, and this is turning to the second point I wanted to make, is that the implementation is absolutely hopeless. Despite the broad and wafty aspirations that the member for Lyne is articulating, let us look at the numbers. The first corporate plan of NBN Co. said that by 30 June this year there would be 1.3 million premises passed by fibre. That was then wound back just a little bit, with the second corporate plan that came out midway through last year saying the number would be 341,000. That was then wound back again, earlier this year, when the company issued a revised forecast that said: 'Whoops! Sorry, we're not going to make either the first number of the second number. We are now going to make a third, much lower, number, which will be somewhere between 180,000 and 220,000—but don't hold us to the numbers because we're visionaries.' And then we have the actual number, as at May this year, which is about 104,000. Members of the House: when we put aside the wafty aspirations and look at the hard numbers on what is actually being delivered, this government's plan is not advancing Australia towards where we need to be—even if we accept the premise from the member for Lyne that we have a bandwidth crisis and we must immediately respond with pressing urgency.

Let me turn, thirdly and finally, to the merits of the coalition's plan which will deliver speeds to all Australians of between 25 and 100 megabits per second by 2016 and 50 and 100 megabits per second by 2019. The coalition's fibre-to-the-node proposal will be rolled out more quickly and with less variability and in a more reliable fashion than the chaotic mess we have seen from a Labor Party which jumped into an over-ambitious plan driven by political motivations without properly analysing what they were getting into. Is it any surprise that when you had a network that was contrived for political reasons the actual execution of it turned out to be completely woeful? So if our policy objective is to build a broadband network which bests supports our prospects of meeting the agreed growth of data that will occur by 2017, the coalition's broadband plan is far and away the best and most reliable way to do it.

Comments

No comments