Let’s run the numbers
In the latest Telecom submission to the MED, Telecom has calculated it would take $1.5 billion to provide Fibre to the Node (FTTN).
(Johnnie points out this might be 1.5 billion Ferrit dollars, normal dollars may buy considerably more.)
But for the sake of argument, say $1.5 billion fixes our domestic broadband infrastructure issues. Building another undersea cable might be around $700 million. So say 2.2 billion for NZ Broadband nirvana.
Because there is significant benefit from being able to better trade with the the world we could debt fund this and not have to spend the surplus (though we could). In some cases debt is good, especially when:
- we are building a capital asset that provides other benefits,
- the capital asset earns revenue that can repay itself,
- there is a benefit for future generations.
So lets say we debt funded 2.2 billion to sort out our infrastructure. At 10% the costs to service that debt would be 220 million per year. (We want to encourage savings. I’d certainly like to get 10% on my money on a Government backed bond.) Add another, say, 50 million for running the network at we have a 270 million annual operating cost.
Compared to what must be spent on the Internet each year in NZ that may not be a lot.
Let’s reality check that:
- There are 1m NZ households online. Say they all pay $30 per month. That is 360m.
- There are 300,000 SME’s in New Zealand. Say 100,000 spend $100 per month on the Internet. That is $120m.
- In the 2006 financial year, total government operating expenditure on information and communication technology (ICT) was $1.1 billion, and total capital expenditure was $0.6 billion.
- Telcom’s last annual revenue was $4.5 billion. Data was $438 million. Broadband and Internet $336 million.
- Exports from the New Zealand ICT industry totalled $1.6 billion in 2006, representing 9.1% of locally generated ICT sales of $17.6 billion.
- Apparently the Southern Cross Cable paid for itself in its first year. Southern Cross announced new sales of US$310m for the last 6 months of 2006 and plans for a major network upgrade to cope with a continuing upsurge in the demand for capacity.
New Zealand is therefore already spending considerably more that $270 million annually to connect. So having an SOE or other entity that provides infrastructure on a cost plus basis is quite reasonable. It appears there is a solid business case.
Why don’t we just fix the problem?
blog
You’re right Rod, at face value it makes sense. It would however need a major mindshift to go down this path. While government is seen as the appropriate funder of infrastructure such as roading, water and sewerage, it is reluctants to undo the changes of the past and invest in something like broadband infrastructure. It also goes against the free market grain. That said you are right - at a back of napkin analysis it would seem to be a beneficial investment for the country - both in terms of downstream economic benefits but also up front substitution benefits. It’s a pity there is no political will to go down that path. regards Ben benkepes.wordpress.com
Ben is right IMO. Politicians here are not used to practical arrangements to resolve a problem, especially if it goes against their inbred beliefs that the market needs to sort it all out. The market is well and good when there is potential for competition, but NZ is a very small and isolated market indeed, which means to do what you are suggesting involves huge amounts of political will, the setting-aside of inflated egos and of course investment in the future beyond next week. Unlikely, given the NZ psyche.
I suspect Telecom will be end up carving itself up.
It’s not so much that the government has to do things because private enterprise won’t do them. It’s that government has to stop PREVENTING private enterprise from doing them.
TelstraClear, for example, would have by now had a good portion of the country wired up with cable TV as the “last mile” (and fiber to that) if hadn’t been made prohibitively involved and expensive for them to do so BY THE GOVERNMENT (local government, mostly).
Government can often do things because when it does them it sets aside the rules that it would make anyone else follow. Most “natural monopolies” are not natural — they are government imposed.
Rod, your numbers are far too simplistic. You’ve forgotten depreciation, operating costs are too low, and what about the revenue build?
Now it’s a constantly rolling upgrade, but let’s say 5 years tops for the life of any piece of the network (given current replacement rates and demands for higher capacity). It takes time to lay lots of cable/fibre. If it takes 2 years to build half the capacity and only when it’s operational that you start earning real revenues, and the balance is built over another 3 years, you’re looking at an upfront loading of huge proportions, and the replacement story starts straight away.
You’ve grossly underestimated what it costs to provide the customer service, network operation, maintenance, accounting,legal and governance for such an operation. As to revenue, apart from us techbiz types, where’s the raging jump in revenues that will pay for it? And when will they arrive?
I’m not saying (yet) that we shouldn’t do it, but let’s get real about what it will cost. don’t have the specialist knowledge to work that out. I have run an electricty network company and advised several telcos, so I have a rough idea - it’s much much more than you’re talking about.
Someone has to pick up the tab. To make your case, get some solid advice from someone specialising in infrastructure ecomomics, accounting and operations.
[...] Drury has posted an interesting estimate of what it would cost to upgrade NZ broadband, as part of his case for a telecom infrastructure SOE. I just wish we could get some solid facts so [...]
http://money.aol.com/news/articles/_a/malaysia-leads-plan-for-first-undersea/n20070427040209990011
“KUALA LUMPUR, Malaysia (AP) - Seventeen major telecommunications companies signed a pact Friday to build a US$500 million (euro360 million) undersea fiber optic cable between Southeast Asia and the United States they claim will be relatively safe from earthquakes and tsunamis. ”
…
“Parties involved in the project include AT&T Inc. from the United States, the British Telecom Global Network Services, Eastern Communications Philippines Inc., India’s Bharti AirTel, Thailand’s CAT Telekom, Indonesia’s Indosat and PT Telkom,
** Telecom New Zealand International, **
Singapore’s StarHub and Australia’s Telstra.”
[...] whole infrastructure provisioning issue as it related to the structural separation of Telecom. Both Rod Drury and Jim Donovan have posted some interesting comments. The fact of the matter is that we will never [...]
IMHO the government won’t take this option, i’ve written about it before on my blog. This regulation isn’t about doing whats best for the country, its a political (that is vote collecting) exercise. The labour government are doing their best to obscure the fact that its failing the country (http://www.stuff.co.nz/4042265a1865.html) and looking for scape goats.
I wonder if 1.5 billion is around the same amount that telecom ships offshore every year, rather than investing back into maintaining the infrastructure….
Either way, the idea is good, getting NZ’s ludite poli’s to DO anything about it is another thing all together. If you tie it somehow to sheep and milk, THEN it may be a go-er. You’d think that National and Act would jump on it, tho - business focus and all that.
(sigh)
Off to the US (3 months) then the UK (a lot longer) - I wonder if anything will have changed when I get back :)
Andy Lark gave an interesting presentation today at the EDANZ conference in which he said that $1.5bill is not a great some of money and in fact less than the marketing budget of some businesses he has previously been involved in. I think that’s a little bit of hyperbole. While unquestionably NZ should invest the $1.5bill in this, a core infrastructure need, the fact is that, given our size and scale, $1.5bill IS a significant amount of money. Best not to forget that
benkepes.wordpress.com
[...] Marmite or Vegemite is the most popular spread. Channel Drury is providing 24/7 coverage of the Broadband debate and TVDPF is tracking a drag queen stalker. I wonder what’s on tv … Trackback [...]
I can’t really comment on your tasty plans to overhaul the net; both you and Donovan make valid points. At the very least, someone should look seriously at upgrading the system, and for god’s sake, make sure the research is done by the right people. For instance, some small business who’s sweating to make a buck and will have their children go hungry if they don’t do the report properly. You just have to consider how more money than what I’d know what to do with is spent deliberating over Transmission Gully to see how nasty it can get if you don’t have people using time smartly.
Anyway, I noticed a very contradictory report by the OECD noting far lower numbers than the link you posted about 1 million NZers having net. NZ Herald says “33% of house have broadband”, and mention we should model after USA’s 42%. The OECD report claims “14% of population have access”. So are to ponder that some families are telling the OECD researchers that they the children are grounded and don’t have net access today?
Seriously, though, I don’t really care if its 14% or 33%… Scandinavian countries are a fairly good economic peer to us and have double the uptake. As someone who lives and breathes the internet, its dissapointing to be reminded that 70-85% of people cannot access the website you’ve built quickly, or at all.
I appreciate that faster broadband is a real issue, and should be addressed, but when the majority either has dial up, or no internet access, we should really be to have a parallel crusade. I’m heartened that the OECD report showed NZ adoption growth rate among the highest, but we do need to be pumelling ISPs to provide “okay” broadband **for dial up costs**.
For the record, “okay” broadband at this time, in my opinion would be 1Mbit for 1GB traffic, downgraded to 256kbit afterwards, for $10-15 a month. Important facets of these numbers are mums dont have to complain their teenager got a $15,000 bill (no excess use penalties) and that ISPs have more expensive products for upselling. You can bet that in 12-24 months, all the numbers I mentioned should double or more, but the thing to focus on, I reiterate, is something to migrate dialup users towards, simply, in safety and in droves.
Lets look at some current pricing:
TelsraClear: $39 for phone line, or $42 to include DialUp, or $62.90 to include Broadband (2Mbit/1GB quota).
Its that extra $20 that’s killing uptake…
FYI. Telecom still charges $29 for a face-slapping 256kbit/200MB connection, and they won’t be changing that for a while until the clean up their Go Large mess.
( http://jetstream.xtra.co.nz/chm/0,8763,204548-203090,00.html )
I meant to say TelstraClear is $32 for phoneline, not $39.
You therefore pay $10 for dialup and $30 for Cable.
Let’s replace the dialup service with the “okay” 1Mbit broadband.
(Which could be considered screaming fast in the eyes of the Xtra sales team)
Rod
The undersea cables aren’t generally a problem. It’s glass in the water - and Southern Cross needed to be solicited around a bunch of telcos to get initial buy-in, nearly went bust twice in the process, and only made sense once they were able to jack the price up.
Where do the cables connect to? That would be MCI/Worldcom/Verizon/Whatever they are these days.
And they charge a horrible fee to onforward your traffic, as you know. Glass in the water is irrelevant - it’s the company connecting you that makes the difference.
Telecom has a POP in LA. They connect to Sprint/MCI. As does TCL. The fees are about as low as you can get, but they are still high given most NZ traffic goes there.
Thoughts?
Jim: if you’re talking about the cable, then multi-strand fibre-optic networks lasts longer than five years. Much longer.
Thanks to local council bureaucracy with silly rules requiring excessive depth for trenching (90-120cm) for instance, you’re looking at between $100-$120 per metre in populated parts of the country.
Juha: Correct - but the actual cable isn’t the big ongoing replacement cost once laid (although the initial rollout is a big upfront cost) - it’s the active equipment (including exchanges and connectors/splitters/reboosters/etc downstream from there). The need for depth is real - power cables at deeper depths get broken all the time, and we’re talking about vital infrastructure (or why are we having this debate). But again,it would be nice to have some real numbers from people who do know this stuff - and that’s not me!
Juha
…and gigabit wireless (non licensed 60GHz and 80GHz spectrum) is now around $35K capital cost for 5km+ range…
I’m in Dunedin - the central city here could be wired for 1Gbps to the 11 or so large buildings for around $500K capital. A 100M national uplink from that would be around $500K annually. 100M of international at today’s prices would be $2-3M annually.
That’s not huge capital outlay for a strong capability.
But those pipes are going to sit idle for 30% of the time. Where is the corresponding shift in business models that either moves us to 24×7 working or aligns NZ business habits with overseas partners?
Wireless is a non starter because of the latency; everyone gets seduced by the speed and forgets that the real world “where the rubber meets the road” speed is more complex than just the advertised speed. Plus you don’t want to go down the unlicensed spectrum route. Those of us who have lived through this will remember the problems when some other provider slams on top of the spectrum you are using. This is why we have licensed spectrums!
I think fibre is the solution; I don’t think the current ADSL1 or the upcoming ADSL2 is the solution if you are talking about all the issues Rod wants to address.
The Southern Cross cable is owned by TCL 50%, Verizon 10% and Optus 40% (or at least it was in 1998). Cost was $US1.3 billion including ground stations. Upgraded in speed to 240 gigabits in 2003 and can go to 1 terabits. So I am not convinced this is where the problem is. We have fibre at work and have virtually no speed problems, but then we are getting to the motorway faster than most of you; thats where the problem is I think. And because ADSL is so integrated with Telecoms backbone that’s where the issue lies I think.
What we can be sure of is that nothing will change quickly.
Dermott,
You have just demonstrated you do not know WTF you are talking about (and I suspect your sample size amounts to 1, or possibly 2).
Licensed and Unlicensed Wireless spectrum works equally well. The key thing is the engineering and knowing where and when to use it. I’m not surprised at all that early implementations failed. Simply a sign of the maturity level of those deploying it at that point in proceedings.
Granted Wireless is not a panacea, but then again neither is XDSLx or FTTx. Like all things, there’s a balance needed, and writing something off as a non-starter is non productive.
There is no silver bullet here. Only lots of bronze ones.
Transpacific is the problem. It’s the remaining bottleneck. At least the average NZ company can do something about solving the (some of the) problem on land in their own small way. Heard of “small pieces loosely joined”?
You need to have at least the weight of the Telecom Gorilla or the NZG to put together a deal like SXC, or even provide credible threat of doing so. Capitulation is a good result.
Precisely. The latency on a 1.25Gbps point to point wireless unit is sub-1ms. The 60GHz and 80GHz gear is naturally range-limited and you don’t get a lot of interference between beams if you engineer correctly. Oh and the 35K price tage makes sure that you don’t get a lot of casual users just lighting one up to see what happens. Plus you have 4 9’s reliability on the link and if you want 5 9’s, that’s two units.
The price per metre is better than fibre in many urban situations. The debate about fibre as such is a red herring. The debate needs to be about putting capacity to where it is needed and how to aggregate and distribute costings for those pipes.
Put it this way - a 2Mbps link that is active now is already outperforming a 100Mbps fibre link that can’t be deployed until 2008, after you’ve got your resource consents and capital and put stuff in the ground.
Jim: fibre deployment has come a long way, and it’s not as complex and pricey as it used to be. What’s holding back a FTTx rollout is more the fact that over short lengths, you get more than adequate performance over copper still, so it makes sense to use that from fibre-fed nodes.
Forcing fibre deep into the ground may indeed prevent Mr or Mrs Crazed Digger from breaking it, but it also bumps up the cost of deployment something horrible, plus makes it far harder to repair should it come to that. Better to lay it shallow and lots of it for redundancy, I say. Cheaper and more reliable that way.