In June last year, Maurice Williamson, in a presentation to the Telecommunications summit, undertook that, if elected, a National government would invest to ensure that 75% of New Zealand homes would enjoy connection to "ultra-fast" fibre-optic cable by 2014.
The speech was somewhat controversial with the industry delegates, principally for its partisan tone, but also for its grounding in reality: the target espoused by Williamson was not, in fact, realistic. Nonetheless, the promise to "ensure fibre to the home of 75% of New Zealanders by 2014" was touted throughout election year.
Key himself enthused to students in at Waitara College that they would be able to "download a movie in seven seconds". That implied gigabit Ethernet via fibre -- a fairly big call when you look at the cable run between Waitara and the nearest regional centre, New Plymouth.
This year, Williamson is not the Minister of Information Technology, and the discussion paper released yesterday by the man who is, Stephen Joyce, sets far more modest goals.
So much so that it is quite possible that the proportion of homes enjoying what John Key described yesterday as "quantum faster" internet (should someone tell him that a quantum is a very small thing?) in the six years from the earliest start data, 2010, will not be 75%, but zero.
Oddly enough, this is more like the goal outlined when the investment policy was first announced, almost exactly a year ago. In that speech, Key promised thus:
National's medium and long-term vision is for a fibre connection to almost every home, supported by satellite and mobile solutions where it makes sense to do so.
Our initial goal is to ensure the accelerated roll-out of fibre right to the home of 75% of New Zealanders.
In the first six years, priority will be given to business premises, schools, health facilities, and the first tranche of homes.
In the paper released yesterday the government's objective is cited as connecting "priority broadband users such as businesses, schools and health services, plus green field developments and certain tranches of residential areas." The window for 75% residential connectivity has been extended to 2020, which, if not exactly realistic, is markedly less of a fantasy than 2014.
There are frank and frequent references to market failure in the paper. "There is also a role for public investment," it says, "where existing private sector investment programmes are not aligned with the government’s desired timeframe for infrastructure deployment." -- because the kind of investment implied by the government's timeframe is not commercially viable.
There are some other subtle, but important, changes. The monolithic structure of a proposed "FibreCo" is gone. Instead, a Crown Fibre Investment Company (CFIC) will invite investment proposals from Local Fibre Companies, with which it will work on the installation of open-access fibreoptic networks in individual regions. The appropriate rate of return will be a matter of negotiation.
Also, last year, National talked about fibre to the home -- now, it's more like fibre past the home. The policy concentrates on "dark fibre" -- this is, cable installed but not fitted with termination equipment, which actually constitutes about half the cost of fibre to the home. LFCs will, under certain circumstances, be able to fit termination equipment, but that will largely be the role of telecommunications retailers.
Which is where is gets interesting. Under the draft regulations announced yesterday, a company that offers retail services will not be allowed majority control of an LFC. The likely upshot will be considerably more onerous than the operational separation required of Telecom by the Labour government. Telecom will be forced to sell its physical networks business, Chorus, if it really wants to play.
So the paper's contention that the policy represents a light hand in the market for government doesn't really wash at all. It's a very significant intervention.
Likewise, the claim that the project will avoid competing with, or duplicating, existing networks doesn't wash. Telecom is in the process of extending the useful life of the copper network maintained by Chorus by spending $1.5 billion installing roadside cabinets to bring DSL equipment much closer to every home -- thus shortening the copper runs and greatly increasing the potential speed of the network.
What signal is being sent to Telecom -- and for that matter, to such companies as Vodafone and Orcon, who are already installing their own equipment in unbundled Telecom exchanges? Should they persist with that investment in existing infrastructure if the government is saying that the copper will be built over in fibre? Do we in fact face the prospect of a freeze in investment by existing telcos?
It would make sense for the fibre running to the new cabinets to be extended out from the cabinets, rather than duplicated -- but that fibre is the property of Chorus. Even if Telecom sells off Chorus, the fibre to the cabinets will not be open-access. We can only guess at what kind of negotiation would be necessary to make that work.
There are some notable omissions amongst the ideas canvassed in the paper. The word "peering" does not appear once, even though connectivity between the local fibre networks will be essential. Will the project-- which almost ejects traditional telcos at the local level -- be subject to the same old business practices at the next tier up of the network? And strikingly, there is no mention of international connectivity. Will Kordia still be allowed to build its proposed trans-Tasman cable? We don't know.
It's also a given that the $1.5 billion in joint investment over 10 years will not amount to half the total cost of the project, or anything like it. This is a $6 billion job, assuming the skilled labour (which is presently either employed or contracted to Chorus) can be readily found.
And I wish the government would stop citing HD TV as a killer app for the new network. We already have two efficient means of delivering HD TV -- Sky's satellite network and Freeview's terrestrial one. Further, the kind of network configuration that provides the very high speeds implied by HDTV (especially on-demand) is quite different to -- and much more expensive than -- one that would simply run cable past your house. It's the difference between sharing a pipe and having your own pipe all the way back to the exchange. John Key's government already seems to have a knack of holding itself hostage to its own soundbites.
But, lest I seem too negative, the idea is, in spirit, visionary. Ubiquitous fibre is as close to future-proof as you'll get. It's just that as it charges towards that future, the government may find it isn't wise to walk all over an existing infrastructure that is not only working now, but working better than it ever has.
PS: We still have some room for this evening's recording of Media7, featuring an interviewing with the departing editor of the Dominion Post, Tim Pankhurst, and a panel discussion on the state of the Official Information Act with Nicky Hager, Haydon Dewes and PSA national secretary Brenda Pilott. If you want to come along to The Classic before 5.30pm, hit Reply and let me know asap.
PPS: Internal Affairs minister Richard Worth has been assiduously removing unflattering news from his own Wikipedia article. There's plenty more here. Oh dear.