While this analyst values his life and peace of mind too much to meddle in Aussie politics, the fact that the opposition coalition is likely - according to current polls - to win the next national election means it's worthwhile taking a look at what it's proposing. (A full party political PDF can be downloaded here)

Putting aside the inevitably partisan alternative costing of Labor’s NBN presented by the coalition, the key changes are the following:

Most of the current NBN plan’s Fiber to the Premises (FTTP) footprint will be switched to Fiber to the Node (FTTN). This will affect 71 per cent of the premises currently targeted with FTTP. FTTP will remain the architecture of choice for areas where copper needs renewal and for greenfield homes and buildings.

The switchover of the cable plant that was negotiated with Telstra and Optus as part of the NBN Plan will be rescinded provided the HFC providers agree to open access to their network on an identical basis to NBN Co’s wholesale prices.

NBN Co will remain as a structurally separate network entity, and the switchover from Telstra to NBN Co will occur as nodes go live to ultimately leave Telstra as a pure service provider entity (with the exception of their HFC footprint presumably).

Effectively, the Coalition is betting the house on vectoring not only delivering its current promise but continuing to improve performance in the coming years. It’s also sacrificing upload as a desirable component of broadband (indeed, the words upload or upstream are nowhere to be found in the document). The coalition seems to view the internet essentially as a broadcast network.

There is no surprise there, and as is pointed out in the document, this is an arbitration made for cost reasons. Also unsurprisingly, the Coalition plan is significantly cheaper than the current NBN plan, about 45% cheaper at $20.4bn vs. $37.4bn.

They key issue with the plan as I understand it, focuses on demand issues. The one element of the current NBN that makes a lot of sense is the full acceptance of an infrastructure monopoly. It may not be the only way to deploy a national broadband fiber network, but it’s one way that makes sense and eliminates demand uncertainty from the infrastructure funding. The coalition reintroduces demand uncertainty in two ways:

It keeps alive existing cable networks and opens up the possibility for Telstra (or others) to deploy their own FTTP network. In areas where an alternative infrastructure exists, the commercial success of NBN Co will inevitably be affected, especially as well managed HFC and FTTP will offer superior performance.

By lowering the performance of the access network, it opens up the possibility for Telstra and other mobile operators to position LTE as an alternative to wireline.

In addition the Coalition plans to significantly lower the NBN Co’s wholesale prices. One wonders about the compounded impact of all this on the revenue side of the equation, a topic that is not addressed in the Coalition’s plan.

Another question mark regards the feasibility of the FTTN deployment at the costs mentioned in the Coalition document. Questions have been raised in Australia on the quality of the copper, but that’s a hard element to assess. More importantly, the assumption that FTTN is systematically cheaper than FTTP can definitely be questioned.

It’s most likely that the Coalition hasn’t done an in-depth geomapped assessment of FTTN deployment costs, and experience in other countries suggests that the low costs of FTTN are largely dependent on targeted deployments. In other words, the numbers quoted by the Coalition are from deployments in select areas where deployment is feasible and deployment costs are low. Nationwide FTTN on the scale the Coalition is contemplating has not been undertaken anywhere (even in Britain). That doesn’t necessarily mean the Coalition is wildly off, but it’s a further level of uncertainty.

Finally, some of the more minor propositions in the document are clearly political in nature and infeasible in practice. The notion for example that people would be able to ask for a dedicated FTTP line and pay for half of it themselves is absurd. Not only would the costs be prohibitive (both to consumers/businesses and to NBN Co) due to the absence of economies of scale but the network management issues would be insurmountable (especially since the FTTP topology of choice for NBN Co is point to multipoint, which only works for mass deployment). In other words, the Coalition here is trying to suggest that if you really want FTTP you’ll get it. It won’t happen.

From day one, the biggest risk associated with Labor’s NBN Co plan was the political risk, and it’s looking increasingly likely that that assessment was correct. Changing courses midstream, especially for large scale infrastructure projects, is never a good idea. Politically however, the Coalition has made the NBN such a big issue that taking a milder stance on revising the NBN Co approach is no longer an option. By choosing a more radical redesign, are they endangering the whole thing?

The answer is likely no. The Coalition’s plan is workable on paper. It would deliver better broadband than what is currently available, if not as good as what the current NBN plan promises. The biggest risks associated with this new plan are its possible economic fragility and the absence of contingency planning if FTTN and vectoring are more expensive or don’t perform as well as hoped.

The core of Labor’s NBN plan though, which is the structural separation of networks and the equal access to a regulated wholesale platform, is retained. By staying on board with that the Coalition offers a plan that is within the continuity of what has been done. It’s less ambitious and a little more risky on the financial and technology side, but it’s also significantly less costly in the short term. That’s an arbitration worth considering, at least.Benoît Felten is the founder and Chief Reseach officer of Diffraction Analysis, a research firm specialised in next-generation broadband. He recently released a report entitled Building the Optimal NGA Service Portfolio.