Copper, fibre, nodes, how do we know best?

It will be many years before an experience-driven judgment can be passed on whether the Coalition's plan for a cheaper national broadband network proves to be a smarter policy than the Labor government's Rolls-Royce of communications infrastructure.

But history and experience tells most of us that you get what you pay for and false economies can prove very costly.

The good news is that the Coalition has provided some estimates on how much its scaled-down model will cost against what it provides and this can now be compared with the government's fibre-to-the-home scheme.

The bad news is that neither can be relied on. Massive infrastructure projects almost never go to budget - especially when budgets are set years earlier and rarely run to time.

Opposition communications spokesman Malcolm Turnbull sits on solid ground when he asserts the NBN is already behind its promised rollout schedule. He is on weak ground if he promises that under a Coalition government there will be no similar hurdles.

For example, one issue not expressly addressed in the Coalition broadband policy statement is the cost of maintaining that part of the Telstra copper network that it plans to retain - the bit from the node to the home. Analysts estimate the copper network sucks up $1 billion in maintenance.

The Coalition plan still will be cheaper but questions arise over how much cheaper.

The real issue is - do we need the Rolls-Royce or would we be better spending less and building a network that is adequate, workable and a lot better than the one we operate with today.

The Coalition suggests this could be answered by a cost-benefit analysis.

In most respects these are useful exercises.

But when it comes to building a network whose uses and usage are unknown beyond a couple of years - placing a measurement on the benefits is near impossible.

The Coalition has taken a fair stab at it based on today's needs extrapolated to account for increased internet usage in the short term.

But they (and we) don't know what else will emerge even over the next five years that will require wider bandwidths.

This is why Communications Minister Stephen Conroy is so fond of using the one-lane harbour bridge analogy. In the 1920s this looked like an ample-sized road to cross the harbour but accommodations were made for future traffic needs.

Even Turnbull acknowledges that some users will want higher speeds (more bandwidth) than can be provided by the fibre-to-the-node home plan it is proposing and that upgrades might be needed.

Under this scenario, it has a plan to co-fund business or private investors to build a better service in particular spots.

Its not a perfect solution and neither is it the Rolls-Royce.

But then that will be reflected in the cost. The Coalition network will cost $20.4 billion - based on today's estimates against Labor's $37.4 billion. But just as the Coalition can rightly claim the NBN's broadband build has been riddled with problems and thus can't be trusted, the voters would be making a leap of faith to think this massive infrastructure project will be smooth sailing under a different government.

From a political perspective one has to question whether the public really cares. There is no ideology contained in choosing which network is better - it's just a matter of whether they accept the positives of being more financially prudent outweigh the detriment of an inferior network.

In a general sense the public just wants a network that works.