Just six weeks since Malcolm Turnbull became opposition communications spokesperson, he’s already sharpened up the Coalition’s broadband policy and provided politically-saleable points of difference from Labor’s national broadband network.

The new policy was reported by Fairfax yesterday. Economics correspondent Peter Martin has conveniently summarised and compared-and-contrasted the two on his blog, although my emphasis would be different.

Coalition policy: Telstra would be split into retail and wholesale companies. The wholesale part — and we seem to be calling it CANCo for ‘customer access network’ — would provide all Australians with broadband of at least 12 megabits per second (Mbps). Where 12Mbps isn’t commercially viable, CANCo gets a government subsidy.

The CAN — that is, the ‘last mile’ or indeed several miles, would use whatever technologies suit, including the existing copper, Foxtel hybrid fibre-coax (HFC) cables, wireless or satellite. Or, indeed, fibre. That’d be CANCo’s decision, and of course they’d start off with the existing Telstra CAN. That doubtless includes a lot of hidden fibre laid over the years.

Labor policy: A new CAN wholesaler is created, NBN Co, delivering at least 100Mbps to 93% of the population using new fibre laid down Telstra’s existing ducts and 12Mbps to the rest using fixed wireless or satellite. Telstra becomes a retail-only operation, using NBN Co as its wholesale provider just like everyone else; NBN Co pays Telstra a one-off fee for every wholesale customer it loses to NBN Co, and eventually the copper CAN is shut down. NBN Co pays Telstra rent for its exchanges, ducts, etc. Telstra agrees not to create a competing higher-speed CAN using Foxtel HFC.

All retail telcos pay NBN Co the same wholesale fee for their customers’ services wherever they are. Since NBN Co can roll out fibre cheaper in more densely populated cities than rural areas, city customers are cross-subsidising the bush. All that happens within NBN Co, which contrasts with the Coalition’s model of direct government subsidy of the commercial CANCo.

The Coalition policy isn’t costed, but I agree with Martin that it probably wouldn’t cost tens of billions of dollars. I disagree with Martin when he continues to use the $43 billion cost of the NBN, since NBN Co’s deal with Telstra reduces that figure and in any event it’s not all taxpayers’ money.

At their heart, both policies turn Telstra into a retail-only operation and create a national wholesale CAN that’s not Telstra-as-currently-constituted — although they take different paths to get there.

The key, unstated, difference is about future network speeds. The Coalition policy is about upgrading the existing CAN to a speed that I already have at my desk. It says nothing about what minimum speed Australians might want or need in the future, except that it’s up to the market to provide. Does that mean 12Mbps is seen as fine for the next decade?

Labor’s policy makes a clear statement about providing speeds significantly faster than today, speeds we might well need in the not-too-distant future — but of course at a significant cost. Turnbull’s National Broadband Network Financial Transparency Bill is all about keeping an eye on that spend. But it’s also about providing evidence of hard economic benefits from those higher speeds.

So far, communications minister Senator Stephen Conroy has been dismal at explaining those benefits. As I said on the Patch Monday podcast, the TV adverts were little more than vague generalities, and Conroy’s example of the smart dishwasher was ludicrous — because it doesn’t actually need broadband speeds at all, and it’s trivial to satirise.

I’ve got my uses for 100Mbps and more. So do other übergeeks. Conroy’s challenge — as it always has been — is to explain to ordinary voters, specifically, why those and even faster speeds are a good investment. Because to voters without a personal vision of what the world might look like a decade hence, 12Mbps sounds fine.