Some things you can't blame Malcolm Turnbull for: the state of the NBN is one of them.
Customers who switch to it (and they are being given no choice – their existing services turn off 18 months after they are invited to switch) are greeted with slower speeds, more dropouts, unreliable phone and alarm services, and movies that continually buffer when they try to watch them between 4pm and 10pm.
Many are worse off than they were before.
Importantly, they are worse off whether they are getting the new Turnbull NBN, which is delivered by a mix of technologies, or Labor's original national broadband network, which in cities was to be delivered all by fibre-to-the-premises.
Australia's foremost telecommunications analyst, Ian Martin of New Street Research, says the method of delivery has nothing to do with it.
"Let's be clear, technology is not the issue in slow speeds," he wrote in the Australian Financial Review this month. "Hybrid fibre coaxial and fibre to the node are well able to handle speeds of 50 megabits per second and 100Mbps or more. In some places the copper component is old and slow but this is not an issue across the board and can be dealt with other than by an expensive upgrade to fibre to the home nationally."
Horrendously expensive and unbelievably time-consuming, Labor's original plan was for a technician to lay or string an optic fibre cable into each home and business in each of Australia's cities and big towns at a cost of $4300 per premises. It was essential for Labor to keep the cost off-budget: $49 billion gets noticed when you are running a deficit. So it decreed that NBNCo, the government-owned company that was to build the NBN, would eventually make a profit.
Spending on profit-making companies isn't treated as spending under the oddly-named Charter of Budget Honesty, which is one of the reasons the current government is building the Western Sydney airport and the Melbourne to Brisbane rail freight link that way. As long as it can pretend they'll one day make a profit it can keep the cost of building them off its books.
For a project as expensive as the NBN, that was hard.
Dodgy assumptions in its 2010 corporate plan helped. It assumed that no more than 16.4 per cent of customers would abandon fixed lines. It's achieving only a 75 per cent take-up, suggesting the real figure is closer to 25 per cent. It assumed the number of households would grow at a compound annual rate of 1.6 per cent per year. Between the last two censuses it's grown by much less.
It assumed that people would want to pay big-time for more speed, as it had to.
Two million or so customers on, it's discovering that eight out of 10 select the lowest speeds possible: 12Mbps and 25Mbps, turning 50Mbps and 100Mbps into niche products.
It expected customers to upgrade as they became acquainted with the wonders of high speed, but given that many were acquainted with higher speeds before they switched, that's unlikely.
And it adopted a curious method of charging the retailers who sold its connections to the public. It billed them twice. First for speed, under a monthly "access virtual circuit" charge set to reflect the speed provided. It's a silly idea. There's nothing to stop it giving everyone the highest possible speed; it's like constraining a car engine until the owner hands over more. And it charges for "connectivity virtual circuits" which reflect how much data the retailer is able to put through at once. Again, there's no reason for this. There's usually no physical constraint on how much can be put through at once.
Retailers wanting to grab market share have been paying for the first – it's an access charge – and skimping on the second. This means they can claim to offer the quoted speeds at a good price, but can't deliver them when their cut-price connections become congested in the early evenings and speeds slow to a crawl.
Their customers blame the NBN, the NBN blames the retailers, and so worried is the government it has commissioned the Australian Communications and Media Authority to inquire into the whole mess.
There's nothing inherent in the NBN that's strangling its speed and giving customers grief; it's inherent in the pricing model it adopted to make it look as if it could make money.
Turnbull's NBN will at least be cheaper. Martin believes that under Labor's scheme the connectivity virtual circuit charge would have been three times higher, making the retailers skimp more and degrade speeds further.
One day it will be sold. That's the policy of both parties. Even under Turnbull's scheme the government will be lucky to get back half of what it spent. Then, when it becomes clear it was never a profit-making concern and the spending gets brought back on-budget tarnishing the financial records of both Swan and Morrison, we'll start wondering why we ever thought we could spend up big replacing what for many is a perfectly good service instead of spending less and zeroing in on the people who needed help.
In The Age and Sydney Morning Herald