Its draft report released in March identified what it said was a classic example of how not to do it.

“The National Broadband Network, Australia’s largest public infrastructure project, was commenced without a cost-benefit analysis having been done,” it said.

“It also appears that detailed analysis of the project was focused, from a relatively early stage, on how best to implement the government’s policy objectives, rather than considering the merits of different options.”

It sounds like Badgerys Creek.

The Commission said the proper process was for the government to first work out what it wanted to achieve (such as moving planes or passengers quickly) and then work out the cheapest means of achieving it. It might be congestion taxes or time-of-day pricing.

Only then, if nothing else would do the job more cheaply, should it build something new.

To spend billions without first having precisely identified the problem and investigated cheaper ways of solving it would be to waste public money.

I am in possession of some back-of-the-envelope calculations done by a Sydney economist who has tried to do that for the Sydney Airport.

It is well located for travellers (much better than Badgerys Creek) but has a curfew and is limited to around 80 aircraft movements per hour.

It only ever reaches 80 aircraft movements during the morning and afternoon peaks. If regional aircraft were encouraged to move to Bankstown and cargo flights to move to Richmond the peaks would vanish.

Regional aircraft need only 1.8 kilometres of airstrip to takeoff. They don’t need to 2.5 and 3.96 kilometres available at Sydney airport. Squandering it on them is “an inefficient use of a scarce aviation asset”.

Regional aircraft typically carry fewer than 80 passengers. A Boeing B747 takes 396, an Airbus 330-300 takes 290. Allowing regional aircraft to use valuable landing slots at peak times means the airport is carrying many fewer passengers than it can.

Regional aircraft typically account for 20 per cent of movements at Sydney Airport.

Most could be banished to Bankstown if its runway was upgraded from 1.4 to 1.8 kilometres and its capacity lifted from 20 to 30 tonnes. A noise levy at the Bankstown and Sydney airports could pay for insulation in the affected homes. The training schools that presently use Bankstown could be moved to Camden.

With commendable understatement the economist says the costs would be “negligible compared to the cost of building a second airport at Badgerys Creek”.

And that’s just the start. Peak period charging could move non-time sensitive customers to other times of the day.

“Applying the laws of supply and demand to address capacity constraints could prove a unique experience for many stakeholders,” the economist says. “Not allowing pricing to ration the shortage of landing and takeoff slots during the peak periods means not using an existing asset efficiently and can no longer be justified.”

And the airlines themselves could reconfigure cabins to carry more customers by allocating seats more efficiently between business and economy class.

Trying all of these things first, as the Productivity Commission suggests, would postpone the need for a second airport. And they might make us realise we don’t really want it.

It’s worth asking who would use it.

The economist writes that Badgerys Creek will become a “spillover” airport.

“Passengers are unlikely to be impressed by the commute time from Badgerys Creek to the centre of Sydney.”

We will continue to want to use Sydney and the airlines will as well, just as they prefer Tullamarine to Avalon in Victoria.

The economist reckons the government has Buckley's of getting its money back for decades.

“It is likely that Sydney Airport will commence pricing policies for the off-peak period to ensure that Sydney Airport does not suffer revenue leakage to Badgerys Creek,” he warns.

I don’t know whether that is right or not, but I would like to know the government had tried other solutions before settling on spending billions building a new airport in an inconvenient location.

As the Commission says, the first step is to identify the problem.

@1petermartin

Follow us on Twitter