Thursday, November 23, 2017

Fake economics: how to make bad transport projects look good

Victoria is spending $5.5 billion building the West Gate Tunnel, another $1 billion widening CityLink, probably $10 billion on the North East Link, $11 billion on Melbourne Metro, $8 billion removing level crossings, and, if the Coalition returns, more than $3 billion on the East West Link.

NSW is spending $16 billion on WestConnex, $14 billion on Western Harbour Tunnel Beaches Link, $9 billion on the F6 Extension, $3 billion on NorthConnex, $11 billion on Sydney Metro South West, $8 billion on Sydney Metro NorthWest, $3 billion on Parramatta Light Rail, $2 billion on Sydney Light Rail, and billions more on Sydney Metro West.

It would be nice to know it was money well spent.

There's a fiction that a benefit-cost ratio above "1" means things are OK.

Here's how it works. A consultant adds up all the costs over a period of 30 or 40 years and all the benefits. If the benefits are greater than the costs, giving a ratio of, say, 1.5, it is said to be worth doing. But if they are less, say, 0.45 (which was the ratio in the first study of in the East West Link), it is said to be a waste of money.

Often the studies are never made public, sometimes they are never conducted (as was the case with the national broadband network) and very often they are conducted as an "add-on"; financial bling to be sprinkled over the project after it has been approved and announced.

Melbourne's $5.5 billion West Gate Tunnel is a case in point. Sydney's $14 billion Western Harbour Tunnel and Beaches Link is another. Internal NSW Transport emails released to me under freedom of information show an analyst complaining that his superiors had as good as completed the business case without access to the numbers.

"How something with no, repeat, no, benefit-cost analysis or traffic numbers can be construed as 80 to 90 per cent complete is beyond me," the exasperated official wrote. "The numbers tell us if the thing makes sense."

And the numbers are sometimes rigged.

A seminar in Melbourne last month on the use and abuse of cost-benefit analysis explored the ways.

One of the easiest is to hike the traffic forecasts. On some toll roads, the number of cars predicted to use them was greater than the capacity of the roads. Out of court settlements were reached between the modellers and investors in Sydney's Lane Cove Tunnel and Brisbane's M7 Clem Jones Tunnel.

Professor Jago Dodson of the RMIT Centre for Urban Research revealed that in the queue at a conference he had met one of Australia's senior transport modellers who had worked on at least one of those tunnels.

"Myself and another colleague were joking. 'You guys all inflate your traffic figures to satisfy your clients, don't you?', we said. He replied: 'Oh no, no, no, we are professionals, we have to sleep at night.'

"Then he sort of slyly looked at us and added, 'But it's amazing how little sleep you can get away with'."

You needn't stop at bulking up travel time saved. Also useful for bulking up benefits is "travel time reliability". It's a concept that makes sense for some types of public transport. You want trains and buses to leave and arrive on time. But it makes next to no sense to count it as a benefit for commuters in cars, who can usually leave and arrive whenever they want. Suspiciously, it forms an important part of the claimed benefits for Sydney's Western Harbour Tunnel and Beaches Link.

And you can get more creative. You can add in the benefits of other projects tens of kilometres away as was done for the West Gate Tunnel. You can add "wider economic benefits" to fill the gap that remains.

The original study of Melbourne's East West Link came up with a benefit cost ratio of 0.45. Then "wider economic benefits" were added to take it to 0.85, then the benefits of other "complementary projects" were added to force it above 1. The wider benefits included the "impact of transport on increasing competition", "competition related user benefits", and the biggest: "agglomeration benefits".

Agglomeration helps productivity because it packs more workers in the one location (although the cost of the accompanying deagglomeration – depopulating smaller locations – is rarely counted). But it made no sense to count them as a benefit of the East West Link. Its whole point was to bypass the city. It makes little sense to count them as a benefit of the Sydney Northern Beaches Link. Most of its users would have gone into town anyway.

But the biggest fudge is the simplest. It's what you choose to compare. By not comparing the costs and benefits of the (much) cheaper rail alternatives to those of WestConnex or the Sydney F6 Extension, the government made their figures look good – but good compared to what? Economics is about choices. Studies that don't examine choices are neither economic nor meaningful.

Each of Australia's two biggest states is engaging in an unprecedented transport spending spree, often with the help of willing partners in the finance industry hungry for access to tolls. Neither can demonstrate convincingly that it is getting value for money.

In The Age and Sydney Morning Herald