Cartoon by Tandberg (source: YCAT)

I’m bemused but not surprised that the Prime Minister is still insisting stage one of Mebourne’s East West Link should proceed.

This is a project where the benefits – in terms of important attributes like avoided traffic casualties and travel time savings – are less than half the financial and economic cost of building it.

Yesterday, as the deal to exit the project was announced by Premier Daniel Andrews, Mr Abbott said the Victorian government is ignoring traffic congestion in Melbourne.

The East West Link is the only answer. Tens of millions of dollars and years of planning have already been invested to get the project shovel-ready.

The East West Link is the only answer to traffic congestion? You don’t have to reflexively oppose all new roads to understand this is a bad project. (1)

The “decongestion” benefits were estimated by skilled professionals retained by the former Victorian government, yet even when combined with other gains the total benefits still fall well short of the cost ( just 0.45 by Infrastructure Australia’s preferred methodology).

If you’re serious about tackling traffic congestion Mr Abbott, please understand that it can’t be achieved by building more infrastructure, whether roads or rail lines.

Gilles Duranton and Mathew A Turner concluded in an important paper published in 2011 in The American Economic Review, The fundamental law of traffic congestion: evidence from US cities, that “both road capacity expansions and extensions to public transit are not appropriate policies with which to combat traffic congestion”.

The key problem in our cities isn’t a lack of motorways; it’s the inefficient way the ones we’ve already got are managed. There’s an enormous amount of “latent capacity” locked up in existing congested major road networks and going to waste – it needs to be released by rationing access to road space.

Congestion could be approached by (say) restricting access to roads on alternate days based on number plates; or more sensibly it could be addressed by charging for the use of roads i.e. by congestion charging.

As I’ve discussed before (Are politicians serious about fixing traffic congestion?), it only takes a reduction of around 5% in the number of vehicles to increase average vehicle speeds by 10-30%. It won’t be at the speed limit necessarily, but it will be fast enough to satisfy the expectations of most drivers.

Congestion charging discourages drivers who make relatively low value trips at peak times. Those who aren’t prepared to pay will make the trip at another time, shift to public transport, or decide it’s not worth making.

According to research by transport analyst Craig Mcgeoch, 17% of morning peak hour motorised trips in Melbourne are for recreation and shopping purposes; only a third are for the purpose of getting to work.

The NSW Bureau of Transport Statistics 2012/13 household travel survey shows 18% of AM peak trips in the Greater Sydney Metropolitan Area are for social and recreational purposes, with another 3% for personal business (26% are commuting with an additional 8% for work-related business purposes).

Congestion pricing isn’t a radical concept: after all, we already charge for public transport and provide concessional fares for travellers with limited means. We already ration other finite resources like water, electricity and gas by charging according to the level of use. Indeed, we usually apply a progressive tariff i.e. the price increases as consumption increases.

By reducing demand at the margin, congestion pricing delays or puts off entirely the need to invest in costly new infrastructure like the East-West Link – see Infrastructure: what to do about the ‘Cleopatra problem?’

Once the set-up costs have been met, congestion pricing also generates revenue that can be used to improve public transport. Indeed, by raising the cost of driving it makes public transport a more attractive alternative.

Requiring motorists to pay for the congestion they cause can be done in a number of ways with varying degrees of efficiency and levels of cost. With suitably sophisticated technology, it would be possible to track the location of every car. Or it could be implemented by a cordon system as in Stockholm, London and Singapore, or via a proxy such as charging for parking.

The vertical equity effects are frequently portrayed as a “deal breaker” in debates about congestion pricing, as if governments don’t already charge for public transport and other essential services.

The equity of a particular measure shouldn’t be assessed in isolation but in the context of the fairness of the entire tax-transfer system. As I’ve noted before (see Is congestion pricing just too unfair?), regard should also be had for the offsetting benefits provided and, of course, for the ability to compensate those who are most vulnerable.

Governments will need to build more urban transport infrastructure but they also need to ensure limited funds are directed at the projects with the biggest pay-off. The scale of the task can be made much more manageable by getting the level of demand right in the first place.

Contrary to the Prime Minister’s assertion, motorways aren’t the “only” answer to traffic congestion; getting governments to manage excessive demand for road space is the key answer.


  1. Mr Abbott provides a good example of the Sunk Cost Fallacy i.e. He doesn’t want to see the tens of millions of dollars and years of planning that have already been invested in the East West Link go to waste.