Melbourne rail network weekday passenger demand relative to crush capacity projected 2031 (source: Infrastructure Australia, Australian Infrastructure Audit, Vol 2)

Infrastructure Australia’s (IA) report on its audit of the state of national infrastructure says the passenger task for private and public transport in Australia’s six capital cities is projected to increase by 58 per cent over the 20 years from 2011 to 2031. The Australian Infrastructure Audit found that serious congestion and overloading will occur in all capitals unless something is done about it.

It would be helpful to understand more fully the methodology used in the report but that isn’t possible; the supporting technical documents which IA says are available on its web site aren’t actually there. Nevertheless, while there’s lots of sensible advice in this report, it prompts a couple of other points I think are worth making. (1)

One is that the sort of disaster scenario shown in the exhibit only tells us how bad things would be if governments sat idly by and did nothing. It’s a bit like saying you’ll stink to high heaven if you don’t have a bath for 20 years. But for all their failings, governments actually do try and do things.

For example, the Andrews government in Victoria has promised to spend $2 billion on roads over the next 10 years and, it seems likely, will greenlight Transurban’s proposed western distributior in Melbourne.

It also says it will, inter alia, build the Melbourne Metro rail tunnel, extend the Sth Morang rail line to Mernda, increase the capacity of the Dandenong rail corridor, and eliminate 50 level crossings. Moreover it’s promised to buy 37 new high-capacity trains and 20 E Class trams.

Even if governments didn’t do anything, the world wouldn’t stand still; it’d adapt. For example, if the rail system ever approached the level of congestion shown in the exhibit, it’s likely some CBD businesses would move to suburban activity centres where they could take advantage of spare capacity in trains running in the counter-peak direction. It might or might not be the optimum solution, but it’s unlikely it would be as scary as the exhibit suggests.

The other point is that providing more infrastructure to meet relatively brief but ever increasing periods of peak predicted demand isn’t the only or the wisest solution. This is an issue I’ve discussed a number of times before (e.g. see Are politicians doing what’s needed to grow our cities?).

While there’ll always be a need to build some additional road and rail in growing cities, the key problem isn’t lack of capacity; rather, it’s excessive demand resulting from political failure.

This insight applies particularly to roads. The main reason they’re congested is access to limited road space isn’t rationed. One consequences is that in the order of 20% of trips in the morning peak are for discretionary purposes like shopping or personal business; these are the sorts of trips that have potential to be undertaken in the off-peak.

Charging motorists to drive in congested periods would have a number of benefits. Most importantly, it would reduce the pressure to provide more capacity. Further, by eliminating less than 10% of vehicles it would get traffic moving faster (although not at the speed limit); increase the relative attractiveness of public transport; and eventually might generate surplus revenue that could be used for other transport projects (see Are motorways the only answer to traffic congestion?).

Similar logic applies to public transport. For example, off-peak fares offering a significant discount could shift demand away from the peak (e.g. see Is rail performing at its peak in Australian cities?). Or requiring city centre property owners to contribute to the cost of expanding rail capacity might encourage some businesses to move to less-congested locations (see also Is building new lines the key measure of progress with rail?).

As I noted here, while governments will need to build more urban transport infrastructure (although hopefully only the projects with the biggest pay-off); the scale of the task will be made much more manageable by getting the level of demand right in the first place. (2)


  1. IA should fix this error forthwith. I’d also note that yet again, a government released an important report to the media well before it was made available to the humble citizens of Australia. The latter had to wait until after the Prime Minister launched the report the next day before they could download it; in the meantime they had to be content with the media’s interpretation
  2. Yes, there’d be downsides from congestion charging that must be considered too, in particular negative equity effects; but they shouldn’t be used to dismiss the idea out of hand; see Is congestion pricing just too unfair?