Myth: Public transport is expensive to run
Fact: There are enormous economies of scale in public transport. To carry each additional passenger costs virtually nothing, whereas to put each additional car on the road system adds to congestion and pollution. What’s important is to ensure public transport is of high enough quality to attract passengers throughout the day, so that operating costs aren’t wasted running empty vehicles. Running good public transport does require real money, but on the whole it costs much less to run public transport well than it does to run it badly.

The argument is sometimes made (mostly in places with even worse public transport than Melbourne) that public transport is a drain on taxpayers’ money, whereas roads are wholly paid for by their users. Since it isn’t practicable to raise fares enough to put public transport on a wholly user-pays basis, the only ‘equitable’ solution, it is said, is to cut back services until a break-even point is reached. An extreme version of this argument was put in the 1990s by then Federal Liberal leader John Hewson, who argued that it would be cheaper to scrap public transport entirely and give free cars to those who didn’t already have one!

In fact, public transport is a bargain for the taxpayer by comparison with a high-capacity road system. The cost of the latter is highly sensitive to the number of road users; each additional user adds more and more to the cost of the system. In economics this is known as the ‘law of diminishing returns’. By comparison, the cost of public transport is more or less the same regardless of patronage, so that if more people use the system, the cost per passenger actually diminishes. This is what economists call ‘economies of scale’. Since public transport revenue increases with more passengers, it is clearly economic to encourage more people onto public transport. On the other hand, it is distinctly uneconomic to encourage more people to abandon public transport and drive cars instead. But this is just what the government has done for the past four decades – and continues to do with its new motorway plans!

Unfortunately, even when this much is explained, the fallacy still arises. This is because there is no shortage of Australian and US cities and towns with low-quality bus services that spend most of their time running almost empty. The cost per passenger of running an empty bus is not just higher than that of running a car: it’s infinitely higher. So it’s not good enough to just run a few buses or trains for the sake of it: the key to cost-effective transport is high passenger volumes.

Ensuring an appropriate level of ‘bums on seats’ on public transport requires not so much money as brains, to ensure that the available infrastructure is used to its maximum potential to provide a high frequency, close-knit transport network. But when brains are in short supply, there is a temptation just to spend more money instead. This is responsible for much of the poor reputation of public transport in the USA and Australia.

By contrast, many systems in Britain, Europe and Canada not only provide a higher level of service – they actually cost less per passenger to run. Vancouver, for example, has for thirty years pursued a strategy of investment in public transport as an alternative to traffic congestion. Its public transport runs at much higher frequencies than ours, and charges lower fares. Yet its budget for roads and public transport in 2005 was $180 per resident, compared to $430 per resident in Melbourne.

It seems to be a quite general rule that it costs much less per passenger to run public transport well than it does to run it badly. The effects of operational inefficiencies, neglect and bad planning leak right through to operators’ bottom lines: on the other hand, a little effort invested in efficient scheduling and service quality to grow patronage (rather than just apologising for failure) pays off many times over in higher cost recovery.

It’s also common to baulk at the high upfront costs of new urban rail systems (in particular), not realising that once an initial investment is made, a high level of service can be provided for relatively little additional cost. By comparison, new road projects can be sold to the public fairly easily, because of the perception that roads cost nothing to maintain. Not a word is said about the vast sums of money that will be required in the future just to keep the road up to scratch.

We’ve found that cities which emphasise walking, cycling and public transport are healthier financially and spend less of their wealth on transport costs. The six cities that came out the best were cities like Zurich, Copenhagen, Stockholm – very wealthy cities now that spend only 4 or 5% of their wealth on transport, and yet they’re the cities that are putting their money into public transport. And the cities still pouring money into freeways use up to 17% of their wealth. Our data would really question that now and say unless you’re building up the rail system in a city like Perth you are not going to help it economically as well as pouring this money into big roads.

—Professor Peter Newman, radio interview, 1998

The road lobby is fond of calling attention to the large ‘hidden tax’ component in petrol prices, as though this proves that motorists are being bled dry by unfair taxes. It is true that of the petrol taxes contributed by motorists, only a small proportion is spent directly on roads, while the rest goes into consolidated revenue. But the argument is also misleading, because the vast bulk of spending on roads comes out of consolidated revenue, mostly via grants to the states. If one draws up a balance sheet comparing road costs on the one hand with petrol tax revenue on the other, the bottom line shows a shortfall, not a surplus. It is general taxpayers who subsidise motorists, not the other way around.

Sydney is far from world’s best practice in public transport. But even there, a 2009 study by the University of Technology (UTS) found that of all modes of transport, trains had the lowest overall economic cost at 47¢ per passenger kilometre, compared with 57¢ for buses and 86¢ for cars. The common perception that public transport is subsidised to a greater level than cars was found to be contradicted by the evidence.

Often Treasury just looks at it in terms of big dollops of dollars for infrastructure that you need for rail systems rather than looking at the long-term costs. I think we’ve built our city around the car to some extent and I think it’s going to take a while to move away from it.

—Study author Garry Glazebrook, Sydney Morning Herald, 17 March 2009

None of this is to deny that public transport in Melbourne is also heavily subsidised. Our public transport has poor cost recovery because, as in most of those cities with low-quality systems, patronage is largely confined to peak periods. This means high costs to provide staff and equipment, most of which are not used for the rest of the day. In addition, the relatively low proportion of full-fare ‘choice’ passengers to concessionary ‘captive market’ passengers mean that fare receipts stay low. Cost recovery will only improve if public transport operators make an effort to attract the ‘choice’ passengers away from their cars, thereby grabbing ‘market share’ from road transport.

Trying to reduce subsidies by raising fares and cutting services is a lose-lose strategy; cost recovery is bound to decline under such conditions, as the last 50 years in Melbourne has shown. Meanwhile, Toronto and Vancouver enjoy more than twice our level of cost recovery, despite starting out with vastly inferior infrastructure on which to base their service.

Few people, John Hewson notwithstanding, seriously argue against taxpayer-subsidised transport. A case can be made for governments supporting both public transport and road maintenance from a combination of general revenue and user charges. The level of support provided to each is a matter for us as a community to decide. Fallacious claims from the road lobby, and appeals to value-laden ideas of ‘user-pays’, only confuse the issue.

Last modified: 21 August 2009