How do you get people to stop driving and start taking public transport?

Not much to cheer about here — mode share of motorised travel (passenger kms) 1945-2014 for five largest Australian cities, public transport vs private transport (source data: BITRE)

Malcolm Turnbull should take a look at the exhibit above, as it shows that public transport’s share of motorised travel in Australia’s five largest capital cities has been depressingly low for decades, while private travel (essentially cars) continues to dominate.

There have been some small blips but no dramatic improvement, notwithstanding the well-publicised drop in per-capita driving and, just the other week, a report that over-80s are now more likely to drive than millennials aged 18-24 years.

Quite simply, public transport isn’t even coming close to winning the “fight” with the car. Back in 2002, Melbourne’s strategic plan set a target of 20% mode share for public transport by 2020. There was a well-publicised patronage surge over 2006-09 due mainly to unprecedented CBD jobs growth, but mode share stabilised since then at just 11%.

I noted recently that simply building more mega public transport projects won’t, by itself, be enough to seriously improve public transport’s mode share.

That proposition probably sounds counter-intuitive and demands further explanation. So consider the impact the following proposed major rail projects would have on car use if and when they’re completed.

Sydney CBD and south-east light rail: This 12-kilometre line is estimated to cost $2.1 billion, however 75% of forecast patronage will come from existing bus users. It’s anticipated existing motorists will account for only 17% of users.

Doncaster Hill rail extension: The Victorian government’s feasibility study for the proposed $4 billion to $6 billion rail line in middle-suburban Melbourne concluded that 98% of forecast patronage on the line would be diverted from existing public transport services.

Rowville rail extension: The proposed 12-kilometre rail extension to Rowville in outer suburban Melbourne is likely to cost something like $3 billion, but the government’s feasibility study found it would reduce the number of car trips on a typical weekday in 2046 by just 15,000; that’s trivial in the context of the number of trips Melburnians currently make by car each day.

The study also concluded it would increase the share of all trips carried by public transport in the metropolitan area in 2046 from 12.6% to just 12.7%.

Melbourne Metro: The latest estimate of the cost to build the planned nine-kilometre tunnel under the CBD is $11 billion. According to the government, the project will provide capacity for an extra 20,000 passengers during the peak. That’s a significant increase in CBD peak capacity (it’s bigger than the West Gate Bridge) and there’ll be network-wide benefits in terms of greater reliability. So it’s worth doing.

But it’s the veritable drop in the bucket in terms of generating mode shift given the size of the total metropolitan transport task.

Although I don’t have formal mode change estimates, the proposed $2 billion to $3 billion Melbourne Airport rail line would replace the existing SkyBus (note Brisbane’s established Airtrain has just 8% mode share). Brisbane’s proposed $5.5 billion Cross River rail line, like Melbourne Metro, is aimed at increasing peak capacity through the CBD.

My point isn’t to deny the benefits mega projects bring to public transport users; most of the current raft of rail projects touted by state governments are worthwhile improvements. Rather, it’s to point out that even very costly rail projects — and this group collectively would cost $25 billion to $30 billion — won’t by themselves have a big impact on car use.

Why would so much money have such a small impact on mode split?

In part, it’s because most new rail projects serve work trips to the CBD, where congestion and high parking charges mean public transport’s mode share is already very high e.g. 80% of motorised trips in Sydney.

Most importantly though, it’s because Australia’s capital cities are highly suburbanised; more than 90% of the population and around 70% of jobs are more than five kilometres from the CBD. Moreover, only around 20% of trips are for the purpose of getting to work.

Even shiny new rail lines can’t compete well with the car for local and relatively short trips in the suburbs e.g. for shopping, personal business, school.

Given the breathtaking cost of retro-fitting rail, it’s implausible to suggest that low-density Australian cities can effect a significant shift away from driving just by building more and more public transport infrastructure.

Nevertheless, while they wouldn’t look much like the Paris Metro, Australian cities can and should have modest but effective metro-style systems that work with our low-density cities. These would be based largely on existing train lines networked with a “grid” of frequent bus/light rail services operating with priority in existing road space.

But if the objective of policy is to effect significant mode shift (e.g. where, say, public transport at least has a larger share than cars across all trip purposes), much stronger policies that actively suppress the competitiveness of private vehicles relative to public transport will be required.

In fact, unless we’re prepared to take road space and traffic light priority away from cars so that buses and trams aren’t caught in traffic, it’s doubtful Australian cities can achieve even a modest increase in the mode share of public transport.

And don’t get carried away by the mode shift potential of higher densities either. A study of US cities by Gilles Duranton and Matthew Turner finds that a 10% increase in residential and job density would, on average, produce only a 1% reduction in driving.

Turnbull is right to talk about investment in public transport and changes in land use. But it’s nowhere near enough; he should go beyond rhetoric and actively push policies, like those recommended yesterday by Infrastructure Australia, that directly address the key underlying problems, especially cars.