Premier Denis Napthine’s Liberal government has promised Melburnians a train line to the airport — it’s an election promise Crikey has set in stone. But does Melbourne really need a train to Tullamarine?

The Victorian government hasn’t disclosed the anticipated cost of delivering the project — it’s rolled into the global $11 billion total for Melbourne Rail Link (which also includes a CBD tunnel). It’s almost certain, though, that providing a rail service between Melbourne Airport and Southern Cross station would cost at least $2 billion to $3 billion.

That’s serious money, so it’s important to think hard about whether or not it would be a sensible use of public funds. In particular, it’s vital to consider if there might be more pressing priorities that funds of this magnitude should be spent on.

It’s a lot more than the $1.2 billion (in 2012 dollars) it cost to build the 10-kilometre Sydney Airport Link, which opened in 2000. And it’s almost an order of magnitude higher than the $0.4 billion (in 2012 dollars) it cost to construct the 15.9-kilometre Brisbane Airtrain line in 2001.

In both those cases most of the funds were provided by private investors at their own risk. That’s much less likely today given recent bad experiences investors have had with transport projects.

Yet despite their much lower capital cost, both the Brisbane and Sydney lines struggled financially in their early years. The Sydney service went broke and the current owner only has a profitable business now because it bought the operation for around a third of the original construction cost. Brisbane’s Airtrain struggled for 10 years and got by on low frequencies and limited hours of operation (up until 2011 the last train ran at 8pm).

Some argue that the government should nevertheless build the line because having a train would provide numerous advantages, including reducing congestion on freeways and redressing the high cost of airport parking.

That argument ignores the fact there’s already a successful public transport service — SkyBus — connecting Southern Cross station to the airport. A rail line won’t miraculously relieve congestion on Melbourne’s freeways any more than expanding road capacity will; what it’ll mostly do is replace one form of public transport with another.

SkyBus provides 10-minute frequencies for close to 24/7. The trip averages 20 minutes off-peak (faster than the proposed rail service’s envisaged 25 minutes) and 40 minutes in peak times. Buses already have a 14% mode share at Melbourne Airport; that compares to less than 10% for Brisbane’s Airtrain and 18% for Sydney’s Airport Link.

While the government promises its airport train will also operate at 10-minute frequencies in peak time, it doesn’t say what will happen off-peak. Brisbane’s Airtrain, which like Melbourne’s proposed line has the disadvantage of terminating at an airport, only offers 30-minute frequencies in the off-peak. Moreover the last service leaves the city at 9.30pm and departs the airport at 10pm. Sydney’s Airport Link has frequent peak and off-peak services, but the airport is a lot closer to the city centre than either Brisbane or Melbourne Airports. More importantly, Sydney Airport is not an end-of-the line station; the line continues beyond the airport to service non-airport travellers.

The financial risks are minimal with SkyBus because it already covers its capital and operating costs and indeed contributes to the cost of improving the freeway. It also has considerable potential to add capacity by increasing frequencies. Speeds can be increased by giving it higher priority on roads at considerably lower cost than building a new rail line. It’s not always as comfortable as a train and it’s slower in peak times, but it’s costing taxpayers nothing.

Like SkyBus, a train isn’t a silver bullet; it won’t do a lot for the great majority of travellers. They’re not going to or from the CBD, but to dispersed destinations across the metropolitan area. They’re still going to kiss ‘n’ ride, drive and park, or take a taxi (especially if they’re visitors), because it’s still likely to be faster and more convenient in the great majority of cases.

There’s a lot of public concern about the high cost of parking at Melbourne Airport, but a train won’t solve that problem any more than SkyBus has. It’s caused by anti-competitive practices, and they need to be addressed directly. A train won’t do much to reduce carbon emissions either — it would be one of the most expensive ways of mitigating carbon imaginable.

And nor is it likely that any future Victorian government would price an airport train at the standard Myki fare. But if one did want to, it could simply subsidise the SkyBus fare (currently $18 one way) directly without incurring the expense of building and operating a new train line.

It’s likely that sometime in the future a train to Melbourne Airport will make sense on operational grounds. This isn’t currently expected until sometime in the decade after 2030, when passenger volumes through the terminal are forecast to be double today’s levels, but it could be much later as it’s difficult to predict future changes in economic circumstances, technology and the like.

When to provide a train, though, is not a decision that can be made on the basis of some rule of thumb; cities differ, so it would require a comprehensive study of the special circumstances of Melbourne. Nothing remotely like that sort of analysis has been undertaken for airport rail in Melbourne for a long time.

The main benefit of an airport rail line compared to scaling up SkyBus would be a faster and more reliable trip for travellers journeying between the airport and the CBD in peak hour. The key question though is whether that benefit justifies the massive capital and recurrent costs it would entail, and whether those scarce public funds might provide greater social and economic benefits if they were instead applied to other projects.

In my view there’s no shortage of competing infrastructure needs that make a lot more sense at this time than the government’s plan to build airport rail, e.g. improved rail signalling. Disappointingly, the Sunday Age reports that the Greens also support building the airport rail line; to its credit, Labor doesn’t regard it as a priority at this time.