One of the difficulties of working out the cost of transport is determining which costs to include. Users of public transport pay a fare for each trip that they use, but this only covers part of the total operating costs and the remainder is paid for by way of government “subsidy costs”. Meanwhile, car users pay a small “variable user cost” to drive (mainly petrol), but also have significant fixed costs (depreciation, finance, registration, insurance, etc) known as “other private costs”, and then there are negative “externalities” (congestion, pollution, etc).
Economic theory dictates that when making decisions about which mode of transport to use, it is the marginal cost that matters (assuming non-price variables, e.g. trip duration, are held constant). Fixed costs, which must be paid regardless of whether you make a trip on that mode, or external costs, which are paid for by others, do not factor in. So which mode of transport has the lowest variable cost, and which mode has the lowest overall cost?
Dr Garry Glazebrook of the University of Technology Sydney quantified these amounts on a passenger kilometre basis and found that while total costs vary between 47c/km to 86c/km, variable costs are just a fraction of this, ranging from 11c/km to 19c/km. Cars have by far the highest total cost at 86c/km, with 38c of that being externalities (20c alone are due to congestion). Car users pay a total cost 48c/km to drive their cars, compared to 11c/km for trains or 19c/km for buses, but only 14c/km of that is variable cost (the remaining 34c/km are fixed costs that are paid upfront). As there isn’t a big difference in variable cost between modes, car owners tend to opt for the fastest option, which generally for them means driving.
Additionally, most car owners have a car “just in case”, though many in the inner city are opting to use car share services in lieu of owning a car, something which encourages public transport use by transferring the fixed cost of running a car into variable costs, but also still providing car access “just in case”.
Given that, on a per km basis, the subsidy to trains and buses is on par with the external costs from cars, it appears fair that a subsidy is provided in order to bring all modes of transport back to a level playing field. However, car use still remains the cheaper option for many people because of the way its costs are structured primarily as fixed, with only a fraction of the total cost being variable. This encourages driving and discourages other modes of transport.
None of these consider capital costs (i.e. the cost of building a new railway line or freeway and purchasing new trains or buses). In the case of cars, it also appears to exclude the cost of maintaining roads. However, this tends to be quite low compared to how heavily cars are used. That is why, despite cars accounting for 90% of all trips, the NSW government spends $900m a year on road maintenance (Source: Daily Telegraph), but $1,400m in operating subsidies to Cityrail alone (Source: Cityrail 2010/11Annual Report, page 53). This does exclude maintenance performed by local councils, but also leaves out the cost of running buses, ferries, and trams, so remains a good rough guide.
A recent suggestion to solve this imbalance has been to link registration costs to kms travelled. This system would use GPS technology to track how far car users drive during peak hour, and registration fees would be linked to that distance. Off peak trips would be free. This way, those that most used the roads would also be the ones who paid the most who used them, and those who don’t use them much would not be subsidising them. Additionally, by shifting costs from fixed to variable, it would use price signals to incentivise public transport while disincentivising car use, without actually increasing the cost burden on the average road user.
Many people counter this by pointing out that they would not or could not shift their travel times, and so would be forced to pay for something that they previously got for free. This is incorrect for 2 reasons. First, this replaces registration fees, and is just a different way of paying for something. Second, it doesn’t have to get everyone to change their travel habits, only a small minority, in order to see improvements. As an example, school holidays see only a 5% reduction in cars on the road, yet often result in a 10%-20% increase in average speeds. Congestion charging could see similar results, and then those who are unable to shift their travel times would enjoy a faster journey. In short, everyone wins.