Should Public Transport Be Subsidised?

This is a guest post by Joseph Corbett-Davies.

A few months back National leader Christopher Luxon kicked off some controversy after his unexpected comment that public transport should “stand on its own feet”, and that it should not be “subsidised or underwritten.” This resulted in immediate challenges from the media and Labour about whether he was really suggesting a reversal of decades of public transport subsidies, maintained under governments of all stripes. 

Luxon has since walked back those comments, but not before pundits called out his hypocrisy as former CEO of a government-subsidised public transport provider—Air New Zealand—and much was said, correctly, about the huge subsidies that private motor vehicles currently receive. But in all this, I thought there was a lack of conversation about exactly why we subsidise public transport in the first place. Why shouldn’t it “stand on its own feet”, without government intervention?

Here are summaries of a few of the common (and not-so-common, but I think strong) arguments for government support of public transport (PT). Where possible, I’ve tried to come up with an approximate dollar amount of subsidy that might be justified in the context of the Christchurch bus network.

Equity and Access

Public transport is a basic lifeline service for those who can’t afford, or are otherwise unable, to get around another way (in Aotearoa, 13% of people over 16 do not hold a drivers licence). Subsidised transportation options help people of all income levels and situations to have a decent standard of living and the ability to participate fully in society. This clearly justifies public funding to allow these bus and rail services to operate in areas where they might not otherwise cover their costs. 

How large should this subsidy be? This is definitely a political question, with no “correct” answer. To get an approximate size of the current subsidy, I’ll guess that maybe a quarter of current Christchurch bus operating costs—around $18m per year—goes towards providing a basic “coverage” network, which makes it possible to get around without a car. 

Coverage vs. patronage tradeoff, from ECan’s 2018 Regional Public Transport Plan

But public transport doesn’t tend to operate in typically poorer small towns and rural areas, instead being focused on relatively wealthy urban areas. In large part, bus and rail systems in major centres cater to well-paid commuters who could probably afford to pay the full cost of their transportation choices. Should we be subsidising public transport even for these people?

Competing Subsidies

A common response to Luxon’s position was to point out that all transport modes receive government support. Motor vehicles in particular receive heavy subsidies, both explicitly—for example government funding from general taxes to build and maintain roads—and implicitly—by failing to pay for costs relating to air pollution, traffic collisions, and congestion (more on that below). Generally we want all transport modes to be on an even playing field, so that people’s travel decisions aren’t distorted by uneven subsidies.

In an ideal world most of these motor vehicle subsidies would not exist, so the appropriate PT subsidy to compensate would be zero. But in our current transport system, we can justify significant government support for public transport just to equal the subsidy provided to other modes. As an example, in Canterbury, spending on road transport is 78% of transport expenditures, but motorists only directly provide 56% of revenues. Essentially, motorists only pay 70% of the costs of providing the road network (and this doesn’t include costs of using public land for roads). If we apply that same subsidy ratio to public transport fare revenue—something like $19m in Christchurch—then we should be subsidising PT by at least $8m per year. Accounting for other car subsidies would no doubt increase this figure.

Transport funding (NZ-wide) and expenditures (Canterbury). Road user revenues and expenditures are in blue. (source)

Carbon Emissions

Another common justification for public transport subsidies is to reduce carbon emissions. All transport emissions are covered by New Zealand’s Emissions Trading Scheme (ETS). The ETS caps the total amount of carbon and other greenhouse gases that may be polluted and auctions off the rights to produce emissions within that cap. As the cap is reduced toward net zero, the price of a unit of emissions will tend to increase, as a smaller number of units is available to the market and demand exceeds supply. This has the nice effect of automatically shifting prices and financial incentives to reflect the cost of emissions—the lower-emissions option becomes the cheaper one.

In the context of transport, riding the bus will become relatively less expensive than driving, as it produces fewer emissions. This will encourage a switch to public transport, which in turn will justify increased investment in PT service and infrastructure. Note that this is not an explicit public transport subsidy, just the desired effect of the ETS.

Given that the ETS exists, do additional public transport subsidies, with the aim of reducing emissions, make sense? This is where things get a bit contentious. I am mostly convinced by the arguments that, because the ETS sets a fixed cap on emissions, any reduction in transport emissions demand due to a PT subsidy will be mostly or entirely offset by increases in other sectors of the economy. There are disagreements on how much this “waterbed” effect occurs in practice, given the complexity of the ETS. But, on balance, it seems to me that public transport funding is an indirect, costly, and quite possibly ineffective way to reduce total emissions.

Traffic Congestion

External congestion costs occur when the addition of a vehicle on a busy road slows down all other drivers on the road. Because congestion grows exponentially, the total delay imposed on other road users is more than the delay the additional driver experiences themselves—so these costs are not fully paid for by road users, either in time or money. Congestion pricing, as is proposed for parts of Auckland, makes drivers pay for some of these external congestion costs, which is probably the best way to fix congestion. Christchurch is not likely to have congestion pricing anytime soon, so are public transport subsidies a good second best option? 

If public transport provides an attractive alternative to a congested road, then PT subsidies can shift more trips to buses or trains. However, a subsidy on its own is unlikely to totally or permanently eliminate congestion. Road space freed up by people switching to public transport will be partially reclaimed, as people make new trips or shift their journeys closer to rush hour in light of the reduced traffic. But even if congestion doesn’t vanish at peak times¹, these new or shifted trips are still valuable to the people making them. The size of the optimal PT subsidy is hard to quantify, but for context a study in 2014 estimated congestion costs of $54m per year across the Christchurch strategic road network. 

Economies of Scale

Economies of scale are a little-known but, I think, compelling argument for public transport subsidies. The real magic of regular transit service, like buses or rail, is that it gets better with more riders. If, say, the number of passengers on a service doubles, and the number of vehicles is doubled to match, then not only does the expanded system accommodate the new riders, it also means that all existing riders have their wait time cut in half, since vehicles arrive twice as often².

Unlike the external congestion costs of driving, someone taking the bus produces an external benefit, as on average their ridership will prompt a small increase in service frequency, reducing wait time for all other passengers. Just as the correct economic response to external costs is a tax or a fee, the external benefit of public transport justifies a subsidy to maximise the overall benefits of the system.

Economists have developed some mathematical models based on this intuition which can be used to estimate the optimal level of subsidy. Herbert Mohring’s original paper on the subject uses some maths to show that for a simple ridership model the government subsidy should equal the total time cost of waiting passengers. In 2019 there were 13.67 million boardings per year on the Christchurch bus network. Assuming an average wait time of 7.5 minutes and a $20/hr value of time, these economies of scale alone justify $34m of subsidy per year.

Summary

Combining these figures gives a conservative estimate for what level of subsidy should be provided for public transport in Christchurch—around $60m per year. This is without including any accounting for congestion and (debatably) emissions reductions.

Factor	Justified Subsidy In Christchurch
Equity and Access	$18m (approx. current funding level)
Competing Subsidies	$8m+
Economies of Scale	$34m
Total	$60m+

In comparison to the $50-55 million per year that central and local government contribute to PT operations in Greater Christchurch, it seems that the current funding level is probably on the low side. Far from needing to “stand on its own feet”, I hope this overview has shown that we should definitely be subsidising public transport in New Zealand.

_{¹  Without congestion pricing, durable improvements to traffic congestion mostly come from providing faster and more reliable transport alternatives — like bus lanes and rail systems that don’t get stuck in traffic.}

^{2 Equivalently, if capacity is increased by adding new routes, then the walk or travel time to reach the service is reduced.}