The case against zero-priced public transport
Zero priced is not “free” — and may be less equitable and more wasteful than charging
Half-price public transport good, zero-price better. That seems to be the consensus view of recent NZ media coverage1 and lobby groups2. They forecast extensive benefits from zero-priced public transport (PT), including:
improved fairness/equity;
mode substitution from cars to PT;
reduced transaction costs (by eliminating fare collection);
increased PT patronage, supporting expanded PT services; and
reduced environmental costs.
I’m going to play the contrarian here, as the opposing case seems to be largely missing from the media. I conclude that zero pricing may be less equitable than positive pricing. Nor is it clear that the other benefits claimed will be realized.
But first, lest you think that I have something against zero pricing in general, let’s explore some situations where zero pricing can perform better than the alternatives.
Zero pricing is sometimes optimal
In An economist staggers into a hospital I made a case for zero-priced emergency care in NZ hospitals, reflecting the merit-good characteristics of emergency health care.3 Here I offer two further examples, drawing again on personal experience4.
Side story 1: when collection costs exceed revenue collected
In the early 1990s the Tasmanian Parks & Wildlife Service started charging for entry to National Parks, which had previously been zero priced. The Service erected toll booths at major road entrances to parks, at which operators collected a fee from passing cars. The fees led to a substantial reduction in visitor entries (particularly from locals making short trips into the parks), yet it failed to raise sufficient revenue to pay for the wages of the toll collectors.
This situation is pretty much the textbook definition of a non-excludable good, in which it is physically or financially impractical to charge a positive price. Zero is the best available pricing option for such goods.
Undeterred by textbook economics, the Parks & Wildlife Service dismantled its toll booths and changed to an honesty-based system with random enforcement. That system generated net positive revenue for park management. Today it operates mostly online, with even lower collection costs.
This demonstrates a more general point: excludability is not an inherent characteristic of a good. It depends on other factors, including the costs of revenue collection.
Side story 2: why zero pricing works for search & rescue
Search & Rescue (SaR) is expensive, costing the NZ Government $42m in 2021/22. But it’s zero priced for those rescued. This sounds like a great deal for them and their families. But, because of the unusual cost structure of SaR, zero is an economically efficient price.
SaR is a bundle of two services. Search is typically very expensive. Land searches for trampers & hunters have run to hundreds of thousand dollars, while sea searches for yachties & fishers have run into the millions. By contrast, Rescue from a known location is cheap. A helicopter pickup with a paramedic crew might cost the public a mere $5000.
Authorities would rather lost and injured people called for help early, undeterred by potential rescue charges. It is worth the public covering the cost of some seemingly unnecessary rescues, to reduce the likelihood of even more costly searches.
Summary: when zero pricing may be optimal
Zero pricing may be optimal:
when collection costs exceed revenue (i.e. for non-excludable goods);
for merit goods (i.e. those that society decides people should receive regardless of their willingness or ability to pay) when it is difficult or impossible to use differentiated pricing (i.e. charging different prices to different groups of customers); or
when suppliers face asymmetric cost structures (e.g. SaR example).
PT is not an obvious fit with these optimality conditions. Clearly, it is an excludable good, as PT fare collection typically raises positive net revenue. And while it could be argued that PT is a merit good, differentiated pricing is already the norm. Lastly, there are no obvious parallels with the SaR example.
Zero priced is not “free”
Advocates prefer the term “free”, as it carries the implication that everyone is getting something for nothing. Economics, befitting its popular designation as the dismal science, teaches us that this is almost never true.
In New Zealand (and many other places), PT fares cover a small proportion (typically around a quarter) of the total costs of provision.5 The remainder comes from local government, other transport users, and central government. The local government contribution is mostly funded by rates — paid directly by property owners, and indirectly by renters.6
Zero pricing nobbles demand management
PT typically has high fixed costs, low variable costs, and peaky demand. Adding extra capacity to cope with increased peaks can be very expensive, sometimes prohibitively so.
Demand management potentially benefits everyone. For example, lower prices for off-peak travel encourages people with travel-time flexibility and price sensitivity to shift their travel to off-peak times. This reduces congestion at peak times, and delays the need for expensive system upgrades. Demand management can also improve the financial viability of off-peak services, supporting more frequent services and benefiting those who need to travel at those times.
Demand management using differentiated pricing is, however, incompatible with zero pricing.
Zero pricing eliminates equity pricing
PT fares typically vary by characteristics of the traveller. Retirees (i.e. Gold Card holders), school children, infants etc. attract lower fares. This type of price discrimination — I call it “equity pricing” — is broadly progressive, as those with a higher ability to pay subsidise those with a lower ability to pay, on average.
Zero pricing eliminates equity pricing, replacing it with ratepayer and taxpayer funding. The progressiveness of NZ’s rate and tax systems is subject to hot debate, which I don’t want to go into here. But claims that zero pricing of PT will improve equity are hollow without a robust investigation of the relative progressiveness of funding alternatives.
However, it’s at least plausible that neither the rate nor tax collection system is as progressive as is equity pricing. Should that be the case, zero pricing would be less equitable than positive pricing.
Zero pricing transfers costs to non-PT users, who may be even less able to pay
Zero pricing of PT means recovering costs through the rates and tax systems. These systems do no discriminate between PT users and non-PT users. Non-PT users will thus be further subsidising PT users. Is this fair? It’s hard to make a case that poor non-PT users should further subsidise rich PT users.
Potential PT users may favour quality improvement over zero pricing
Market research, such as that reported in NZTA’s 2022 research note Impact of Half-Price Public Transport Fares, indicates that value for money and affordability are not overriding factors in choosing PT. Depending on the question asked, other service characteristics and quality-related factors can dominate price.
Zero pricing can induce waste
The marketing literature emphasises the special "pulling power" of a zero price, relative to low non-zero prices.
But when a price is zero, people find different, and sometimes wasteful, ways to use the service. My favourite example7 involved apartment dwellers in the Soviet Union keeping a gas burner alight all day, even when they left their apartment. The reason? Gas was zero priced, but they had to pay for matches to relight the burners. One consequence was more frequent apartment fires! Waste, by definition, is antithetical to reducing environmental costs.
Zero-priced PT could lead to some people treating buses and trains as mobile lounges. This could, in turn, make PT less attractive for other users. The effect on total PT use could go either way, but note that behaviour of others and personal security have greater impact than do value for money and affordability in the graphic above .
Zero pricing can induce more demand than a system can reasonably supply. As Michael Cameron wrote in The Great Walks may be free, but they are not free:
When a good or service has no monetary cost, there will almost always be excess demand for it — more consumers wanting to take advantage of the service than there is capacity to provide the service. Excess demand can be managed in various ways — one way is to raise the price ... Another is to limit the quantity and use some form of waiting list (as is practiced in the health sector). A third alternative is to degrade the quality of the service until demand matches supply (because as the quality of the service degrades, fewer people will want to avail themselves of it).
Neither waiting lists nor service degradation are obviously better options than positive prices for PT.
Lowering PT revenue isn’t an obvious way to expand PT services
If the overriding goal is to expand PT use, then lowering the revenue available to pay for PT infrastructure and operations will likely work against that goal.
Indeed, raising prices could be an even better strategy. If demand is more sensitive to service quality than price, then potential users may be happy to pay more for an improved service, as opposed to paying less for a degraded one.8
Place-dependent PT prices get capitalised into land prices
Reduced PT fares lower the cost of living in those places well served by PT. Who benefits? Those living in such places, who make use of PT. And if it’s now more attractive to live in those places, then we would expect the price of land to rise, along with rents. Who benefits? The existing owners of land, whether they use PT or not. Who loses out? Those priced out of the places with reduced PT fares, including renters. Understood this way, zero-priced PT may redistribute from the less well-off to the well-off.
Pricing is a very powerful tool — but real power comes from matching solutions to problems
Zero pricing might sound wonderful, like something is being created for free. But not every socially valuable thing can, or should, be zero priced. Each case should be examined on its merits.
There are solid arguments against zero-priced public transport. In particular, zero pricing can induce waste, and necessarily comes at the expense of demand management and of equity pricing. There is a much stronger case for positive pricing.9
By Dave Heatley
[Minor error in paragraph under Place-dependent PT prices corrected 14 May 2023.]
See, for example: Free fares campaign; Free public transport for everyone; Restore passenger rail; and Now is the moment for free fares. (Note that some of these groups are advocating zero prices for all PT users, others for a subset of them.)
The many uses and definitions of merit good can be confusing and contradictory. I use the term to describe goods (typically services) that society decides people should receive regardless of their individual willingness or ability to pay. Merit-good supply can be implemented by zero pricing that good for all potential consumers, or just for those with restricted willingness or ability to pay.
Full disclosure: I was a volunteer for NZ Land Search & Rescue between 2009 and 2022, and wrote software for the Tasmanian Parks and Wildlife Service visitor charging system in the early 1990s.
“The costs of local public transport services are funded between three main parties, i.e., users of the services (through fares), regional councils and central government (through WK [Waka Kotaki/NZTA]). The total gross costs of some $1300 million pa (2018/19) were funded approximately 28% through fares, 31% by regional councils (which recover these costs mainly through regional rates) and 42% by central government (recovered mainly through general taxation).” From Ministry of Transport (2022) Domestic Transport Costs and Charges Study. Public transport use fell substantially from April 2020 due to Covid, and is yet to recover to 2018/19 levels. In addition, the Government’s “temporary transport cost reduction package”, which took effect on 1 April 2022 and was recently extended to 30 June 2023, further reduces the proportion of costs recovered from fares. A reasonable guesstimate is that fares are currently covering less than 14% of costs.
What is the incidence of rates in rental housing markets? I’d appreciate a pointer towards any research on this topic. My assumption, at least in supply-limited rental markets like most of NZ, is that landlords, when setting rents, (largely) pass rates charges onto renters.
A further example: Spammers face a zero marginal cost to send spam emails, so send far too many of them from a social welfare perspective. See, e.g., Robert E. Kraut, Shyam Sunder, Rahul Telang, & James Morris. (2005). Pricing electronic mail to solve the problem of spam. Hum.-Comput. Interact. 20, 1 (June 2005), 195–223.
Bus services between my home and Wellington city are less frequent, less reliable, and travel via longer routes, than before fare prices were halved in April 2022. While I’m not attributing cause and effect, I would much prefer the previous service level, and be happy to pay higher fares for it to be restored.
I’m not arguing in this post that current PT fares in NZ (whether at “half” or “full” levels) are optimal in any sense — I haven’t done the work to offer an opinion on that.
Great article. One way to resolve almost all the issues you raise would be to to introduce peak pricing and make it free at other times. For example if the present bus fare for a worker travelling to work between 7:30-9:30am and returning home between 4:30-6:30pm was $3 per trip, it could be raised to $6 per trip representing peak-demand pricing. Travelling outside these times would be zero priced. This would provide equity pricing and shift demand to other times of the day. It would also improve the service at peak periods, as people would expect a better service as they were paying more for it, which is important for people getting to work. As there are fewer alternatives for workers demand would still be high and drive a revenue stream to expand the service.