In this recent post on Auckland’s transport funding gap, Peter Nunns espoused the merits of time-of-use transport pricing as a way of increasing the productivity of the transport system. Peter came up with this analogy to help highlight the merits of time-of-use transport pricing:

Let’s say you’re managing a factory. Your machines are running at 100% utilisation ten percent of the time, and 20% utilisation the rest of the time. This is constraining your ability to produce more, so you ask the chief executive for money to buy more machines. His answer, if he’s got any sense, will be: “Get knotted. You need to manage your workflow better.”

As some of you may know I also support time-of-use transport pricing (articulated in posts here and here). Notwithstanding my general support, I do accept there’s important design and implementation issues to work through. For this reason I’m relatively relaxed about timing and would prefer that – instead of rushing headlong into a particular solution (as the NZCID would have us do) – we instead took the time to have a decent/informed public debate about the concept. I hope that such a debate would end with the majority of people supporting the idea in principle, which would then enable political/technical leaders some space to work out exactly what we should do, how we should do it, and by when.

In this post I simply wanted to reflect on two of the dissenting comments raised in response to Peter’s post. For example, one person commented:

My understanding is that NZTA and the government are principally opposed to a discriminatory tax on assets that have already been paid for.

This statement suggests that time-of-use road pricing is a “tax”. From what I understand, taxes are typically used by governments to fund any number of activities that are largely unrelated to the activity being taxed. Income tax and GST, for example, are used to fund a whole manner of things, such as education and health. In contrast, revenue raised from transport activities (at least in New Zealand) are hypothecated to operating, maintaining, and improving the transport system (the MED website provides a breakdown of the various duties, taxes, and levies that are applied to liquid motor fuels in New Zealand for those are you who are interested).

The second part of the aforementioned comment  part inferred that time-of-use transport pricing was wrong because the “assets have already been paid for“. I don’t think this is a credible argument for several reasons.

The first reason is that it presumes new roads do actually pay for themselves, insofar as the fuel duties paid by drivers are adequate to cover the lifetime costs of construction and maintenance. This is patently not true of local roads, where costs are part-funded by rate-payers, not road-users. I suspect it’s also not true of many of the RoNS, which would certainly struggle to cover their costs. Hence, it’s not really the case that new roads are paying for themselves, it’s just that their financial ass is being covered by old roads that have more than paid for themselves. I don’t know when roads actually “break-even” (and suspect it varies a lot from place to place), but suspect that many of our recent improvements don’t come close.

But the more critical issue with the suggestion that the “assets have already been paid for” is that the costs of a road do not end with construction and maintenance.

More specifically, roads incur ongoing congestion costs, which tend to arise at  peak times. Now I do understand that congestion is a less tangible economic costs than construction and maintenance costs, which must be funded directly out of the transport budget. In contrast to these costs, congestion is an external, non-monetary cost. When we say “external” we are referring to the fact that the person who chooses to drive in congested conditions does so without having to bear the additional delays they cause to other drivers that are stuck in the queue behind them. Hence congestion is an external cost that arises from our individual decisions to drive. Despite being less tangible, it’s nonetheless real – as any commuter will know!

So rather than being a tax, time-of-use transport pricing is, I think, better seen as a targeted user-charge. One which seeks to place the costs associated with travelling at peak times at the feet of the people who are making those decisions. And given that much of our transport budget is currently being spent on transport projects that are designed to cater for people that are travelling at peak times, then it makes sense to charge more for these trips than for trips that take place at off-peak times.

Having raised those two issues, the same person went onto state:

The charge is a mechanism to force people to change the mode they use to travel. In the case of congestion charging it makes PT more attractive in comparison to driving … In effect you are shifting the equilibrium in favour of the poor using PT and the wealthy driving.

My first issue with this statement is the presumption that the primary benefit of time-of-use road pricing is that it changes existing behaviour. Naturally modal shift from cars and to PT would occur and result in less congestion. Such a benefit can be thought of as a “static efficiency“, in that the benefit arises from improving the efficiency with which people currently travel. In saying that the experience with time-of-use road pricing overseas is that 80% of people keep driving, i.e. the majority of people kept doing what they have always done, and there is very little change in PT use.

So I would argue that the primary economic benefit of time-of-use road pricing lies less with it impacts on people that are currently driving, and more with how it impacts on people’s future decisions about where they live/work/play; it impacts on our future land use and transport decisions.

Benefits from more efficient decisions being made in the future can be thought of as “dynamic efficiencies“. By sending a price signal about the relative scarcity of road space at peak times, time-of-use road pricing will progressively encourage people and businesses to make choices in the future that help them to avoid situations the need to drive at peak times. Thus, the dynamic efficiencies of time-of-use transport pricing will accrue progressively over time, by discouraging people from making decisions that result in inefficient transport and land use outcomes.

I would suggest the dynamic efficiencies of time-of-use transport are more important in a city like Auckland, which is expected to grow rapidly over the coming decades.

As an aside, the potential for time-of-use road pricing to deliver dynamic efficiencies is a major reason why I don’t buy the line that “we need to invest in alternatives first” before we consider implementing a time-of-use transport pricing scheme. The reality is that we already have the primary “alternative”: We simply need people to exercise more discretion about where they live/work/play to ensure they don’t end up driving so much at peak times. And the longer we go without time-of-use transport pricing, then the fewer people who make transport/location decisions in consideration of their true costs and the harder it will be to implement such a scheme in the future.

Consider, example, the Auckland Plan’s proposed greenfields development around Hobsonville and Pukekohe. Do you think such locations would be equally attractive with time-of-use road pricing? And do we think that implementing time-of-use transport pricing will be more or less easy once these suburbs have developed and lots of their residents are now driving elsewhere to work? Personally, I think the answers are “no” and “much less easy” – which is why I suspect the dynamic efficiencies of time-of-use transport pricing would quickly dominate the static efficiencies associated with (relatively small) mode shift.

The second issue that I would like to address is the suggestion that “the equilibrium will shift in favour of the poor using PT and the wealthy driving“. The point that is being made here is that time-of-use transport pricing is likely to result in our transport system being prioritised for high-value travel. And it is true that the latter is positively correlated with income.

In saying this, I think it’s important to consider the following issues:

  1. Income is not the primary/sole determinant of people’s willingness-to-pay for travel. I would argue that many other factors determine how much someone is willing to pay for transport, e.g. the purpose of a trip is more important than someone’s income. Speaking from personal experience, I know that if am travelling for work purposes then I am more prepared to pay more than if I travelling for recreational purposes. The key point to note here is that willingness to pay varies greatly depending on the reasons why someone is travelling and while income is likely to play a role (if only for framing their perspective on relative costs), it’s not as straight-forward as the comment above makes out.
  2. It makes presumptions about how a scheme would operate. As mentioned in some of my earlier posts, the degree to which a time-of-use road pricing scheme impacts on low/high income households depends very much on its design and wider policy decisions about how the resulting revenues are used. If the scheme was designed to be revenue neutral, for example, then the additional revenue could be used to lower fuel excise duties and/or improve public transport – in this way resulting in a situation where low income households were much better off (remembering of course that low income households tend to drive less, especially at peak times, and own less efficient cars – hence they pay more fuel tax per kilometre).

Before I wrap up, I should say that I think the whole time-of-use transport pricing debate could be flipped on its head if it was presented as 1) revenue neutral and 2) linked to lower prices for off-peak travel. In fact, what we should actually be discussing is not “higher peak charges”, but higher peak charges that are used to fund lower off-peak charges.

Finally, some of you may have noted my preference for “time-of-use transport pricing” rather than “time-of-use road pricing”. The reason I prefer the former terminology is that it emphasises the general concept, rather than its specific application. Indeed, very similar arguments apply to the use of public transport: Maybe we should consider charging public transport users more to travel at peak times, with the additional revenue in turn helping to fund lower off-peak travel? I think so.

Share this


  1. Great post, I would be supportive of this measure providing these 2 things occur.

    1) Other transport modes reflect Peak/Off peak pricing. That would make this fair
    2) Assuming we are able to get rid of some of the projects creating the funding shortfall (there are a lot of projects that are unnecessary, both rail & PT). I would like to see a decrease on the petrol tax or some other levy making it revenue neutral. This would also make much more politically tenable.

    Lastly one problem might be the collection of the tolls. This will only work if the collection of tolls is cheap though hopefully new technology will help us with that.

    Also that last bit comes off as passive bullying to me, anonymity is fine.

    1. I think that the ideal system is a mix of road tolling and fuel taxes. Road tolling would be the main source and deals with managing demand and raising enough revenue to cover road operations etc. Fuel taxes would then be massively reduced and become primarily based on countering the negative impacts caused but using fuel

  2. Yes, I for one have long been in favour of differential pricing on the Transit network. Off peak journeys should be at a substantial discount to peak ones. The cost of adding additional peak services (especially more buses onto our cluttered roads) is very high yet there is capacity on off peak services to spare. Although, these are, largely inadequate. It also seems likely that this move could be revenue neutral as we know that Transit uptake is price sensitive. So a large discount to off-peak may all that is needed, without any rise in the already high (by international standards) fares at peak.

    1. The need for off peak fares on PT (and roads) was a point I made to the Mayor a week or so back.

      You need off-peak rates to encourage people to self-manage their travel better – why schedule a meeting at 8:30am across town, if you can do it at 10:30am – the transit system, your sanity and the transit users will thank you for the later meeting time if you do, and maybe if you say that the $’s stack up too then thats a win/win/win for all.

      Right now all PT travel has the same price whether it be Monday at 8am, or Saturday morning at 10am.

      Which puts me off using the trains/buses in the weekend – as they’re too expensive and infrequent in the weekend.

      Especially when I can drive to town, park all day, and then drive home at times I set and still be better off financially than if I did the same trip using PT where I have to work with its schedules.

      And when the new EMUs arrive, the cost of operating the trains (and hence the whole PT network south of the bridge) should plummet.
      So reward those who use PT off peak rather than punish.

      1. Yes and in Brisbane the off-peak discount provides a good reason for people to hang out in the city after work for a beer/coffee.

    2. Good point and something I also think is a great idea (as per Vancouver). I don’t tend to use PT in the weekends as it is a)expensive and b)a very infrequent service (every hour – Te Atatu). I can cope with one but not both. Perhaps making it much cheaper would go a long way in the time between now and the roll out of the new contracts / services over the next 2 1/2 years?

    3. “Off peak journeys should be at a substantial discount to peak ones.”

      We’re sort of doing that already with the Gold Card. I know the government are picking up the tab, but end-user pricing is being used as a mechanism to fill up off-peak buses and trains.

      1. The answer is “yes” SuperGold has that effect in a relatively indirect way. TransLink’s approach in Brisbane has been to offer 20% discount on any trip between 9.30am – 3.30pm and after 7pm.

      2. Except that you have to have a Goldcard to get said discount which is 100% to the user.

        And I don’t imagine the off peak discount we’re talking of going forward is “free”.

        For the trains I’d envisage something like 30-50% off the “peak” fare rate as about right.

    1. “snow flake” ignored my previous requests that he not abuse his anonymity. In light of this I thought it was reasonable to pre-empt his comments. Perhaps not?

  3. Weren’t you going to write a post specifically about equity issues in network access pricing in that series arising from Long Bay congestion, Stu? Based on your comments above about income and ‘choice’, that might be useful.

    1. Yes, and Peter and I were actually discussing how we would approach this post over a beer last night. We thought it was better to first focus on responding to the above comments, before writing a specific post on equity issues. Our general theme is that inefficient transport pricing is not the best way to address equity issues, which are better addressed through 1) direct tax and transfer (short term); 2) labour market protections (medium term) and 3) education (long term).

      Thanks for the reminder though!

  4. It should be easier to sell this idea in Auckland than anywhere else in New Zealand. Every motorway user is paying for all of the lanes, not just for the one lane they are using. Tell off-peak travellers (especially low-paid workers such as cleaners) that they are currently paying for all those lanes they don’t need because if those extra lanes cost 10% of the transport budget then 10% of the petrol tax on every litre gets spent on those peak lanes.

    1. Yes, although I suspect that capacity expansions account for more like 25% of the total land transport budget.

  5. You could argue that we have time-of-use pricing for road users now. It is just that the currency is time. Think of it like this, if you needed to go somewhere by car and you had the freedom to pick any time of day, would you pick 8:30 on a Monday morning?

    1. And Brett you’re quite correct, but it seems that the rate isn’t high enough to change behaviour (everyone seems to think that someone else should change their pans!). So it seems there is insufficient elasticity in timing of needs, and as we constantly observe, modes.

      Still we need to improve options if we are to improve flexibility in travel choices.

    2. Yes that’s a fair point. Although to put a question back at you: What type of trips will start to occur in pervasive congestion? My suggestion is very low value trips. So congestion actually “prices” high-value trips off the road, which I don’t think is a very good outcome.

  6. The question of low income workers gets raised quite a bit in these discussions. How much of an issue really is it given that presumably many blue collar workers don’t work in the CBD and probably don’t work regular 8-5 office hours? If you work shift hours, work at a suburban industrial area, work as a CBD cleaner outside regular commuting times how much impact would higher peak time pricing really have? Possibly lower pricing at off-peak times and on roads in the counter peak direction could reduce the cost of transport for some low income workers. The remaining low income CBD workers such as retail workers may already be high users of public transport given the price of parking is probably a big disincentive to driving.

    1. I agree with you, And the follow up question is should we accept wholly inefficient transport pricing because relatively few low-income workers may be adversely affected by a time-of-use price? Or should we look to adopt other mechanisms that can help to mitigate the effects of the charge on these households for a period of time.

  7. Here is one thing that Jarret Walker has brought up, should we call this sort of thing congestion pricing (be it on road or PT), or decongestion pricing?

    1. “should we call this sort of thing congestion pricing (be it on road or PT), or decongestion pricing?”

      “Decongestion pricing” is probably the right term, but it sounds too much like a medicine that tastes like crap.

      And we all hate taking medicine that tastes bad – no matter how much everyone tells us that it is “good for us” right?

      And “congestion pricing” – thats sounds like an additional tax/punishment (which it is) – but one that I as a car driver believe other people will pay (but not me).

      So on the surface, congestion pricing is the winner for the connotations they both have, even if both terms means pretty much the same.

      1. Do you really think it’s fair to describe time-of-use charging as a tax? Rather than a more targeted user charge?

        Do hotels “tax” their customers? I think not.

        1. Stu,
          My use of the word tax is deliberate and is part of the perception versus reality for the term.

          From the perspective that if you believe you have already paid for a service and are then charged “extra” for using it.
          Then yes it seems its like a tax.

          For roads, there is that perception that I (as a road user) already paid for the road, so why do I have to pay (again) to drive on it?
          Not saying its right – but thats the view many people have of road or whatever pricing.

          As for Hotels taxing customers?

          In the US (for example) many do, its called variable room charge – they charge you what they think you’ll pay.
          In Manhattan, NY some of the hotel rooms have a plaque on the wall showing somewhere (usually at the back of the wardrobe!) the minimum and maximum rates the hotel can charge for the room. And there can be a substantial difference (500 dollars a night difference) between the two. Is that variable rate not “taxing” your customers?

          Also, in the US as well, its common to be charged a “service charge” supposedly to cover tips and the like to staff, so you don’t have to tip them. And despite which you are already having to tip the waiters and the like as you know that the service charge never makes it to the workers.

          So if that not also a “tax” by another name and made by the hotel?

          Yes you could chose another example than a hotel, but the principle of paying more than once is the sticking point.
          “User pays” is similar.

          – my argument with many of the user pays services the Council (or Government) provides, is that they are user pays in principle but not in reality.

          Its truly “user pays” only if you when you don’t use it – you don’t pay anything, ever.

          Example of this is water – its in theory user pays, except there this “daily charge” which is a fixed fee no matter the amount of water you use. Which in my case exceeds the user pays by a good long way.
          So even if I don’t use the water, I still pay for the possibility of using it.

          Same for power, and of course roads – through rates and registration fees – even if you never actually drive on the road, you pay a lot for the possibility of being able to.

          So this is an example of the local council having it both ways – user pays, and fixed charges.

          And If there is one nagging suspicion about this pricing stuff, is that while it may start out as purely one or the other,

          Sooner or later politicians will meddle and then we’ll end up paying twice for the same thing – like the service charges at US hotels.

          So not only do you have to fight the perception (if not reality) that it is a tax, you also have to fight the perception that we’re being charged ore than for the same thing.

          Not easy to overcome that perception/belief and that why the pricing system has to be clear and transparent above all else.

          And to come full circle, “congestion charge” is a more honest description of what it is.

          “De-congestion charge”, while possibly more accurate has an air of “political double-speak” about it.

          We all sense (innately, if not explicitly), that the “decongestion-ness” of any “(de)congestion charge” is only as good as the amount of money collected after collection expenses (the tolling system efficiency) and the uses that that money is actually put to.

          And no one trusts that politicians will spend that money for the benefit of all. Rather we suspect it will end up as directed spending.
          There will be vested interests (say freight operators) who want their congestion charge money used to build “freight lanes” or something that directly benefits them. Or the AA wants their motorists to only funds more motorway improvements etc.

      2. We already have the perfectly good word “tolling” which is the traditional way to describe charging people to use a road. “Decongestion” sounds pretty weaselly, so I’d prefer “congestion toll”.

  8. this can be a possible solution to Auckland Transport
    a) During peak hours only cars/buses run on the Motorway. No commercial vehicles who can always work their logistics.
    b) Subsidy on Taxi ride during peak hours. This will encourage commuters to use Taxis- which are often running empty.
    c) Nominal tax on private cars during peak hours only. Subsidise Tax for private cars having a passenger- to encourage car sharing/pooling.
    These are just my thoughts.

    1. I disagree with the idea of removing commercial trips as many are hoigher values as it is hard to police.Also carpool trips obviously have a lower cost per personlarady, why subsidise cars when they could be buses.

  9. Yes, I tend to agree with Sailor here: Commercial trips are the ones that should have priority; they are much higher value than commuting to work. It’s easier to move people on a bus than pineapples.

  10. Totally agree with this post! I very much like the ideas of static and dynamic efficiency.

    With regard to the “road users have already paid” argument, I have always found this a bit odd. In addition to the very good arguments given, I can’t understand who this “road users” person is? Road users are not a unified being in any meaningful sense. For example I use the Harbour Bridge 10 times a week, but have never coughed up a cent for it (well maybe a bit for maintenance). I was never asked to buy out the shares of my recently retired neighbour, who used did actually help pay. So I can’t see how road users own it in a meaningful sense. No one would suggest that electricity users own the power companies, or that students own the universities.

Leave a Reply