There’s a lot I agree with in Stu’s post yesterday about being careful how we look to grow the public transport market and focusing on low-hanging fruit before trying to convince rich people to give up the BMW’s by building super-expensive light-rail lines everywhere. However, there’s an interesting area where I’m not sure I do agree with Stu – and that is in relation to what emphasis we should place on making public transport faster. Here’s what Stu says:

Before wrapping this up, I think it’s also worth mentioning that some aspects of this discussion are related to an earlier post on generational differences. That is, because most of our transport decision makers (including myself) fall into the 19-65 age-group there is a natural tendency for us to propose solutions that address our needs, rather than the needs of our users. This can result, for example, in a undue focus on high-speed services. For their part, PT users seem to not value speed – or more accurately “travel-time” – as much as other attributes, such as frequency, reliability, simplicity, and affordability.

There are some really important discussions and debates which fall out of this issue and come down to the fundamental reasons why people choose either one mode of transport or another. Should we focus on improving speed of service if it comes at the cost of reduced convenience of stops (such as spacing bus stops further apart)? How important are fast services compared to simplicity – like the debate over whether there should be express bus services or not? How important is increasing speed, if it comes at quite a high cost and therefore might require an increase in fares to reflect that investment (or an increase in rates or petrol taxes or foregone investment elsewhere)?

Stu’s arguments are very similar to those made by Jarrett Walker in the book Human Transit.In Human Transit Jarrett critiques much of the focus on speed on the ground that it’s generally people who mainly drive (and therefore understand the concept of improving speed) thinking that public transport works exactly the same way. Of course public transport is more complex in the sense that other issues like reliability and frequency matter a lot as well. Along with other, more difficult to quantify matters such as simplicity and ease of understanding of a PT network, quality of waiting facilities and so forth.

Perhaps what’s really key here is to focus on improving public transport speed as actually meaning improving the time it takes to get from your door to where you’re going, including wait times, including transfer times, including how long it takes you to walk to the stop and so on. In this sense, the actual speed your vehicle goes is going to have a fairly tiny influence on the speed of your entire trip (i.e. how long it takes to get from A to B). What’s going to matter a lot more are things like:

  • How frequently does the service come? (i.e. if I turn up randomly how long am I likely to have to wait)
  • How long does it take for people to board the service? (this matters a lot for buses when they’re stopping to pick up passengers all the time)
  • Does the service get stuck in traffic congestion or does it have a dedicated lane?
  • Does the service have to wait at traffic lights all the time or is there a clever pre-emptive phasing system?
  • Does the service take a straight line from A to B or does it go all over the place down every back street imaginable?

At risk of falling into the trap that Stu outlines above, it is the excruciatingly long time that public transport takes for most non-commuting trips which puts me off using it for pretty much anything other than getting to work. Even for getting to work, catching the bus is far slower than driving would be (probably at least twice the time), but as I don’t want to shell out for parking each day I catch the bus.

By contrast, in cities where public transport seems to be used for a wide variety of trips every little piece of the system seems dedicated to making your trip time as short as possible. Frequencies are high, dedicated infrastructure is provided to separate the service from congestion (whether that be bus lanes or rail infrastructure), routes are straight, traffic lights turn green when the bus/tram approach them and – yes – the services are fast. In a successful PT system the weighting given to all these competing factors (frequency vs speed, simplicity vs speed etc.) varies by the area being looked at. In inner suburbs frequency and simplicity are perhaps more important than sheer physical speed because a greater proportion of the trip is likely to be waiting for the bus/train to turn up. For longer trips speed becomes more important because you’re on the service for much longer.

I’m guessing that perhaps Stu’s position is not as different to mine as you might think – because it comes down to defining what is actually meant by “speed”. In my mind we do need to make public transport a lot faster. However the most important ways to do that in the vast majority of cases won’t be through making the vehicles travel quicker when they’re at top speed – instead it’ll be things like better frequencies, straighter bus routes, faster boarding times and the most important of all…


Share this


  1. This is really important and clearly with all of those changes and especially the last the bus network can be significantly improved but it also underlines why trams in traffic, or just more buses on existing streets, are no substitute for fixing the Rapid Transit Network: The grade Separate core of our Transit Network, rail and the busways.

  2. Good points, I really enjoyed a chapter in Sizing Up the City which talks about how PT users value their time: quite differently to car drivers (and to the people who wrote the Economic Evaluation manual). To anyone who missed it: Russell, M. (2010). Convivial Public Transport: Six Theories About Travel Time and Social Wellbeing. In: Howden Chapman, P; Stuart, K and Chapman, R. Sizing Up the City: Urban Form and Transport in New Zealand. Wellington: Steel Roberts Publishers. p.117.

  3. Dave call me a crazy optimist but I’m pretty confident that we are living under the last government that thinks there is political capital in opposing the wishes of the country’s biggest city. They will shift to arguing about which project to do first and other details but we are all but over the auto-only age. Auckland’s scale is an undeniable fact. Even if its through gritted teeth electoral calculations will be made. Then we will get news bites like the one above.

    Politicians who are slow to work this are looking sillier and sillier by the minute, eh Mr Brewer.

    1. Yes. The census that came in the mail this week will reveal further shifts, which will tilt the electoral calculus towards Auckland. They’ll redraw the electorate boundaries, and Auckland will probably gain two seats.

      The country loses population to emigration rather evenly, and gains it to immigration in a rather concentrated fashion. Add to that a long-established urban drift, and you see Auckland’s relative political power increasing for quite some time into the future.

      1. Besides starting with a random swipe at “super-expensive light-rail lines” and some supposed rich man stereotype from the 80s in his BMW (I’d go Audi if you want to update that put-down by the way), this post ends with the rather shouty outburst that we need, and I quote: “…A WHOLE HEAP MORE BUS LANES!” to make PT “a lot faster.”

        How that relates to balancing the different variables of a qualitative and quantative assessment of PT as a service delivery, which seems to be where this post was trying to go, I don’t get.

        So, first of all, where’s the evidence to support that claim that we need “…A WHOLE HEAP MORE BUS LANES!” to make PT “a lot faster.”???

        And show me a great city where PT has been optimized through the provision of a network highly geared towards buses, high speed buses and bus lanes – and I mean a city we can aspire to emulating – not Bogota. Where is that city?

        1. Agree – and not Curitiba either – that place is pretty average at best. A focus on bus lanes is basically a cop out – for mainly third world cities that are too cheap to invest in other forms of transport.

          1. Hmmm, more like important part of the transport mix, but no substitute for proper RTN. Good balance of cost effectiveness and outcome, if joined up.

            But yes Curtiba and Bogata are overstated as models for cities like AK

          2. The thing is because we’ve made a hash of AKLs PT infrastructure and need to reboot it that at least gives us one huge late-to-the-game advantage – we can leverage the best of what other cities have done and sidestep the worst.

            I’d love to see a post looking at which cities we should model – yes, we have our own unique set of challenges from the geology up – but let’s take advantage of the massive amounts of time, money and energy other places have put into researching, building and implementing their systems.

            Berlin? Seoul? Singapore? Madrid? Dresden? Who’s got the goods?

          3. Ben The Great Upgrade (as I’m calling it);
            1. electrification of existing rail network with 10min frequencies (turn up and go)
            2. Total redesign of bus network also with turn up and go frequencies
            3. Integrated fares and transfer stations
            Is going to make a huge difference and is all underway.

            But I agree what it will leave us with is a good platform from which to actually transform the city with step change projects. Starting with the CRL.

            But also #2 does need better priority for buses on our streets to really work. And this process does need to work in order to help get buy in for the really big changes.

          4. Bus lanes in cities that we’d like to emulate… how about Vancouver or Paris? They have a heap of bus lanes (which Ben seems to be getting confused with megacapacity BRT).

          5. OK, cool, so Vancouver is the best suggestion we have for a city that prioritizes bus (ie motorized) transportation in the mode mix?

            I haven’t been to Vancouver but a quick look around the ‘net reveals that it seems the locals aren’t too enamoured of their transit system – so much so that they their very own website set up – The Vancouver Public Transit blog, a little in the spirit of this one. Here’s their ‘About’ page:

            “The Vancouver Public Transit blog (VPT) is a site born from frustration with the state of public transportation in BC’s Lower Mainland. Late, missing, and overcrowded buses are the status quo. Bus drivers in the Vancouver region reported a record 19,000 pass-bys in 2006. The shortage of transit vehicles has recently been exacerbated by the breakdown of the new trolley buses.”

            Ok, looks like that page hasn’t been updated for a while, but still, makes a point.

            Vancouver City has set a target of 2/3 of all trips to be PT by 2040. Getting rid of diesel buses is part of that plan. However the problem they have locally is the City doesn’t own the transit companies so they can only influence the direction, not determine it.

    1. One of the primary benefits of bus lanes is improved reliability, rather than simply faster journeys. And that’s important because it means you don’t need to allow so much dwell time between trips, which in turn flows through into running more in-service hours across the board. Bus lanes can be congestion free without being high speed, e.g. the median bus lanes used in European cities such as Paris and Amsterdam are good examples of non-BRT bus priority.

  4. I would say that speed is definitely an issue. For example, even the express bus from Rosebank to the CBD is a 1 hour trip. At just over 12 km, that’s *running pace* for a particularly fit and motivated individual. I’ll use it in a pinch, but that’s a serious time-tax.

  5. Assuming AT get the rapid / frequent transit plan in place that will help a lot but that is a way off still. What I think it all comes down to the fact that the network has to work together which it currently does not. My personal example is the Onehunga train station and the fact that the bus stop serving this is very poor (There is no shelter just a narrow footpath) and none of the bus services line up with the trains. When I can I walk to an from the train station to Mangere Bridge as the buses usually arrive 30 plus minutes after the train. Anyway one solution would be to stop the buses that go all the way from Mangere and Otahuhu from carrying on into town an just turn around at Onehunga and get them to feed the train network. As the buses are only slowed down once they leave Onehunga and head into the CBD. I would suspect there are a lot of other bus routes that would be the same. I realise this is basically what the new regional transport plan is looking to do but I do wonder why is it going to take years to implement?

    1. But the plan to fix these issues is underway and not that far off say compared to the CRL which won’t be operating this decade, whereas within three years we will have a brand new RTN and FTN. Would be lovely today, and should have been operating for decades, but there you go; people vote funny.

  6. Speed is a factor. I came across a letter I had published in the Herald in June 2003 (I guess John Roughan wasn’t editing the letters to the editor then) where I noted that it took a bus 45 minutes to travel from the town hall to the railway shed on Beach Road during peak. Yesterday I took a City Link bus from Vulcan Lane to Pitt Street off peak and it took 20 minutes so things are really not improving. In both instances it would have been faster to walk. Without grade separated bus lanes and traffic light priority, these Link bus services can only be seen as a long-running, slow, joke.

    1. Your example suggest that bus speed has nearly doubled in the last 10 years! (vulcan-pitt is about the same as queen-beach, no?) And the city link is free to boot!

      1. Town Hall to Beach Road rail shed was 2.16km, so the speed was 2.8km/h. I recall that it was raining heavily and it was around 5:00pm (I was hoping to catch the second last western line train of the day or otherwise I’d have had to wait a further hour). Vulcan Lane to Pitt Street is 1.35km, so the speed was fractionally faster at 4km/h. This was at 3:30pm on a fine, sunny, day; because there were so many AGGS pupils boarding I alighted at Pit Street and walked to the K Rd stop faster than the bus.

    2. One of the good things about the City Link is that it isolates much of the slow/unreliable parts of the city centre (e.g. Queen Street) from the rest of the network, which in turn means that the latter can function more efficiently. So in some ways you could say that the City Link is set-up to be slow and unreliable, but in doing so it means performance improves across the rest of the network.

  7. Speed is needs to be considered as a whole sum made up of:
    Time from home/work to bus, ferry or train stop, waiting time (ie frequency and convenience – whether a timetable needs to be consulted), speed of journey and crucially, reliability.
    The reliability is obviously doubly important where connections need to be made.
    That single transferable ticket also becomes a critical part of the overall package.

    There is one other factor which I am not sure is given enough prominence – that is whether the time spent en-route can be used productively.
    This is where a speed disadvantage for PT can be off-set against the ability to use the time better than one would be able to while driving.

    Obviously rail and ferries can deliver significant advantages for commuters from distant suburbs with tables, space to spread out to work, lap-top power and so on.

    However, with wifi and a stable ride, standing passengers on “inner metro” type modes can also potentially use the time, with i-phones, i-pads and the like. Delivering a stable ride is the big factor here – buses can deliver but need large scale expenditure on a full busway to achieve this. Weaving in and out of bus-lanes which are inevitably filled with pot-holes isn’t going to deliver that stable ride.

  8. The time taken to board passengers is ridiculously long in many instances. The driver selling tickets has got to be reduced to an absolute minimum.
    Also one simple (not easy, but simple) way to speed up bus progression would be to copy NSW and make it compulsory to give way to a bus indicating its intention to leave a stop. All good drivers do this already of course (which in Auckland means that 95% of them don’t).

    1. The sale of tickets by the driver is ridiculous and another example of how NZ slavishly follows the practise in the Anglophone world, especially the UK. I have almost never been able to buy a ticket on an urban bus in Continental Europe. There are either ticket machines nearby or a newstand/kiosk where I can buy.

      We should transition to the normal European practise of plain clothes inspectors so that all the doors can open and the passengers board as quickly as possible. There can be an interim period where the driver can sell tickets but it is not his responsibility to make sure everyone has one.

      Eventually, we phase out onboard sales so the driver can concentrate on driving and also be better protected behind a screen.

      This can only happen with integrated ticketing and will require a campaign to encourage AT HOP cards. We also need to start installing ticket machines at major stops and contract dairies, cafes and other shops to sell tickets. Maybe also have a system whereby people can buy a ticket on their smart phone that then sends them a QT scan by email/via app. If queried by an inspector, they just show the code, the inspector scans it and Robert’s your father’s brother.

      I know that souds like scifi in Auckland right now but if Bucharest can get that system working with the FUBAR that is Romania, I am sure we can manage it here.

      1. Sydney moved to no ticket sales on buses (weekdays, 7am – 7pm) about two years ago. Inevitably some people “learned the hard way” and the drivers did have some discretion in the first month or two, but I think it’s working well enough.

      2. Very true – the UK system is very rubbish – just look at the recent satisfaction figures which are poor. The outsourcing they have means that the mostly multinational bus operators often find it cheaper to offer a terrible service and pay the fines etc – i.e. little motivation for being on-time or even having clean buses. Applies to the trains too.

        Is it necessary to still have onboard cash based tickets? This is the age of online payments, auto-top ups, direct debits, smart phones, and frequent service stations, dairies etc. Time for tough love. The time and fuel savings on cash handling and driver safety alone would be massive compared to the initial capital cost.

    2. I’ve found that the regular bus users are now used to forming two lines, one for the cash and one for the HOP but that people who don’t regularly take the bus (and thus won’t have correct change, have large backpacks and want to ask the driver questions) regularly block the “card lane”. I would suggest (from easiest to hardest)

      1. A bit of paint on the floor of the front entrance. Dived the doorway into 2 with a dollar sign on the left lane and a Card or hop logo on the right lane.

      2. Ask how journeys can be speed up. If AT is prepared to spend hundreds of millions to save a few minutes for car drivers why not (for an example close to my heart) encourage drivers coming from Dominion Rd to the Civic to drop passengers off in Wellesley St rather than waiting (for what feels like 5 minutes) for a complete cycle of the lights to go 100m (in the wrong way for the 80% of passengers header down Queen St).

      3. Get rid of monthly and other cards that must be shown to the driver.

      4. Simplify the HOP saving. Something simple like 50 cents per trip for 0-2 stages and $1 per trip after that. I see people who catch the bus every day and still pay cash.

      5. Please start getting more aggressive with people who drive/park in bus lanes. I see people who park every day in the same place for 5 minutes to pickup their morning coffee. In the meantime several buses have to wiggle around them in busy traffic.

      1. Simon,

        I saw today (the first time in a long time), AT clearing a parked car in the T3 lane on Remuera Road – and there was an enforcement camera just up the road to boot, proving that AT “really do care” at some level about PT users. Mind you it was taking some time to do, so the 5 minute parker may not get prevented, but again its visible enforcement thats required to make most people honest.

        But, yes change is a coming.

        Parkers in Bus lanes is my pet gripe as its makes a complete mockery of the whole Bus lane/T2/3 or whatever lanes to have people parked in them to such an extent that its better to not bother with having the lanes in the first place!

        Agree HOP card savings are not enough benefit for some people to have them stop paying cash even for daily use. 10% is nothing, when AT (and NZ Bus i nthe AT HOP card case) are sitting on the float on those cards earning interest (or more to the point/more likely – reducing their 15%+ interest overdraft quite substantially)!

    3. buses pulling out of a bus stop have right of way in Italy as well. I don’t understand why in New Zealand we always have to try to do things differently when they’re proven to work somewhere else

  9. Or 1/2 hour by bicycle along the cycle path.

    I would not want to live further than cycling distance from work (18 minutes for me). Apart from the exercise, commute time is a waste of my life and I have better things to be doing.

    If a transfer between services is involved, frequency is twice as important.

  10. Transport planners should include focus on providing express services from outer suburbs that at peak times ensure PT beats private car.

    Should every EMU rail service stop at every station for example? This doesn’t happen overseas. Getting AMETI Pakuranga/Howick/Botany rapid bus transit sorted will be a great start as would a NW motorway dedicated busway initiative.

  11. Stu’s original post said nothing about the impact of speed on operating costs. In Melbourne the “Think Tram” project that has provided pre-emptive green lights for trams at many intersections was promoted to government as a way of reducing the need for additional trams. Because tram journey time was increasing through traffic congestion, more trams would be required to maintain the service frequency. When government was offered the choice of purchasing about an extra 20% more trams at $16m each, or a traffic light priority project, they chose the latter.

    For Auckland, the best way to get good value in bus contracts is to fix the bus priority and off-board fare purchase first. AT could do a cost-benefit analysis of these priority measures, and how much it could reduce bus operating costs. Such measures might be necessary to achieve the government’s target of fares as a proportion of costs. Perhaps when ratepayers see the benefits of lower subsidy levels or higher service levels, they may be more supportive.

    1. +10. Let’s get on with the cheap low hanging fruit, as the big hits are being planned/negotiated…. should do them anyhow.

      I guess the time to really incentivise people off cash fares is when we have the one real HOP card [to rule them all] in operation.

    2. Good point – and speed is important to operating costs. I was commenting purely on what matters from a users perspective, although the two are obviously related insofar as if you can increase operating speeds and reduce operating costs, you can then free up resources to improve service in other ways, e.g. reducing fares.

  12. “because it comes down to defining what is actually meant by “speed”. In my mind we do need to make public transport a lot faster. However the most important ways to do that in the vast majority of cases won’t be through making the vehicles travel quicker when they’re at top speed – instead it’ll be things like better frequencies, straighter bus routes, faster boarding times and the most important of all…”

    To create an aviation or nautical analogy:

    You and Stu are using similar terms loosely and then end up comparing “Speed over Ground” (SOG) and “Velocity Made Good” (VMG).

    The difference between these two?
    Well, SOG is how fast you are travelling that instant – relative to the ground underneath you. i.e. the bus going 50 Kph or 40Kph etc

    VMG is how much progress you’re making towards your **destination**
    i.e. how long until I get to where I need to be (not just to the bus stop where you get off the bus or train – but the ultimate destination – your home, the shops or work etc where you actually are going on this journey).

    So, what Mr A says, and I agree with him, is that PT needs more VMG and not just more SOG,
    Or to put it another way, using the old adage “more haste and less speed.”

    If you increase SOG but not VMG then all usually do is to make PT users “hurry up and wait”.

    And how do you get more Haste? Pretty much everything Mr A listed.

    But the point is by interlocking your modes and reducing wait times between mode changes (as well all the stops along the way)
    – you increase the VMG on your journey, without necessarily needed to increase SOG.

    To be sure there is a trade off between optimising these two, but with careful design of the PT system as a whole you can ensure superior VMG for most journeys, while only moderately increasing SOG.

    And once everyone realises that the VMG obtained by using PT (once its grade seperated ala bus ways or railways) is better than the alternatives (e.g. car on surface streets), then they’ll shift to it in droves – even if the total distance covered is longer by PT, if the VMG is better than the alternatives then why wouldn;t you use PT – most other things being equal?

  13. i suggest rather than outright speed, the issue for many people is is the presence or absense of delay, a lot of people live in the now and being stopped rather than moving is a curse

    the prime example of this is the driver who pushes past a cyclist at the most awkward point only to stop at the next set of lights and be overtaken by the cyclist

    so speeding up boarding, bus priority and signal pre-emption all can contribute to maintaining momentum and the feeling of making progress, saving a couple of minutes isn’t really going to motivate people to catch PT, but maintaing movement can reduce frstration

    1. Speed and apparent speed are both important. I cycle to uni at least in part because the bus is too slow… meandering, infrequent, and uncomfortable. I do use the bus when cycling is not appropriate [rain, don’t want to be sweaty] but really I have a very low level of bus love. And speed improvements would help a lot, as would modern especially electric buses. But then I’d really like them when riding and walking!

    2. A Wellington, train, example, if I may… Trains follow SH2 between Ngauranga and Petone – about 5km. What really annoys is the 70km/h speed limit, and temporary speed restrictions which put the train to a crawl for some patches.
      At 70, the train is still going slower than rush-hour traffic most days and the temporary speed restrictions just make it look retarded – a train with the exclusive use of it’s corridor, and it still ain’t that quick.
      While speeding the train up to 120 won’t do a lot for the journey times, I feel it would project a really good image for PT to those in cars – one that trumps everything the car has, and makes the train commuters feel like their choice is justified.
      I agree that it’s not outright speed that we should be aiming for all the time, but there are some cases where literal higher speeds could be used to grow PT.

Leave a Reply