Todd Littman of the Victoria Transport Policy Institute has put together a really useful critical analysis of the way we measure congestion and how this affects the general approaches taken to reducing congestion – typically seen (especially when undertaking cost-benefit analyses of transport projects) as the main justification for transport expenditure. The article is a fairly easy read, if somewhat lengthy, so I won’t go through all the details. Here’s the abstract: To keep this post from getting too technical and detailed, I’ll largely focus on the pictures and graphs that are in the article. This is a particularly relevant graph, because it means that unless our transport demand modelling has been updated to take account of the pretty radical change to traffic volumes (this is the USA, but we’ve seen similar trends in New Zealand) in the past few years, chances are we’re going to be enormously over-estimating traffic volumes (and therefore congestion) in the future. There is a real danger of failure to see change, even after it has happened, by people who have spent their careers under one orthodoxy. And I believe that is what is happening now at the MoT, if the reports I hear are true, of officials there arguing that there is no point in changing policy just because of recent trends.  Well how long do ‘recent trends’ have to continue before they become the relevant? 5 years? 10 years?: Some of the dip is obviously due to the economic situation over the past few years, but other aspects might be more structural – relating to demographic change, changing land-use patterns, significantly higher fuel costs and general cultural change. Littman’s article notes that, should this down-turn in traffic volumes prove to be long-lasting and structural, it really changes the way we approach the issue of congestion in the future:

It made sense to invest significant resources in roadway when the basic roadway system was first developed and automobile travel demand was growing rapidly. During that period highway projects provided high economic returns, consumers reaped large benefits, and there is little risk of overbuilding roadway capacity since it would eventually fill. But once the road system matures, so there are high-speed highways connecting regions and a well-developed network of paved local roads, the marginal benefits of incremental roadway expansion tend to decline.

Transport planning and financing practices will need to change in response to reduced growth in vehicle travel demand and congestion problems, and increasing demand for travel by alternative modes. This will require reducing emphasis on congestion problems and roadway expansion and increasing emphasis on other planning objectives and other types of transport system improvements. 

Most of our long-term thinking is still based on the assumption that traffic rates will grow and grow. To an extent, in a place like Auckland where the population is growing quickly, that is likely to be somewhat true – but it seems reasonable to expect that per capita vehicle travel may well have peaked, meaning a slower rate of traffic growth in the future. This is a good thing, as it means we can hopefully get away from an endless game of adding capacity to the road system only to see it fill up over and over again, hoping that perhaps this one last time the extra lane might solve congestion for good (like a drug addict wanting just one last hit?).

Of course, in the real world (rather than the fantasy world of Ministry of Transport policy analysts and NZTA traffic engineers) we understand that  building more roads simply leads to inducing more traffic – meaning that more often than not you find yourself back to square one rather surprisingly quickly (ever noticed that the widest sections of Auckland’s motorway network often also seem to be the most congested?). This is well illustrated in the diagram below: Now it’s probably a bit unfair to say that there’s no benefit from allowing more vehicles to travel at the time they want to travel (as induced demand provides for) even if it means the road is still congested. But what Littman’s research shows is that there are diminishing returns when it comes to the value added – in that a lot of the people “induced” to using the road at peak times wouldn’t probably be quite willing to travel at other times, or via other modes, to avoid the peak congestion. The article goes on in quite a lot of detail about how we might better focus on reducing the impact of congestion – through building public transport, pricing roads and undertaking smarter land-use policies. There’s enough detail in those to probably fill a future post, but there’s one last graph that certainly caught my attention, because it looks at the wide variety of different costs – related to travel – that individual people face: What this graph shows really cleverly is that, in the broader scheme of everything, traffic congestion doesn’t really cost each of us that much money. Sure, it’s an annoyance, but then so is losing $2,000 a year in depreciation on our vehicles, having a cost to society of almost $2,000 a year from crash damages, paying (through higher rent, prices or lower salaries) parking subsidies and so forth. Yet for some reason our transport policy seems to be utterly obsessed with minmising the cost to us of what’s really quite a minor matter – only the sixth most important issue in terms of cost magnitude.

So automatic, so deep is the group-think at the MoT and NZTA, not to mention their close buddies in the road lobbies like the Road Users Forum and the AA, that the only possible answer to all questions of  movement is to expand the road space, that these institutions are spending uncritically on projects with ever decreasing returns on the investment. Of course the pressure from their political masters is currently very great too and it all adds up to a deeply unsophisticated and wasteful approach to infrastructure investment and a less efficient and poorer city.

Of course, for some road users (freight movement and business travel comes to mind) congestion and unreliability have a much greater impact. But there really is something incredibly stupid and naive about over building one system that forces everyone, no matter how reluctant, into a car and in the way of these vital road users. Under this ideology it’s almost like the MoT and NZTA are really in the congestion business. Where is the analysis that shows that getting lower value trips onto public transport through investment in providing quality services wouldn’t be more cost effective than trying to squeeze ever more traffic onto already clogged roads? Especially because, as this study shows, congestion isn’t even the biggest burden of auto-dependency.

Share this


  1. I thought Julie Anne Genter’s speech a few weeks back in parliament picked up on quite a few of these issues really well. At least we have some hope that there’s someone in parliament who “gets it”. I wonder how influential she will be over the course of this parliament in changing mindsets on transport issues?

  2. Just while we’re talking about numbers;-

    The professionals in this country who work up number estimates for the future have been doing an astonishingly poor job lately, what with Treasury being so rosy 5 seconds before the election and so gloom ridden straight after, with Bill English admitting he has no idea how much we’ll get for selling our state assets, ATEED’s guesses on RWC crowds, Auckland Transports- the same- and it’s got me wondering-

    Will Auckland actually grow as fast as predicted?

    How do we know these estimates are any better than the random “dartboard monkeys” efforts listed above?

    They could all be done by the same people for all we know. Anyone got a headsup?

  3. If the tailing off of traffic growth does turn out to be a long term trend, I predict it will take about 20 years for the road planners who’ve built their careers on one orthodoxy to admit it.

    It’s been almost 40 years since the first oil crisis (1973) brought wide attention to the problems of cars in big cities. You wouldn’t know it, would you?

    As has been said in another context, ‘the dominant paradigm in science doesn’t change because scientists change their minds. It changes because old scientists die and young ones think differently.’

    1. Well if it does take 20 years, we’re at least 5 into it already!
      I do fantasise that because the engineers at the top of the motorway industry are clearly imagination free quants one day the numbers may break through to them in ways that other more sophisticated forms of reasoning don’t. Dreaming?
      But also, to be fair, any outbreak of sense at the Ministry would surely be slapped down (fired) by their extraordinarily small minded and backward looking political masters, so really the best hope is change at the ballot box. And that, gentle readers, can happen quite suddenly. And change there will mean big change in transport as the two major parties now have widely different policies unlike for the last 50 years. So onward!

    2. I predict it will be when fuel prices really start to bite, and roads become noticeable emptier as a long term trend.

      On that note – I’m puzzled that Todd’s last graph, helpful though it is, does not appear to identify fuel costs explicitly. I can’t see that it is absorbed in “vehicle ownership” or “operation” entirely, and he has a specific item for “fuel externalities”. I’m guessing the average NZ urban consumption figure is between $1200 and $5000 (allowing for very different urban locations & vehicles), based on and general observation. Any predictions what annual cost will be the tipping point?

      I also heard mention once of an official statistic recording VKT or fuel purchases recorded at Meshblock or Area unit level. Does anyone know if / where this exists?

      1. Yes I was puzzled by that figure too. Seems too low.
        But also, ironically, the coming fuel price squeeze will go along way to reduce congestion by simply pricing the poor off the roads. We know this because it happened in 2008 when there was a sudden 10-15% drop in vehicle numbers, since recovered, probably through the expedient of buying petrol to get to work instead of breakfast for the kids.
        Not, of course, a bad outcome for those able to afford higher petrol prices easily [now who could that be?], who are also the same group able to afford newer more fuel efficient vehicles, so transport policy is very much a social equity issue.
        Amazing how determined the government is then to ensure that this group has no alternative to driving, now even attacking the slim ability that local councils have to pick up the slack in this area.
        In terms of tipping point it very much depends on how suddenly it occurs. A ‘shock’ being more visible. But the slow and insidious upward creep of the costs of auto-dependency help disguise its effects as other social problems. Parents are blamed for the problems faced by their children, housing unaffordibility is carefully discussed separately from transport poverty and so on.

        1. “We know this because it happened in 2008 when there was a sudden 10-15% drop in vehicle numbers, since recovered”

          The fuel price today is the same as it was in 2008, but the decline in vehicle use hasn’t eventuated this time. I would suggest that people are becoming more accepting of higher fuel prices, and reducing their expenditure elsewhere, so that they can keep driving. In 2008 is was a shock. In 2012 people are planning better.

          1. Ah, Geoff, yes they have adjusted because what choice do they have? One thing people are clearly trying to do is to move into the inner suburbs as reflected in property and rental prices, but that can’t go on [finite supply].

            And does this mean they can go on adjusting?; I don’t think so, but I do think we are going to find out.

  4. You really need to finish the electrification project. The Perth experience should be repeated, admittedly from a higher base. When not just your piddly little diesel trains are full, but six car electrics running into Britomart from all lines at peak maximum capacity, it should become obvious, as it has over here, that something has to change.

    Not that motorway construction ceases; just that rail gets a better looking, and ideally the bus enthusiasts get away from wanting to run trunk bus routes that compete with rail and focus on feeding rail better.

    1. Electrification is underway and yes the hopeless deadend at the heart of the system will remain obvious to anyone who cares to look even when it is complete. The argument around the continuation of motorway building is both that we cannot afford to build both a new transit system and keep building motorways [unlike in WA, possibly], but also that each new bit of motorway, on an already lavish system, further embeds us in suffocating auto-dependency.

  5. Thanks Patrick

    Question: lots of talk on this blog and elsewhere about an Airport/new southern line but one question I have is how do these proposals boost central city capacity (or take it away).

    I would suggest (again drawing from Perth) it might be helpful to build a whole new line from airport to city. Ignore the ‘loop’ aspect or any of that stuff. Just another corridor to get to the central city. You could substantially reuse your existing south line for some of it but really make sure it is fit for purpose as a new corridor.

    Notwithstanding the excitement over having fixed the Onehunga branch, it doesn’t look suitable to be part of a new high capacity corridor.

    1. Well, I would say the lack of will from central government to spend on a 1.5km rail link would indicate there will be few funds available for a brand new line.

      Even so, I’m not sure where else you could have a rail corridor from Airport to CBD. The natural line between Onehunga and the airport goes through the main residential areas of that part of Auckland, does it not? So its seems logical to pass through there than run through somewhere else. My thoughts, anyway.

      1. @KLK, the most effective airport line would be one from Wiri. This will enable the widest possible catchment, with trains able to run from Britomart via both Newmarket and GI, and from Papakura, to the airport. A line from Onehunga to the airport, will be quite ineffective as an airport rail service. Furthermore, a line from Wiri is easier and cheaper to build. It is also the favoured option by both KiwiRail and Auckland Transport, largely for the reasons I mention. You get much more bang for your buck with the Wiri option.

        The Onehunga-airport line should be built later, as part of an Avondale-Airport route, connected with Onehunga along the way. You’ll get more bang for your buck with that project as well.

        Building Onehunga-Airport first is actually the least effective option of all, so I find it odd that it is the preferred option for some. They really haven’t thought it through.

        1. That is the case if you are only concerned with Airport to City trips and have no interest in providing services throughout Mangere. A suburb with very poor connection to the rest of the city and also very badly cut up by motorways. In my view it would be a bad mistake to only build another terminating spur line here and not complete the link between Onehunga and Manukau.

          1. Yeah, I think the connection from Onehunga had 3-4 new stations servicing the residential areas?

            Looking at a map, between Wiri and the airport, there doesn’t seem to bemuch of a need for any stations, for a line that would be in any case 2/3rds the length of the Onehunga link in any case.

            Even taking into account the extra distance, seems more bang for your buck with Onehunga (and thats before considering what looks like a problem for a Wiri link in getting across the water on Puhunui Rd. Mangere bridge was future-proofed for that).

          2. KLK I think the thinking is to just get to the airport as cheaply as possible, and stations are all add to the expense and the south route is also short and therefore cheaper. But you would hope that there is no longer any doubt about rail uptake so do we still have to do things so meanly? Just thinking that this is a bout the airport is a mistake in my view, anyway how about the service pattern with three terminating trunk lines off the Southern Line? Frequency for each terminus is going to be poor and/or the spine will be over served.

          3. I guess you can always stage the Airport Loop. Build the eastern connection to Puhinui first and the bit through Onehunga later. Or vice versa.

          4. Yes take a leaf out of the motorway builders’ book, build a little then lobby like mad to ‘complete’ the line or network, still a new spur from Puhinui would be more problematic service pattern wise than extending the Onehunga spur…. main thing now is to protect both routes.

            The other important issue is how well designed the line is, and especially how well integrated the airport station or stations are with the terminii, especially important now as there are plans to move or rebuild the domestic terminus.

          5. “That is the case if you are only concerned with Airport to City trips and have no interest in providing services throughout Mangere”

            It’s the Onehunga-Airport route that is only concerned with CBD-Airport rail travellers. The Wiri-Airport route is the one that will create a viable airport rail service, by enabling trains from Newmarket, GI and Papakura routes to get to the airport.

            Providing services to Mangere comes at the expense of a viable airport service. So it comes down to whether or not you want a new suburban line for Mangere residents, or an airport line relevant to the rail network as a whole. Since it is a the region that wants airport rail, not just Mangere residents, then Wiri is the responsible way to go.

            Put it this way – if you build Onehunga-Airport, you may have tracks to the airport, but you won’t have the airport rail service that the region wants.

          6. Fair point, Geoff.

            The ironic thing is that, as a link from Wiri is just linking the airport to the network, on the face if it it supports the argument from opponents that “we don’t have the travelling numbers to support it”. This despite, as you say, providing access to a greater part of the wider rail network.

            The Onehunga link has, again, on the face of it, the stronger argument for the link as it would provide a RTN for a PT starved part of Auckland. yet it probably does not have the regional advantages the Wiri link does

    2. Ricardo there is plenty of unused capacity on the existing rail network once two obstacles are overcome. One is the need for higher capacity rolling stock and that is funded and ordered and due to be delivered from next year on. The other is the deadend at the Britomart Terminus. Happily Auckland Transport are proceeding with the solution for that too, the City Rail Link, with Council funding up to actually digging the tunnels. There is time for a change to a less brain dead government in Wellington or for this one to begin to see sense on this matter so there is good chance that around the end of this decade we will see this condition being met too.

      There is no capacity issue on the Southern Line except at Britomart so once that problem is addressed it would be crazy and crazy expensive to duplicate it as there is no need. In fact an important argument for the South Western Line is the fact that it is mostly already there. Yes the Onehunga line will need to be double tracked, we love to build things cheaply then do it all over again later, but the main thing is that the ROW is there.

      It is also important to note that the this line is not just about the airport and even nor will the airport terminals be just about travelers. The airport is a centre of employment and the surrounding suburbs home to workers for the whole region, and it is an area that suffers a great deal of severance from the existing transport infrastructure that it hosts, a passenger line through here that links to the city centre, the regional hub of Manukau City, and the rest of the network. Designed well this line will offer a cost effective expansion to the network and be of great value to the city as a whole.

      1. The Southern Line isn’t quite free of capacity issues. Remember, in addition to squeezing passenger trains through at five minute frequencies, there are also some 32 to 36 freight trains per day to get through, many of which run through the day and some during peak hours. This is why the third main is being built, although it’s entirely a KiwiRail project at this stage, meaning the third track won’t be available for AT trains to use, and won’t be electrified. To my way of thinking, this is a mistake. It should be a KR/AT joint venture, so that in addition to freights, express EMU services can also use the third track to overtake all-stop services.

        From next month, Wiri is going to become a problem point, albiet a minor one at this stage, on the Southern Line, as Manukau services entering and leaving the Southern line will cause Southern line services to stop either side of the junction. Expect some delays here from April 16th.

  6. There’s been a fair bit of inflation since 2008 though, so I would say that current prices aren’t (in real terms) up to 2008 prices.

    As Patrick says, it’s more the change in price that influences behaviour change (i.e. shift to PT) than the actual price. Plus, traffic on state highways declined in 9 of the 12 months last year – even in a recovering economy, suggesting a link with higher petrol prices.

    1. Sorry that post was in response to Geoff’s above which said:

      “The fuel price today is the same as it was in 2008, but the decline in vehicle use hasn’t eventuated this time. I would suggest that people are becoming more accepting of higher fuel prices, and reducing their expenditure elsewhere, so that they can keep driving. In 2008 is was a shock. In 2012 people are planning better.”

Leave a Reply