An excellent recent article on CityLab discusses the important concept of ‘induced demand’ – the phenomenon where additional road capacity encourages more people to drive and fills the road back up again – as well as asking the important question of why is so much money still being spent on road widening when we know it doesn’t work.

For those not familiar with induced demand, it is an incredibly important concept in transport planning, and has been known about for many decades:

Economist Anthony Downs is often credited with first articulating this “iron law of congestion” in 1962, as construction crews were hacking interstates through American cities. Downs published a seminal paper with a stark warning: “On urban commuter expressways, peak-hour traffic congestion rises to meet maximum capacity.” In other words, adding lanes won’t cure snarled traffic; the additional car space inevitably invites more trips, until gridlock is as bad as ever.

Not only has induced demand been known about for a long time, it has also been measured across many countries again and again, highlighting the futility of trying to solve congestion by building wider roads:

Downs’ iron law applies not only to U.S. cities, which have grown more traffic-jammed despite billions of dollars in fresh pavement, but also to those around the world. Highway expansions in Norway and Britain haven’t reduced congestion there, either. The principle now meets little opposition among economists and urban planners. “It’s widely accepted,” says John Caskey, who teaches induced demand as part of his urban economics course at Swarthmore College. “For economists interested in urban transportation, there isn’t really any debate.”

And yet, billions upon billions of dollars continues to be spent on road widening in the hope that the next road might actually – finally – achieve the desired goal of reducing congestion. Despite somewhat better rhetoric than most recent governments, the current government’s transport plans are still very tilted towards road building rather than public transport, walking and cycling:

Beyond what’s in the NLTP shown above, the government is also ploughing billions from general taxation into the NZ Upgrade Programme, which is mostly funding new or wider roads like Penlink, Papakura-Drury, Tauranga Northern Link and Otaki to North of Levin.

Yet to many even this vast expenditure on more and wider roads is not enough, despite decades of proof that more and wider roads won’t actually help fix congestion. For example, a recent newsletter from the AA pushes hard for even more spending on this failed strategy:

It is clear that a significant rebalancing of the transport programme is required, in the form of a much greater focus on improving conditions for general traffic. That would entail scaling up and bringing forward:

  • New road projects on the periphery of the city (in particular, around high-growth areas)
  • Targeted widening on sections of the motorway network
  • Small adjustments (such as dynamic traffic lights, peak-period clearways and re-configuring lanes at chokepoints) right across the network

Reinstating Mill Road, and Stage Two of the Papakura to Drury South motorway widening are essential.

There is no logic to continuing to spend billions of dollars on solutions that have been proven over and over again to not work. So why on earth does this still happen? The article suggests a number of reasons.

Firstly, there is very little accountability in place to actually properly assess whether past investments actually achieved what they set out to do, and even fewer ways that these lessons get applied to future projects.

Part of the problem is a lack of accountability, says Beth Osborne, director of the advocacy group Transportation for America. The federal government doesn’t penalize states for getting their congestion mitigation estimates wrong, as TxDOT did in spectacular fashion with Houston’s Katy Freeway, widened in 2011 to as many as 26 lanes at a cost of $2.8 billion, half of which came from federal funds. The next year, an article in the Houston Chronicle declared the project a success: “What was once a daylong traffic jam is now for the most part smooth sailing.”

But by 2014, most peak-hour commutes on the Katy took even longer than they had before the expansion.

Federal inaction “allows TxDOT to find no induced demand in their project,” Osborne says, “even though their last 25 projects didn’t work very well because of it.” TxDOT’s website currently forecasts travel time savings of up to 84% from its $7 billion proposal to expand I-45 in Houston, which many local residents and officials oppose. (TxDOT did not respond to a request for comment.)

Matt Hardy, the program director for planning and performance management at the American Association of State Highway and Transportation Officials (AASHTO), agrees on the need to look at historical outcomes. “We don’t go back very much and see if the project provided the benefits that were expected,” he says. “I wish we did.”

The Katy Freeway in Houston

Over the past 20 years huge parts of Auckland’s motorway network have been widened – most of the Northern Motorway north of Akoranga, almost all the Northwestern Motorway, the Southern Motorway through spaghetti junction and then from Otahuhu down to Papakura (soon to be Drury). Yet before Covid-19 these motorways were as jammed up as ever. Even massive projects like the Western Ring Route and the Waterview Connection, delivered only a relatively short amount of relief and again were starting to fill up again before the most recent lockdowns. But there’s no accountability for this, and barely even any monitoring of how successful (or otherwise) projects have been. Waka Kotahi stopped even bothering to undertake post-implementation reviews in August last year.

Secondly, there are a lot of vested interests in wanting to ignore induced demand. In New Zealand, there is a tight relationship between the motor vehicle and freight lobbies and transport policymakers, which undermines evidence-informed public policy.

Whole organisations – state departments of transport in the USA, Waka Kotahi and to a lesser extent Auckland Transport and other councils in NZ – have a huge interest in pushing to continue to build big roads that encourage driving, which generates more revenue that can help fund the next generation of projects.

Gas taxes further distort decision-making, because collected revenues often go straight to state departments of transportation — just as they did 90 years ago. By implication, an initiative that reduces driving (such as a transit expansion) could result in a lower state DOT budget, while one that induces more driving could lead to a bigger one — even if thicker traffic means that drivers themselves don’t benefit.

Eric Sundquist advised state DOTs on strategic planning for a decade before recently becoming an advisor to Caltrans, the state transportation department of California. “I remember working with one Midwestern state DOT director who showed me his dashboard, which included a measure for vehicle miles traveled,” Sundquist says. “The state’s goal was clearly more VMT. When I asked why, I was told that more VMT would provide more revenue to maintain the state system.”

Such an attitude does not surprise Jose Gomez-Ibanez, a professor emeritus at the Harvard Kennedy School, who has studied transportation funding for decades. “States get rewarded for ignoring induced demand,” he says. “There’s a lot of self-interest involved.”

Waka Kotahi’s revenue mostly comes from fuel taxes and road user charges, which literally means that reducing travel and fossil fuel use – essential to tackling climate change – undermines that funding stream. Even if there are many well-meaning people in Waka Kotahi, this is a really bad incentive structure for an organisation that needs to play a critical role in transforming our transport system over the coming years.

Thirdly, and perhaps most importantly, induced demand is a bit counter-intuitive. It also takes a while to kick in, which means that the big new project appears to have been successful initially. There are also lots of hypothetical ‘counterfactuals’ about what might have happened without the project which are never truly possible to know.

“The whole notion of induced demand is a little counterintuitive,” says Sundquist. “Highway building seems like a straightforward thing to do. And it does make cars go faster — for a little while.”

Caskey agrees. “I understand that if someone who drives everywhere sees traffic congestion, their immediate solution is to keep paving things. It’s sort of a natural reaction.”

That makes the topic of induced demand a challenging one for policymakers. “Everyone with a driver’s license thinks they understand transportation,” says Rosenberg. Meanwhile, politically potent contractors, unions and automakers have an incentive to keep the public confused.

There are some positive signs in recent times that key decision-makers understand how induced demand makes building new and wider roads to fix congestion futile. The balance of investment is slowly but surely shifting towards focusing more on projects that actually reduce vehicle travel demand through mode shift, and we have actually seen major roading projects like Mill Road and East West Link cancelled without much of a public outcry. Yet in other ways it still seems we have quite a lot way to go – especially when it comes to the more ad hoc transport decision-making governments appear to increasingly be making, like the NZ Upgrade Programme. Despite decades of evidence to the contrary from around the world, it seems that many people still think that building more and wider roads will actually solve congestion.

Share this


  1. Most infrastructure projects become political due to the scale of investment, and induced demand is a medium to longer term effect to the entire transport network, rather than purely project based.

    As a result our politicians are after the short term relief gains that align with their political terms, as solving the actual problems involve greater investment into alternative modes of transport.

    Unfortunately organisations and consultants have been trained into the short term lowest cost outcomes over years of political direction into an engineering industry that does have a bit of an old boys attitude to it.

    In saying that the public perception is slowly changing and the demand for alternative transport options are becoming louder. As a result I would expect wider transport network induced demand to take more of a focus into transport projects BCRs going forward.

    1. Individuals within the organisations and consultancies have decades of evidence that should be shaping their advice and practices. That this is not happening is a collective failure of ethics.

    1. Reminds me of the “being moderate for the sake of moderate” political mentality.

      Even with NZTA’s blatant car bias and manipulation of their reports and findings, the government and media still appears to be making decisions on the assumption that the middle ground between two extremes will be the “best decision”. E.g. I have American friends who think Biden and the establishment Democrats compromising with Republicans somehow means “more gets done” than if they insisted on pushing through progressive policies – even though that gives the Republicans too much leverage & results in poorer policy outcomes.

      The NZ left is (somehow still) associated with favouring public transport & active modes. The NZ right is associated with favouring roads & cars. Compromise “moderate” philosophy therefore dictates that we need to build both in equal measure, despite the abounding evidence against new roads (and for reducing space for cars in cities).

    2. Yes, that’s spot on, John.

      And that is an outrageous ad on in many ways, which shows there’s no genuine intent to change the system through the governance tools available.

      1. The ad for head of NZTA requires someone who’s in the current system, but understanding of climate change and other environmental issues is only in the list of other optional attributes. Should there be a campaign directed at Michael Wood to get that changed before the appointment is made? What would it take to succeed?

  2. None of the above. Induced demand is ignored for two simply reasons. 1/ Roads are funded based on travel time benefits. Sure safety benefits get included and so do vehicle operating costs but travel time benefits make up around 80% of the benefits that get claimed. So this leads to the building of an assessment framework designed to find travel time benefits.
    2/ New trips get valued at half the benefits of existing. That is because the benefit is the area under a compensated demand curve and nobody seems to have gotten past straight line assumptions.
    The net result of 1 and 2 is that people assume a fixed demand a lot more than they ever should have (if ever they should have). It doesn’t mean new roads don’t have benefits as the new trips are because people are gaining utility. But it means the benefits claimed are not the real benefits.
    The whole thing is a kind of demonstration of the MacNamara principle where you only value the thing you can count. Dumb really but that is how it is.

    1. “2/ New trips get valued at half the benefits of existing. That is because the benefit is the area under a compensated demand curve and nobody seems to have gotten past straight line assumptions.”


      This alone is quite disadvantageous to high growth modes like biking and PT. And is also crazy. This assumption that existing users are inherently more important than new users is insane to me. Just because its less clear who will be benefiting and they can’t be as loud, doesn’t mean they’re less important trips or people. Doubly so if the old users are disadvantaged for improving the lives of vastly more trips / users.

      1/ Roads are funded based on travel time benefits. Sure safety benefits get included and so do vehicle operating costs but travel time benefits make up around 80% of the benefits that get claimed. So this leads to the building of an assessment framework designed to find travel time benefits.
      Imagine how much safety we could buy without having so much capex tied up in improving travel times. Median wire ropes could become almost ubiquitous across the state highway network.

      1. The half value is not a value judgment, it is an economic theory issue. A consumer surplus is the area change under the demand curve. If the demand line is linear then the consumer surplus is a triangle and the area is there 1/2 x p x q where p and q are price and quantity. The existing area is the rectangle p x q. Benefits are like a consumer surplus but only for wealth effects and with the substitution part removed. (a Hicksian demand curve).

        But the real point is the funding allocation method results in an assessment method. If infrastructure was funded based on predicted cow pats then someone would come up with a method that predicts a hell of a lot of cow pats. Funding based on time is the problem. The real benefit of travel is the things you can do or money you can earn when you get there.

      1. Actually it works well. I’m going to start using the term. Interpretation of its exact meaning might depend on tone, and accompanying level of nose flare or eye roll.

  3. “Induced demand is ignored for two simply reasons.”

    The third reason is that Joe Public don’t understand it so it is impossible to get buy in from them

    1. But why does Joe Public not understand it?

      If the government started giving away hamburgers, set up a hamburger restaurant in every town in the country, hamburgers were free (but paid for out of taxation), then we’d expect a high demand for hamburgers to develop quickly, wouldn’t we? And demand for other types of food would decline. That seems easy to understand, surely?

      1. But if the government had been giving away hamburgers for decades and everyone was eating pretty much only hamburgers and the government had been tearing out all the other farmland and food processing facilities and food stall to try to make and distribute more hamburgers to the point where most other food was really difficult to find and they made rules that required every house to have a hamburger storage shed and businesses had to set aside an acre for grazing land but STILL people had to wait for an hour just to get a hamburger because there simply weren’t enough, one could be forgiven for thinking that we wouldn’t have to wait so long if only someone would make more hamburgers.

    2. I think many in the public understand the general gist of it but are deliberately excluded by the technical-rational processes that dominate the sector. There’s limited public participation in modelling processes, and this allows the select few experts to reject criticism and refuse to allow improvements for how the processes can be improved. Decision-makers have little understanding of the methodological limitations inherent in transport modelling advice because that’s the way the modellers like it.

      1. Exhibit A – Miffy’s technical explanation above.
        However Miffy’s cowpat example is both much easier to understand and just as true. This is the type of comms transport planning needs to get through to Joe Bloggs.

  4. especially when it comes to the more ad hoc transport decision-making governments appear to increasingly be making, like the NZ Upgrade Programme

    You know I was scoffing at those criticizing elected officials, saying its no place for them to ‘pick winners’. I was thinking, of course its their place to pick winners, they’re the elected leaders, their job is to make decisions. But with MOT and the treasury saying building Otaki to North of Levin was unrecommended and the govt just blasting on ahead with it, I’m not so sure. When you have a 1.5B project on your hands for a BCR of <0.2 and the government is furiously nodding its head, surely you have a bit of a re-think. The opportunity cost is just so high.

  5. The new argument is that induced demand exists but it’s a good thing because more people and goods are going where they need to go.

    1. The counter arguments are, yes,
      but remove any travel time savings from your calculations, see what the they say the BCR is now.
      and if we’re going to induce demand we might as well do it on a mode that doesn’t have so many ill side effects.

  6. Your graph at the top indicates that Coastal Shipping is extremely poorly financed. Seeing as new ultra-mega-container ships are now on their way to NZ and can’t afford to stop more than once (if at all) in NZ, clearly Coastal Shipping is an area poised for large amount of growth. Can we get the iron law of induced demand to help here? If the NZTA was to purchase say, 5 new ships, to get the coastal shipping industry a big boost, would’t that have the effect of getting massive amounts of trucks off the roads and thereby lessening the need for more roads?

    1. I don’t think they should bother with induced demand rationale but we definitely need coastal shipping to be in our control, designed to suit our goals for labour conditions, coastal environment, road safety and resilience. Deciding to rely on international shipping to provide the local legs was one of the most stupid transport decisions NZ has made, which is saying quite a bit.

    2. Recently we have seen new railheads being established at Massey road and Wiri. Hopefully this has relieved congestion at Southdown. Inland ports can be used to shift congestion away from city areas we see that with the recently announced relocation of the freight hub from the Ashburton to Fairton. Coordinated container movement to and from inland ports to import/export ports via rail and coastal shipping is what we should be striving for. If the price is significantly cheaper than road transport its probably what will happen. Maybe the Govt has to chip in to get a home grown coastal shipping established if the trend of bigger ships servicing fewer ports continues. Is coastal shipping is be used to transport freight other than import/export within New Zealand as well. Its very hard to tell because when you see a truck or a train drive past with a container on it you can’t tell what’s in it or where its going to. In fact you can’t even tell if the container is empty or full. I do see some identifiable containers of materials like cement, LPG and liquefied gases like oxygen, nitrogen and CO2 being loaded onto coastal shipping so I expect other freight is being carted as well.

      1. Stuffed that up. Should read.
        Is coastal shipping being used to transport freight other than export/import within New Zealand as well.

  7. Isn’t there a rule of thumb ,that the”benefit” of any roading “improvement “,lasts as long as the time taken to build said improvement before status quo resumes. Building a motorway system in any major metropolitan area,certainly improves traffic flow,imagine Auckland without motorways. The Waterview Tunnel and Upper Harbour projects in my view completed the Auckland Motorway network,sadly without PT provision.
    The folly once the whole network is established, is in expanding said network,based on the case that the initial project worked well,so expanding it will make it better. Getting this message across to the commuting public is key,having a transport agency who actively undermines anything but road building isn’t helpful.

  8. I seen this message many times with really extreme examples of 26-line motorways, nobody bothers to explain the idea further: does this mean no motorways shall be built at all, no road may be widened and new public roads should not be built at all? Same question as with population density?

    Is there a limit for good? Kowloon had extreme population density and no roads at all. Was it good? Dharavi slums has not many roads. Is it any good?

    Any links for further reading?

    1. If you put “induced demand” into the search bar here on GA there are quite a few posts that come up, and generally have good links to other papers and articles.

      Decisions made at each particular time need to further the goals of the city, which can be wide-ranging, so there are no universal absolutes.

      1. Majority of posts similar to this one make an impression that “no roads” is the absolute.

        I’ll search, thanks.

        1. In NZ we’ve invested in one mode at the expense of the others for decades, so evidence-based reasoning for the NZ situation will tend to use that absolute, yes. Other places, not so much.

        2. Majority of posts similar to this one make an impression that “no roads” is the absolute.

          Practically none of the posts make that impression.

          If people get sucked into that line of thinking – that any suggestion away from car dependent planning means all roads will be stopped and your car taken away from you – they really need to visit their own biases and how it shapes their thoughts and responses.

  9. The vocal minority that go. On talkback radio seem to have a disproportionate influence upon policy. Stuck in traffic and frustrated about their commute.

  10. Congestion is the prize of a thriving economy. We won’t take that prize away, but we can offer people alternatives. One of the blankets thrown of what is really going on is to lump all “improvements” together in one category. Improvements to priority modes, safety and emissions do not have to offer extra car capacity. It’s better to offer people economic and social benefit without car trips where possible (15 minute neighbourhoods, PT, bike paths). Its also time to stop asking Autoholics Anonymous for opinions on investment and broaden logistics from just trucking.

  11. We also see the reverse situation to often where infrastructure actively discourages other modes. Our local supermarket has plans to redevelop, including adding a new access road for delivery trucks that will enter the road right next to the town centre’s pedestrian crossing, to enable this they also plan to remove the footpath that goes from street to shop door and replace it with more parking.

    1. Honestly? That’s shit planning. Where is council when you need them? Why do they have a planning department and why do we make submissions to council plans if this sort of thing isn’t stopped by the council?

  12. Why is induced demand still ignored?

    a) Technically the issue can be addressed, its purely political it hasn’t.

    b) It needs a holistic systems approach (landuse and transport) not the piecemeal approach we currently have

    c) Continued subsidy of vehicle travel cross-subsidises greenfield development which drives further demand for road capacity & then more greefield development as a cycle.

    d) Land use laws, policies, rules & regulations allow further car dependent development – e.g. low density housing. Planners write rules without having to bear the costs of those rules & nor are each of those rules subject to full impact assessment.

    e) Ratepayers cross subsidise road users

    f) Vehicle drivers dont pay all the direct costs (e.g. due to ratepayer subsidy) let alone the externalities

    g) Population growth will always drive the need for some capacity increases but these would be much lower with full road user pricing (including externalities)

  13. Has anyone OIA’d anonymised KPI sets for the senior exec teams at Auckland Transport and Waka Kotahi? The results they get are so consistently bad for safety and carbon emissions that they must be incentivised to be working entirely on other indicators.

    1. That would be good, Anthony. It wasn’t so long ago that I heard that Exec staff KPI’s still include regressive measures like “productivity” on the road network. Any chance you could put in a request? You might get it earlier than I will.

  14. Matt, a comment was brought up on another platform, that I think is a good point.

    Where you say:
    And yet, billions upon billions of dollars continues to be spent on road widening in the hope that the next road might actually – finally – achieve the desired goal of reducing congestion.

    Now I assume you mean that this is what the public think the goal is, not that this is the primary goal of the NZTA, but it might have been worth clarifying the premise of the argument.

    It can’t be the NZTA’s primary goal, only the most dense transport engineer would think that the widening would help congestion over the long term. The NZTA not clearing this up is the real issue I think, they know that it wont help past a couple years with congestion, but they put that it will help with congestion on every project page, not saying how much, not correcting news outlets that say it will “bust congestion” etc, not correcting their model that says it wont not help overall, or will be a long time before the congestion is back or above pre widening levels.

    Its not that they’re crazy, its more like lying through omission, or letting the public believe what they want to believe to support the projects they want to do.

  15. Where is Patrick Reynolds in all this? I thought he was invited onto the Waka Kotahi board, in-part to counter this entrenched and out-dated thinking.

    1. The board are quite impotent, they’re fed hand picked, sanitised, summary information from the organisation, so they never hear anything except exactly what the executives want to give them.

      They are also specifically expected to avoid engaging in operational or tactical matters, their role is strategy, apparently.

Leave a Reply

Your email address will not be published. Required fields are marked *