CityLab have put together an excellent summary on the concept of induced travel demand, that we have discussed many times over the years. It’s a really good piece because it pulls together a lot of different points on induced demand into one place.

It starts off with a great example of induced demand, from Texas:

With 26 lanes at its widest point, the Katy Freeway in the Houston metro is the Mississippi River of car infrastructure. Its current girth, which by some measures makes it the widest freeway in North America, was the result of an expansion project that took place between 2008 and 2011 at a cost of $2.8 billion. The primary reason for this mega-project was to alleviate severe traffic congestion.

And yet, after the freeway was widened, congestion got worse. An analysis by Joe Cortright of City Observatory used data from Houston’s official traffic monitoring agency to find that travel times increased by 30 percent during the morning commute and 55 percent during the evening commute between 2011 and 2014. A local TV station found similar increases.

The Sisyphean saga of the Katy Freeway is a textbook example of a counterintuitive urban transportation phenomenon that has vexed drivers, planners, and politicians since the dawn of the automobile age: induced demand.

I-10 Katy Freeway, Houston, Texas
Katy Freeway in Houston, Texas. 26 lanes and still congested.

So what is induced demand? At it’s core, it’s a very simple concept – that people respond to incentives. If you make it easier to drive at peak times, more people drive at peak times. The CityLab piece explains how there’s both short-term and long-term versions of induced demand:

Induced demand is often used as a catch-all term for a variety of interconnected effects that cause new roads to quickly fill up to capacity. In rapidly growing areas where roads were not designed for the current population, there may be a great deal of latent demand for new road capacity, which causes a flood of new drivers to immediately take to the freeway once the new lanes are open, quickly clogging them up again.

But these individuals were presumably already living nearby; how did they get around before the expansion? They may have taken alternative modes of transport, traveled at off hours, or not made those trips at all. That’s why latent demand can be difficult to disentangle from generated demand—the new traffic that is a direct result of the new capacity. (Some researchers try to isolate generated demand as the sole effect of induced demand).

Initially, faster travel times (or the perception of faster travel times) encourage behavioral changes among drivers. An individual may choose to take the new highway to a more distant grocery store that has cheaper prices. Trips that may have been accomplished by bike or public transportation might now be more attractive by car. More distant leisure and business opportunities might suddenly seem worth the trip. In aggregate, these choices put more cars than ever before on the newly expanded road, increasing net vehicle miles traveled (VMT) (and greenhouse gas emissions).

In the longer term, roadway expansions make an impact on the human and economic geography of an urbanized area. Businesses that rely on trucking are more likely to locate near these new roads. With those new jobs, and access to countless more via the higher capacity road, housing developments and shopping centers spring up nearby. Urban form responds to existing infrastructure: Roadway capacity expansions spawn autocentric development patterns that utilize the new roads.

These short- and long-term effects eventually bring the expanded road back to its self-limiting equilibrium—in other words, back to capacity, fulfilling Downs’ Law of Peak Hour Traffic Congestion.

In short, immediately you might see people change the time, route or mode they are travelling by. This might have some benefit to them – they’re able to travel at the time they’d prefer and perhaps by a faster transport mode, although this individual benefit will quickly reduce. In the longer-term, you might see more dispersed land-use patterns as these more far-flung areas are now within a “reasonable” commute time.What’s particularly interesting though is how induced demand can effectively negate the travel time savings benefits from new roading projects within a decade. Remember that the vast majority of the supposed benefits from roading projects is travel time savings.

In this paper from the Victoria Transport Policy Institute, author Todd Litman looks at multiple studies showing a range of induced demand effects. Over the long term (three years or more), induced traffic fills all or nearly all of the new capacity. Litman also modeled the costs and benefits for a $25 million line-widening project on a hypothetical 10-kilometer stretch of highway over time. The initial benefits from congestion relief fade within a decade.

One question that comes up a bit about induced demand is whether it also applies to public transport. The answer, of course, is yes. Auckland is a great example of how improving our PT system over the past decade has resulted in many many more people using it. However, for public transport, induced demand is almost always a good thing. It supports more frequent services, which means less waiting for customers and ultimately it can justify major infrastructure improvements that deliver much faster and more reliable travel times. In contrast, travelling by car only gets worse as travel demand increases.

Despite induced demand being well accepted in the transport profession, the extent to which it is taken into account in business cases and other transport assessment processes remains very unclear. I know that some transport models simulate some forms of induced demand (routes and modes) but not others (land-use changes and travel at different times of day). Furthermore, it still seems as though the travel time savings benefits are assumed to be maintained over time, whereas Todd Litman’s graph highlights they are usually gone after 10 years.

Share this

41 comments

  1. Great post, Matt, but you’re being too cautious with “the extent to which it is taken into account in business cases and other transport assessment processes remains very unclear”. This is what we know, from an OIA I sent to NZTA:

    “It is industry standard that newly created trips are the least likely response to new infrastructure and it is not normal practice to include them… The Transport Agency is not aware of any models or projects that include newly generated trips in addition to other induced traffic.”

    So every project in NZ has travel times calculated as if the number of person-trips is the same with the new project or without it. That completely buggers the business cases and has resulted in huge wastes of money being spent on the projects that induce traffic.

    1. NZTA, Transfund, Transit NZ, National Roads Board (different names same views) are the reason we had to use a fixed matrix for so long. Their view seemed to be that a variable demand would over-estimate benefits for a lot of projects (obviously those outside Auckland). If a project results in the same number of people travelling but for a shorter time there is a benefit. If a project results in more people travelling but for the same time that is also a benefit (but not as big). Why do people divert to a new route, mode or origin/destination? Because the benefits to them outweigh the costs.
      The real issue is if the new project increases travel time beyond the level benefits of the new travel. That is why we need real world models that include all of the effects.

      1. >>Why do people divert to a new route, mode or origin/destination? Because the benefits to them outweigh the costs.

        And because they don’t have to think about the long-term disbenefits to society of more car travel, something that also doesn’t seem to be counted in the models.

  2. This post explains a lot, a road’s self limiting equilibrium and Down’s Law of peak hour traffic congestion, are these ever evaluated or taken into account by NZTA road planners?

    1. Yep, and it’s all due to decisions made by ignorant politicians, or by them pandering to the ignorant public. Ignorance ain’t all bliss.

      1. Traffic modellers and engineers are required to advise clients of the limitations of their analysis and warn against inappropriate use. The limitations of the models used in traffic modelling are well established, as is research into the damage cause by their misuse.

        Has anyone else seen any evidence that the NZTA modellers were advising the politicians with appropriate warnings? On the contrary, are they likely to do this when part of an alliance with the road construction companies? Of course the road construction lobby-influenced political parties (of both left and right, historically) have had their own agenda, but I believe in the last 15 years the time has been right for the professionals to step up and intervene, using the ample international research to describe how their models should be used – which is as a useful moment-in-time analysis of where the traffic is likely to be, and most definitely not as a tool to calculate the business cases.

        1. Tbf, I have yet to see the NZTA need to twist a politician’s arm to get a road project underway. In fact, it was far more likely that the NZTA was given grief for its caution and delay. “Vision” often trumps calculation.

          I also have seen that the constituency that lobbies for more roads is a lot wider than just road construction firms.

          I think Ari’s succinct comment, and John Barker’s prelude, capture the essence of the problem. Road construction projects are the gifts that keep on giving when you are politicking: big money to show you care, a “useful” physical monument, a few local jobs for a while, and a series of press opp’s for your trouble.

  3. That picture up the top is incredible. Incredibly sad, that is. All those thousands of cars, and only one bus, also stuck in the same traffic. There’s really no hope for America with that sort of attitude to road design.

    Can you imagine what it would be like to be stuck in that jam?

    1. AFAIK, top picture is from China. It’s a toll booth at Chinese new year or something when everyone flees the city to go back to their home town to celebrate.

      1. Yes, the Great China National Highway 110 traffic jam, beginning August 14, 2010. It was 100km long, and lasted 2 weeks. Many people were able to move their vehicle 1km per day, and some were stuck for 5 days.

  4. Completely agree with all this. But against the stats is the question of whether people perceive they’re living better lives because of changes to roading infrastructure. Obviously the people in homes displaced by the project are not – and while projects like Waterview get built, and completion is repeatedly delayed, the economic benefits the project was designed to produce are eroded by increased travel times, that need to effectively be recouped once the project is complete.

    I hated the notion of Waterview’s colossal cost and disruption – yet, in hindsight, I think it’s a success. It’s taken the heat off Dominion Road, allowed the southern half of the isthmus to connect via what is effectively a ring road, and splits traffic heading south, at least until the choke valve at Manukau. And so far, it’s not full – in fact its rigid speed limits and enforcement are a fine example of Katsuhiro Nishinari’s 40m rule in action – everyone’s at a fixed speed, no one can overtake and people ease back to maintain smooth progress. Which is why I’m bewildered by the 100Km/h kilometre between St Lukes and the tunnel, across all lanes. And I have to say, I think tunnel users largely ignore it, with few bothering to accelerate much past 90.

    Avoiding projects that create self-generated demand is one thing, but expecting the rule to work in reverse – that by deliberately making the roads impassable to cars you somehow generate all the positive outcomes of a shift to PT – is idiotic and politically perilous. Because there’s a far more general rule in play here – build it, and they will come. If you don’t – and despite planners’ schematics we haven’t – then they won’t. People can’t ride on trains that aren’t there. They just get angrier and uglier.

    1. There is a hysteresis that affects the rule working in reverse, yet it does still work. Providing the necessary better PT to enable people to make the change can’t happen in isolation from the physical situation: that PT needs space, and so there will be initial disruption to traffic flow. Tolerance of this is required if we are to change trajectory towards that of compact cities and away from that of the ones shown in the pictures above.

      1. Agree – let’s get on with it! It’s not just Waterview that can take the term of people’s natural lives to complete – check out Franklin Road – and no trams there! Not sure about loops of hysteria though…

        1. I have hoops of wisteria in my garden that are quite calming even though I think their growth pattern follows the rules of induced demand.

  5. I think there is a difference between induced demand and population growth demand. I’m guessing a lot of our congestion problems in Auckland are more related to the latter. Either way, there is no way they can build more roads to fix it…

    1. We need to lay this myth to rest. The different places you can accommodate growth are:

      – on the outskirts of the city, where the distances involved in accessing amenities and cultural facilities leads to car dependency, or
      – within the existing urban form, where PT already exists but can be improved by the larger pool of residents.

      The former increases vkt. The latter decreases it, even with population growth, because the PT is improved for the existing residents too.

      We could have been decreasing our vkt in the last couple of decades. The road building and greenfields development prevented this.

      1. On the outskirts you can accommodate people where land is cheaper so some houses can be built small and affordable. In the centre you can accommodate people on high priced land which then gets bid up even higher as you have to use rules to restrict development of the outer land because otherwise people will opt for the outer as a substitute.
        So the result is high priced land goes up in price even higher. The restricted supply means only wealthier people get a look in. Wealthy people want bigger houses/apartments.
        The result is a city that is liveable if you have a lot of money. Everyone else has to leave or face housing poverty.
        But hey think of how much fun it will be for wealthy people to ride on light rail.

        1. Bollocks. If you accommodate people on the outskirts where land is cheaper you are inducing traffic but have been externalising the costs of that traffic onto everyone else. You are also laying waste to huge areas of land for car infrastructure because cars need space.

          You seem to think the “baseline” cost of a house is that of a house on the outskirts, but that house on the outskirts is subsidised so this is a fallacy.

          We cannot justify car dependency. We cannot justify increasing vkt. Our planet cannot cope. Greenfields growth causes car dependency and increases vkt. We cannot justify it.
          The “baseline” cost of a dwelling needs to be set at that of a home built where agricultural land is not ruined and where transport can be provided sustainably.

          1. Which is fine if you can afford to live in a city like you propose. In the US they have cities like that, but they also have cities for poor people that you would never want to live in. They have an apartheid system based on wealth.
            We have ridiculous rules that limit development in inner areas but even if we got rid of them land would still be expensive. It has to be to justify apartments. Then there is the added cost of building high. Then all the infrastructure to support those developments has to be provided. Yes they are subsidised too. But under the current dogma we add up everything in the outer areas but exclude everything in the inner areas. It is cheaper to provide water and wastewater services to Redhills than the CBD, Sure transport will cost more but you get better amenity. Value matters more than cost.

          2. Taking Patrick’s figures from his chapter in the Big Questions book:

            “The nation spends about $5 billion in taxes and rates every year on land transport infrastructure, mostly roads, and we, the users, then spend another three times this, around $15 billion a year, buying, insuring, repairing, and fuelling them. Additionally, a dizzying amount of land is lost to more productive use by being reserved to store them several times over; where we work, shop, play, and live.”

            Your greenfields developments really save us more than $20 billion a year because the land is cheaper? Show me the figures, miffy.

          3. Are you seriously suggesting that if we didn’t build houses in outer areas we can avoid the cost of all roads everywhere in the country? That would only work if we all live and work in one big building. Roads don’t just exist to serve outer suburbs, they link our cities, farms, ports. Some people even walk or cycle on them or ride a bus.

            If your plan for the future will only work by taking away other people’s choices then it probably isn’t a very good plan. Intensification for all growth only works if you regulate a hard edge to the city and if you are prepared for the hardship it creates.

          4. “Are you seriously suggesting that if we didn’t build houses in outer areas we can avoid the cost of all roads everywhere in the country? ”

            No, but then I didn’t even attempt to include the “dizzying amount of land is lost to more productive use by being reserved to store [cars] several times over” Nor did I even attempt to include the costs the road network imposes on the planet through climate change, pollution of our local environment, social ills, public health ills.

            So of course roads are needed. But the costs imposed by them are far in excess of $20 billion a year too.

            Auckland covers so much land already. We have a lot of intensification to do for catch up.

          5. Yes Heidi but what are the benefits of those roads. If it exceeds $20 billion then I have no problem with that cost. Houses cost billions too so does healthcare and education and food.

          6. Its not just roads, its all the carparks too. When everyone lives in far flung homes and owns multiple cars then all development citywide requires spaces for all those cars… a space for your car at home, at work, at play, everywhere. This is the major loss of productive land, on top of roads. I would be curious to know how much of aucklands privately owned land is currently used for carparking. Does anyone have an estimate on this?? And we can then calculate some costs…

        2. “some houses can be built small and affordable”

          But are they? Awful lot of oversize McMansions get built (because that’s what makes developers a good (subsidised) profit under the current rules).

          (Maybe that should be “a lot of awful …”.)

          1. That is the result of land constraints. When we allowed lots of subdivision we got lots of different sizes of house to meet the market. Once the MUL became a means of stopping development rather simply coordinating development then only wealthy people could afford a section so the market responded with big houses only. Maximising profit is what markets do. Rules bite poorer people most severely.

          2. Agree miffy. The main thing that is stopping people having the house that they want (whether that be big/small, affordable/mansion, sprawl/central, modern/old) is council planning rules. The council have no chance of knowing what everyone wants – so they end up only representing the most audible people (generally the rich).
            Why not leave it to the people to decide what they want?

          3. Isn’t a lot of the McMansions & for that matter a lot of expensive oversized apartments also due to previous lack of rules restricting overseas investors & also bank lending rules meaning the demand was virtually limitless. Empty houses people sit on for capital gain. I suspect with recent changes & market signals this is having an effect.

        3. “In the centre you can accommodate people on high priced land which then gets bid up even higher as you have to use rules to restrict development of the outer land because otherwise people will opt for the outer as a substitute.”

          But this is the opposite of what we do now. We make it really hard to build in the inner areas and really easy to build greenfield. I realise there is the MUL but that is nothing compared to the rules that prevent dense inner city living. Minimum lot size, maximum coverage, minimum set back – all of these act as massive brakes on inner city brownfield development. especially in inner suburbs.

          And when good quality dense housing is built, it is gobbled up by people who don’t want to live a long way from the centre.

      2. I would say everyone should have a choice where they live – for example a big house in sprawl, or a compact central house. But each option has its downsides, such as a bad commute or having no back yard). And you shouldn’t expect the taxpayer to pay to fix these downsides.
        When people out east or on the shore complain of the bad traffic I can’t quite work out why – its been bad for as long as I remember, Auckland’s population has been growing for as long as I remember, and there are no easy fixes. You chose to live there knowing this.

        1. Nonsense. The reason that traffic is bad in the north and east is the poor provision of public transport, not because of historic traffic numbers. It comes down to a refusal by central Auckland councillors to provide the same public transport infrastructure north of the bridge that they deliver to themselves, not peoples choices of where they live.

          This current council is as useless as the previous ones in this regard, and just take a look below at what is being planned to fix this (nothing).

          https://www.aucklandcouncil.govt.nz/plans-projects-policies-reports-bylaws/our-projects/proposed-10-year-budget-projects/10-year-budget-2018-2028-transport-projects%E2%80%8B%E2%80%8B/Pages/default.aspx

          Everything north of the bridge has literally no significant investment in public transport, everything south of the bridge has at least electric trains. That’s why people in the north and east complain about bad traffic.

          1. Oh FFS get over yourself. The North Shore is smaller than Greater Christchurch yet 10 years ago you got the Northern Busway. https://en.wikipedia.org/wiki/Northern_Busway,_Auckland

            “The busway became fully operational in 2009, with some final sections being completed with little publicity, for around NZ$290–294 million: $210 million for the busway and $84 million for the stations.The project was funded by Transit New Zealand, ARTA, Auckland City Council and North Shore City Council.”

            I think someone else maybe due some rapid transit funding….

  6. So Jimbo asks why people complain, I take the time to respond and I get trolled. Nice one Brendan!

    To the facts, the busway is very good but the problem is that there are no transit stations at any North Shore urban centres, so for many the busway is hard to get to. All of the stations are by the motorway and we haven’t extended the busway out to where people live yet. Frequent buses are few compared with central Auckland. For me (Devonport) the busway is two bus trips away (45 minutes) and those buses are 30 minutes apart.

    If you don’t think 30 minute wait times plus double-transfers is not going to induce traffic then I wonder what you think will? I stand by my view that poor public transport provision is traffic-inducing and I can tell you that I have first-hand experience of it every week.

    1. What you describe is a function of 100+ years of planning, rather than current funding mechanisms. The reason we have good HR south of the bridge is as much to do with NZR choosing routes for long distance trains in and out of Auckland over 100 years ago as it is to do with recent spending.

      I agree it is one of the main reasons that the Shore is very car dependent, but from my observation that is exactly how North Shore councils and their voters have wanted it.

  7. Complaining about Greenfields is fine, but we don’t exactly have affordable apartments as an alternative. There are currently social costs from artificially restricting the supply of land; there always will be. They’ll be exacerbated if planning rules don’t allow for a supply of decent density. But going on and on about the ‘social costs’ of greenfield when it’s the only safety valve we have for housing supply-relate price rises while ignoring the social costs of high house prices is the definition of having your cake and eating it too.

Leave a Reply