Removing Roads Speeds Urban Travel

Yes, you read it right. More roads leads to more congestion and slower travel times, even if the amount of cars/drivers remains the same. It seems that the Nash equilibrium, which drivers achieve simply by acting in their own self interest is not the most efficient way to move people across our cities. An article in this month’s Scientific American gives us the details:

Conventional traffic engineering assumes that given no increase in vehicles, more roads mean less congestion. So when planners in Seoul tore down a six-lane highway a few years ago and replaced it with a five-mile-long park, many transportation professionals were surprised to learn that the city’s traffic flow had actually improved, instead of worsening. “People were freaking out,” recalls Anna Nagurney, a researcher at the University of Massachusetts Amherst, who studies computer and transportation networks. “It was like an inverse of Braess’s paradox.”

The brainchild of mathematician Dietrich Braess of Ruhr University Bochum in Germany, the eponymous paradox unfolds as an abstraction: it states that in a network in which all the moving entities rationally seek the most efficient route, adding extra capacity can actually reduce the network’s overall efficiency. The Seoul project inverts this dynamic: closing a highway-that is, reducing network capacity-improves the system’s effectiveness.

Although Braess’s paradox was first identified in the 1960s and is rooted in 1920s economic theory, the concept never gained traction in the automobile-oriented U.S. But in the 21st century, economic and environmental problems are bringing new scrutiny to the idea that limiting spaces for cars may move more people more efficiently. A key to this counterintuitive approach to traffic design lies in manipulating the inherent self-interest of all drivers.

A case in point is “The Price of Anarchy in Transportation Networks,” published last September in Physical Review Letters by Michael Gastner, a computer scientist at the Santa Fe Institute, and his colleagues. Using hypothetical and real-world road networks, they explain that drivers seeking the shortest route to a given destination eventually reach what is known as the Nash equilibrium, in which no single driver can do any better by changing his or her strategy unilaterally. The problem is that the Nash equilibrium is less efficient than the equilibrium reached when drivers act unselfishly-that is, when they coordinate their movements to benefit the entire group.

The solution hinges on Braess’s paradox, Gastner says. “Because selfish drivers optimize a wrong function, they can be led to a better solution if you remove some of the network links,” he explains. Why? In part because closing roads makes it more difficult for individual drivers to choose the best (and most selfish) route.

The article goes on to discuss the phenomenon known as shared streets, where all traffic lights, signage and lanes are removed from roads and space is shared equally between cars, trucks, bikes and pedestrians.  One would think that anarchy is the result, but in fact traffic flows and safety both improve. Who would have thought that the French method of driving was optimal, or even safe?  😉

Finally, there is a discussion about zoning that requires developers to provide x number of parking spaces per square metre of building, and how this seemingly good idea subsidises car use and makes congestion in our cities much, much worse.

In a misguided effort to reduce congestion, planners in the 1950s required developers to provide a minimum number of free parking spaces-a strategy that “completely ignored” basic economics, Siegman says, referring to how lower prices increase demand.

Now limited urban space and concerns about global warming are inspiring city planners to eliminate these requirements. In San Francisco, for example, developers must restrict parking to a maximum of 7 percent of a building’s square footage, a negligible amount. Although downtown employment has increased, traffic congestion is actually declining, Seigard says. With fewer free spaces to park, drivers seem to be switching modes, relying more on mass transit, cycling and just plain walking.

While I am not advocating the removal of any roads in New Zealand, (I was thrilled when they finally completed spaghetti junction in Auckland), the building of new roads is clearly uneconomic. Let’s fix up and otimise the ones we have, while providing real alternatives that reduce our dependence on foreign oil.

34 Comments Posted

  1. TRevor29But fixed route public transport can only serve the needs of a small percentage of the total population.
    It does not serve the majority of transport demands in our cities and certainly cannot deal with the transport needs of our small cities, rural towns and rural countryside.
    These arguments tend to focus on the [assumed} efficiency of the vehicle rather than the efficiency for the user.
    Container ships are the most efficient transport vehicles but if you are exporting live mussels to California you will send them by air because for you air transport is more efficient overall.
    Similarly if public transport requires that I make four modal shifts and takes three times as long as a trip in the car then the car trip is much more efficient for me than public transport – regardless of the “efficiency” of the train or whatever over one leg of the journey.
    And if I have a passenger in my car then it eats buses and trains for vehicle efficiency and that performance will actually improve faster than the public transport fleet. Have you looked at the cost of replacing and expanding a whole fleet of buses and trains with all the support systems?
    Why would you do it? What is the appeal?

  2. Owen, I was quoting Jarbury’s earlier comment and making the point that in Japan the generation that is growing up with virtual face to face to sovial contact is eschewing private car ownership. There is evidence that this trend has been reducing rail travel as well as car sales.

  3. A business as usual increased public transport system will use more energy and generate more greenhouse gases and more pollution.

    If the increased public transport system uses electric trains and trams, then it may use less energy. Even if it uses electric buses (or flywheel storage or compressed air powered buses), the increased energy use is likely to be small, and if we can generate our electricity from renewable resources then there will be less pollution and less greenhouse gases.

    It will take far longer and far more money to convert our private vehicle fleet to run on electricity or to run on rails.


  4. The Korean example shows the value of orbital freeways, ring roads or bypasses or whatever you want to call them.

    The original De Leuw Cather motorway design dropped off all the CBD destination traffic on the ridges around the CBD and did not use Grafton Gully at all.

  5. bjchip,
    I should have said sub-thread which is where I came in re spaghetti junction.

    The loading of private cars on non work trips is higher than on public transport so what are the gains of having this massively expensive public transport system?

    More energy, more greenhouse gases, more pollution and much more cost.

    Where’s the beef?

  6. The incomplete 6 mile Cheonggyecheon freeway was replaced with 68 miles of busways and 100 miles of orbital freeways. The freeway was incomplete like Auckland’s motorways before spaghetti junction was completed, and it was parallelled by two six-lane arterials. The fact that the freeway removal/river restoration project was the brainchild of the founder of Hyundai, one of the masterminds of production line ship building, is a pretty good indicator that they knew that the 1970s Cheong Gye Cheon Freeway was redundant in the new radial freeway system then under construction. The subway system was also being extensively expanded.

    Looking into the history of this freeway it turns out it was built when there were very few cars in Seoul and the main motivations were to cover the sewer and erect a symbol of progress. A decade later when a full freeway system was being planned to cope with traffic growth it became obvious that the only way to get the freeways traffic out of the CBD was to tunnel under the city to connect to a southwest freeway but they never had the money to do that. The new freeway plan was intended to stop the city sprawling outwards by redeveloping the inner city. Turning the sewer into a river parkway was the centrepeice of the plan. With the new subways, busways and orbital freeways the unneccesary and decrepit freeway could be removed with no ill effects.

    There are a lot of examples from all around the world where sections of the road network have been removed with positive effects on traffic flow. Engineers have generally favoured reduction in intersection turbulence as the explanation for the improvement. But obviously it was the drivers taking the “short cut” who were creating the turbulence, hence entirely consistent with Braess’s paradox.

    In studies of both planned and unplanned freeway closures in California it has been found that the majority of drivers switch to alternate freeways or surface streets, increasing there distance travelled by 7% but without affecting travel speeds on the other freeways and streets. Again supporting Frog’s argument. However only 2% switched to public transport, much lower than the percentage who simply avoided travelling altogether.

  7. Apologies, I rather thought there was a missing general-specific condition. There was. The thread started out with an observation about a Korean road network so I wasn’t focusing on Auckland.

    Dead right Auckland isn’t NYC and there is virtually no capacity in Auckland to use public transit to go to general destinations or function without a car except in an economically and personally crippled manner. A good bike would be useful but the place needs a lot more public transit before it works as well as one of the majors.


  8. What makes you think public transport is useless for non-work trips Owen? If a city has an excellent public transport system people can and will use it for “errands” and “leisure trips” which actually make up the majority of trips. Furthermore, if streets are made better for cycling and walking then a lot of car-trips could be eliminated altogether – particularly with mixed use higher density neighbourhoods. Apparently about half the trips made in Auckland are under 10km in length.

  9. I was writing about Auckland (spaghetti) because that was the topic of the thread.
    I said only a few percent of Auckland ‘s trips are commuter trips to the CBD.
    Auckland is not New York or Hong Kong.

    And even in New York public transit is not used for commercial trips.

    When the topic is New York I use New York stats.
    But this thread was about Auckland. And so are my stats and statements.

  10. Owen

    Depends on density. In NYC one can very easily live completely car-less and many people do. Public transit is used for everything. Same thing in Moscow or Boston or London.

    Take care with blanket statements. It depends on the public transport system’s character too.


  11. Most vehicle trips are not work related and hence are not amenable to public transport other than taxis or the new iPod/GPS?Internet based ride sharing systems such as Avego.

  12. Owen, you are correct. Unlike most other cities in the world, you cannot divert the freight traffic going through Auckland because there is ocean on either side. State Highway One as it stands is a necessary motorway and would need to be kept over the long term.

    With regards to the other 95% of trips, I would imagine that you could see a massive bus network spring up. Of course, it might also make Auckland’s CBD an attractive option, although that is already occurring with the public transport improvements that are being made (in line with what happened with Brisbane in the 1980s and Perth in the 1990s).

  13. The idea that people using state highway 1 would divert to surface streetsor use transit reveals a lack of understanding of the use of this highway.

    40% of the traffic is through traffic and downtown Auckland is not its destinatiion.
    And a large percentage of all traffic (sorry I cannot remember) is commercial traffic including heavy trucks.
    So you really think you improve anything by having them trundling through urban streets. And the commercial traffic cannot be diverted to public transport.
    These are complex networks having to cope with complex demands.
    ONly a few percent of metropolitan Auckland’s trips are commuter trips to the CBD. YOu have to deal with all the other 95% or so.

  14. Could someone answer a question for me. When that Seoul Motorway was undergrounded, how much did public transport patronage increase by? Did conditions on those services get worse, and by that, I mean, were commuters treated even more like cattle than before?

    If anyone is going to put barriers across Spaghetti Junction, at the very least make sure that there are enough vehicles with the alternative options. The rail service, which already suffers from overcrowding, would hit breaking point. So would bus services as well.

  15. Frog, It is equally true that the “general principle” that no more roads mean no more congestion is false.

    And that the “general principle” that more PT means less congestion must also be false.

    ie, there is no general principal that can be applied to every link in a land transport system.

    At the heart of this research conclusion is the notion that reducing driver’s choices reduces conflict and consequent turbulence thus improving traffic flow even when capacity is reduced. That was core argument for introducing limited access motorways – to rliminate the conflict between local and longer distance traffic. Many American freeways have even been divided into seperate freeways within the same corridor, one set of lanes for through traffic and another set for city bound traffic.

    I think the most important thing to be gained from this type of research is that each link has to be assessed to be assessed for it’s impact on the whole network and not just it’s immediate vicinity using specific traffic flows not just the general flows used in the developing the master transport plans of the 1950s. That is probably even more important when considering links in a multi-modal land transport system. Unfortunately the interaction with land use decisions is currently too poorly understood to be able to do that. GIGO rules, still.

  16. Owen, I would not argue with the notion that closing spaghetti junction would mean chaos. Nor was I arguing that point. The point was that the “general principle” that more roads mean less congestion is false, and that is the primary driving principle behind our current roading binge.

  17. I think putting a barrier in the middle of State Highway 1 would create a problem initially. However, over time people would use a variety of alternative transportation means, they’d decide to stay at home more or catch the train. Over time congestion would lessen.

    People interested in this concept should read more about “woonerf”, the Dutch principle of mixing traffic and pedestrians together to make life safer for everyone. It sounds counter-intuitive, but it has worked incredibly well. A “shared space” is being introduced to Exhibition Road in London – outside the Natural History and Science Museums.

  18. This behaviour of networks has been known some time and has been modelled with water and electronic analogues. Networks like this show some elements of chaotic behaviour. As a general rule a bridge is either dramatically underdesigned or dramatically under used.
    However, it is a serious error to draw from this example of reducing choice reducting congestion to a general principle. IF you put a barrier across the Auckland State highway 1 at spaghetti junction I can assure you that congestion will increase because so much of the traffic is not interested in Auckland at all.
    Also, what we designed for Auckland in the sixties was a genuine network with circumferential bypasses. These have never been built which means that an accident on the existing through route brings the whole thing to a standstill.
    My own view is that completing the Western Bypass at Waterview (without tunnels but the same surface road that Phil Goff’s electorate had to put up with, when combined with land use regulation which allows land use to churn in response to the change in the road network may well obviate the need for a second harbour crossing.
    So by adding one element to the road network you can reduce the need for another element somewhere else.
    Auckland is under-roaded. Just like Los Angeles.

  19. Sorry BJ. This is one of those situations where the original design ‘promised’ dramatic improvements but the compromised design has only produced moderate improvements. It is only when the performance of the compromised bypass is compared with the performance estimates in the original proposal that we arrive at Toad’s conclusion that “It didn’t do what it was supposed to do”.

    This is similar to Auckland’s old central railway station not doing what it was supposed to do and therefore being a huge waste of money. That situation occurred because it was designed as part of the Morningside deviation, and that deviation was never built.

    Unfortunately the fact that the performance of a network is always greater than the sum of it’s parts means that opponents of any transport network can always argue convincingly that completing the network is just pouring good money after bad. Then when the incomplete network performs miserably they advocate some other grandiose network, which will also fail because of lobbying from it’s opponents.

  20. Dunno about that Kevyn, can you specify what changes to the design happened as a result of ‘objectors’? I think the impacts here were more about peripheral issues – heritage buildings and so on. The major design changes (such as ditching the second Terrace tunnel) seem to have been mainly cost related.

  21. Kevyn

    I still don’t know how it is “not doing what it is supposed to do” as my experience with it has been positive in all respects.



  22. Toad, The Wellington inner city bypass didn’t do what it was supposed to do because objectors ensured it wasn’t built the way it was supposed to be built. Tantamount to objecting to having the kitchen wired with extra plugs then complaining that the upgrade was a waste of money because whenever you use the microwave and frypan at the same time it trips the overload cutout on your multi-plug adaptor.

  23. All good. Funnily enough, as one who lives on an (at times) very busy road, I reckon the biggest help with traffic pollution occured when the Oil Companies went over-the-top greedy with their pricing.
    Traffic was immediately cut by two thirds and people were much more likely to slow down to conserve fuel.
    The Companies tried to reverse their fiscal error, but too late. People have changed their habits and lifestyles to include less use of fuel – much less.
    So the prices have been hiked back up (anyone seen the media reports describing why?) and, paradoxically, we have greedy oil companies to thank for the biggest single change in pollution levels so far. They have helped enormously, no matter out of the most venal of motives.

  24. The anarchy argument is an interesting one as could be seen with the power cut that hit Wellington last week. It was interesting to note how less stressful and aggressive the flow of traffic was through the city, as well as how less risky the behaviour was of pedestrians fed up with being crowded on intersections.

    I don’t know that a lot of that could be taken as being standard behaviour under that situation for extended periods of time, when uncertainty of what is happening would have been a significant influencer on people’s behaviour at the time.

    Food for thought all the same.

  25. Oh forgot to say – my commute is optimised for low stress and fuel consumption. Its by no means the fastest way in, but I get to drive at a constant velocity due to no congestion, rather than much start/stopping.

    Lesson: Not everyone uses the same optimisations, or indeed has the same goals when planning a route.

  26. I read the title as something like “removing speed limits increases traffic flow” 🙂

    The allocation of parking to buildings ratio only works if there are sufficient and adequate alternatives. Public transport in Christchurch is limited, and only works in limited hours. So we have a major hospital with thousands of staff who have no place to park. Well, there is a staff car park, but I know a member of staff there who lives by me (ie a fair way away) and is number four hundred and something on the waiting list for a place, and will have retired before getting a parking capability. So every day, petrol is wasted by staff looking for somewhere to park.

    How good for the environment is that?

  27. Toad ? Did it not improve travel times to the Hospital and the Airport? I’ve certainly found those to be more accessible since the altered pattern took effect. That’s a subjective impression, I don’t make enough trips to be certain and I certainly didn’t make any attempt to measure things ahead and I don’t know what it was advertised as being able to do either. Perhaps it falls short of the target.

    However, it is certainly not worse.


  28. I would say jumping to grand, indisputable conclusions with little or no evidence presented, or debate, is fairly troll-ish, maybe 60% troll…

  29. Frog

    So a troll is somebody who does not agree with you?

    I guess it is true, socialist are not much more than fascists in drag.

  30. That statement was intended to head off the trolls like BB whjo would instantly jump to the conclusion that Green policy was to remove roads, not just stop the building of new ones. If the research justified removing a road or two here in NZ, I’d consider it. But I don’t think anyone here has done the numbers.

  31. frog said: While I am not advocating the removal of any roads in New Zealand…

    What about the Wellington inner city bypass frog? It didn’t do what it was supposed to do, and a long park there would be kind of neat – although not as neat as all the old buildings that were demolished to make way for it.

  32. I think you’ve just highlighted very well why its increasingly difficult for us to solve problems that present themselves to us.

    You have just quoted a bunch of stuff coming from the past and from other places to come to a conclusion that any viable thinker would find looking out his own door or window, or god forbid, sitting in a bus stop for half and hour.

    But of course, solutions must come pre packaged from lofty institutions, otherwise all you decision makers would have nothing else to do other than… actually make decisions.

    The education system is so pervasive in parliament and decision making that instead of actually going on a field trip and making up your own mind about what you actually see you have made our lives dependant on the multi choice.

    The researchers go out scanning the globe for whats already been done and collecting it all up and offering it as the only viable possibilities. God forbid that NZ’s decision makers would ever try out an idea they found themselves by good old fashioned understanding.

    I’ve determined that one of my most important skills is the ability to sharpen knives, chisels etc. Why? maybe because it’s a metaphor for life.

    A single blade can do many things and the ability to work it through a variety of grinding and polishing mediums so that the finished edge can be viewed only with a magnifying glass to ascertain its sharpness is about ideas. Ideas are tools and one is far better having an idea made from the best materials to do the most amount of work, within ones particular enviroment, and being aware of how to keep those ideas sharp and appropriate to the work required than one is able to have a workshop full of specialised tools that only do a single job and then require specialised tooling to keep sharp… but hey, its so easy these days to go out and buy a new one to be used only as long as it’s needed.

    Get my drift?

Comments are closed.