Rush-hour traffic in Boston. (John Nordell / The Christian Science Monitor /FILE)
Does closing roads cut delays?
By Eoin O'Carroll | 10.06.08
File this one under “intensely counterintuitive.” A recent study has found that closing off certain streets can actually relieve traffic congestion.
Using Google Maps, a trio of scientists – Hyejin Youn and Hawoong Jeong, of the Korea Advanced Institute of Science and Technology, and Michael Gastner, of the Santa Fe Institute – looked at traffic routes in Boston, New York, and London. Their paper, titled “The Price of Anarchy in Transportation Networks: Efficiency and Optimality Control” [PDF] and published in the jounral Physical Letters, found that, when individual drivers seek the quickest route, they sometimes end up slowing things down for everybody.
It all hinges on something called Braess’s Paradox (and yes, I appreciate the irony of a Wikipedia entry that challenges the wisdom of crowds), which states that adding capacity to a network in which all the moving entities rationally seek the most efficient route can sometimes reduce the network’s overall efficiency.
The authors give a simple example of how this could play out: Imagine two routes to a destination, a short but narrow bridge and a longer but wider highway. Let’s also imagine that the combined travel times of all the drivers is shortest if half take the bridge and half take the highway. But because each driver is selfishly trying to seek the shortest route for himself, this doesn’t happen. At first, everyone will go for the bridge because it’s shorter. But then, as the bridge becomes backed up, more drivers start taking the highway, until the congestion on the bridge starts to clear up. At that point more drivers go back to the bridge, which then becomes backed up again. Eventually, the traffic flow settles into what’s called the Nash equilibrium (named for the beautifully minded mathematician), in which each route takes the same amount of time. But in this equilibrium the travel time is actually longer than the average time it would take if half of the drivers took each route.
Note that this still happens even if – indeed, especially if – all the drivers have perfect information about what all the other drivers are doing, such as with a GPS that gives real-time traffic updates.
The authors compared the Nash equilibrium time to the socially optimal travel time, and dubbed the ratio between the two “the price of anarchy.” In their study of the Boston area, which looked at travel times from Harvard Square to Boston Common, the price of anarchy at peak traffic times made for a journey that is 30 percent longer.
But the price of anarchy drops if you close a few roads, because individual drivers are less able to selfishly optimize their routes. In their analysis, the authors identified six streets in Boston and Cambridge: By closing those streets, they say, the optimal collective travel time would decrease between the two points.
At first blush, this study seems dissonant with findings that traffic flows can be improved by increasing vehicular anarchy. As I noted a few months back, there is considerable evidence that removing all traffic controls – lights, signs, road markings, and even the distinction between streets and sidewalks – can actually make traffic move more smoothly, as well as cut down on the number of accidents and increase the area’s economic vitality. The idea behind these “shared streets,” which have been successfully deployed in many European cities, is that the lack of traffic signs makes you take personal responsibility for directly negotiating with the pedestrians, cyclists, and other cars around you, instead of, say, gunning it through an intersection just because you know you have the light.
But maybe these two traffic models have more in common than it first seems. Both encourage individuals to drive more slowly so that everyone gets to his destinations faster. Both favor a holistic approach to traffic, one that designs from the perspective of the overall flow rather than that of an individual driver. And both open up more space for pedestrians.
It’s not too difficult to imagine a city designed with these principles in mind. Fewer roads with slower but smoother traffic. Spaces that can easily be converted to car-free zones to suit the needs of the network. And fewer opportunities for people to drive like jerks. Sounds like a nice place to take a walk, actually.
<< ‘Smart’ cars: How green? How safe? | MainComments
2. Evan | 10.07.08
Speaking of anarchy and closing roads, I have a video that might interest you. It’s about anarchists at the Republican convention last month. Check it out here:
http://current.com/items/89375810_lockbox_an_anarchist_tale#
3. Eoin | 10.07.08
As both commenters have pointed out, I’m using the term “anarchy” pretty loosely here. I imagine that most people who subscribe to anarchism as a political philosophy probably don’t have traffic jams in mind when envisioning the direction they think society should take.
4. Dave Reid | 10.07.08
This is an interesting report, and I feel it connects with the concept of “induced demand”. Because it is the flip side of closing roads i.e. expanding roads. Often when freeways are expanded congestion and travel times actually increase.
5. C. van Empel | 10.09.08
The idea that fewer and slower streets may not be the perfect application of Braess’ Paradox. However, more streets, that are slower and equally weighted may provide a better example of the Paradox. In other words, streets that have more lanes are more attractive to drivers because drivers believe those streets move faster than narrower streets with lower speed limits.
BUT if most streets had roughly equivalent numbers of lanes and speed limits, and there were more streets from which to choose (a gridded network of two- and four-lane streets), most streets would appear equally attractive and drivers’ choices would be less predictable than when there are distinct design differences between streets.
The additional advantage with the system of more streets, as opposed to a system that has fewer streets, is that when there is a wreck or some other obstacle, such as a lane closure, drivers are not stuck on the selected street, but can easily navigate around the obstacle by turning on other streets in an intelligible network.
6. katie | 10.09.08
As a Bostonian, I’m curious, what roads near Harvard Sq. were closed? There’s a big difference between closing off, “short-cut” side roads to routes like Mass. Ave and Mem. Dr. and actually closing say, Memorial or Storrow.
7. Eoin | 10.09.08
Great question Katie. It doesn’t say in the paper, but the sci-tech site Ars Technica did a story on it, and their graphic says that the streets are Staniford St. and Charles St. in Boston, and four sections along Main St. in Cambridge between Mass. Ave. and the Longfellow Bridge.
8. L. Montgomey | 10.10.08
Where does “crossing traffic” come into play? I prefer to take a safer route that allows me to make a right turn whenever there is no traffic signal to protect a left turn.
9. teqjack | 10.10.08
Well done, but hardly news other than finally applying something that was accepted decades ago. My memory is admittedly vague, but I recall an article in the late Sixties showing a circular route and that adding a diameter-following road cutting the circle would actually increase travel time.
10. Seth | 10.10.08
This is all fine and dandy…if you’re a parcel! I would rather spend 10 extra minutes in my car making decisions than have those decisions taken away from me. There is empirical static units of time and then there is perceived time. Ten seconds sitting on a hot stove is a far longer time than ten seconds kissing your beloved goodbye. One of the most primal human needs is the need to make choices. Part of the reason socialism and communism failed is that they ignore that very human need. Central planning may be less messy or more efficient on paper but people will be made less happy overall by having their choices restricted. If we ban private modes of transport this might be good information when designing public transportation routes but limiting choice will create other problems.
Someone pointed out that the real way to approach Braess’s Paradox is to get rid of (or lessen) the apparent variety of types of avenues. Make grids of avenues that are roughly equal in their capacity so that one route doesn’t appear to be more efficient (i.e. the highway across town appearing to be a quicker route than the surface streets).
11. Eoin | 10.10.08
Seth, your assertion that “people will be made less happy overall by having their choices restricted” is contradicted by a number of psychological studies. As a number of researchers (particularly Sheena Iyengar and Mark Lepper) have observed that a number of choices above a certain threshold lead to anxiety, poorer decisions, and sometimes no decision at all.
This idea was popularized in a 2004 book by Barry Schwartz titled “The Paradox of Choice.”
12. Graham | 10.13.08
Alternative Network design should be a investigation area for these students.
There was a fine article by Dr William Zuk (spelling?)who mathematically refuted the rectangualr grid system so futilely employed in the USA.
The ideal network is not found in an orthagonal grid system. Once the “ideal network” is found - two dimensionally - maybe we can advance the system we have built for congestion, by stepping into one that is better for our future. No matter how much we theorize as how to incrementally improve it, without honestly examining alternative geometic networks with an open mind, we are stuck with our present failure.
Decisions of choice are being aided by Garmin these days!
However I fully support the name ANARCHY.
13. Brian Fowler | 10.14.08
Without going into any detail, here’s a piece of the puzzle that I don’t see addressed. Any discussion about network layout is completely futile without addressing how the intersections operate - traffic control and traffic signal phasing/timing. These are generally the features that control flow, not link design or “network architecture”. They are also the features that have the greatest impact on people’s perceptions of efficiency and safety as they travel. There are also the issues of traffic volume and access design - having a huge impact on the efficiency of flow in a roadway network. An oversaturated network that provides the most capacity may not be the one that provides the quickest travel times in undersaturated conditions. For traffic engineers that must deal with all of the complex realities of “operating” a network, this discussion just seems to be another attempt to oversimplify something that is far more complex than most people have any real sense of.
14. Reed Agleson | 10.29.08
Decrease capacity leads to decreased travel time?
I like it. Let’s close all the roads so nothing can move and travel time drops to zero. Problem solved.
Giving arguments for special case exceptions to general truths (increased capacity in most cases yields decreased travel times) to all off the instantaneous experts bred by the internet age is like “giving whiskey and car keys to teenage boys”
This article will be used by neo-ludites everywhere to argue against increased capacity on highways when demand has long since increased to fill it. Thanks for making a functional surface goods movement system even harder to acheive.
15. Eoin | 10.29.08
Reed, I think I make it clear in this story that Braess’s paradox only applies under very specific circumstances. It would be deeply condescending avoid nuanced stories just because some people might interpret it the wrong way. When writing a story, I assume that the reader is smart.
Trackbacks/Pingbacks
Leave a Comment
We do not publish all comments, and we do not publish comments immediately. The comments feature is a forum to discuss the ideas in our stories. Constructive debate - even pointed disagreement - is welcome, but personal attacks on other commenters are not, and will not be published.
Tip: Do not write a novel. Keep it short. We will not publish lengthy comments. Come up with your own statements. This is not a place to cut and paste an email you received. If we recognize it as such, we won't post it.
Please do not post any comments that are commercial in nature or that violate copyrights.
Finally, we will not publish any comments that we regard as obscene, defamatory, or intended to incite violence.
1. urbanodelacruz | 10.07.08
“this study seems dissonant with findings that traffic flows can be improved by increasing vehicular anarchy.”
I guess you have to differentiate between “driving behavior anarchy” vs. “route choice anarchy”