Posted on Monday June 6th by Eric Jaffe | 1,435En el mundo del transporte, seguir nuestra intuicion puede perdernos. A primera vista, nadie sospecharia que eliminar una carretera importante puede hacer que mejore el flujo del trafico - no obstante eso es exactamente lo que sucede (o sucederia) en algunos casos
In the transportation world our intuition can lead us astray. On first thought, no one would suspect that removing a major road can improve traffic flow — yet that’s exactly what it does (or would do) in some cases. The flipside of this contrarian coin is that building a brand new highway often fails to alleviate the congestion that inspired its construction in the first place. Economists Gilles Duranton and Matthew Turner of the University of Toronto offer an impressive and empirical explanation for this concept in an upcoming issue of the American Economic Review (full paper via Google Docs).
Duranton and Turner analyzed loads of data on traffic, infrastructure, and travel behavior from metropolitan regions across the United States and found that “vehicle-kilometers traveled … increases proportionately to roadway lane kilometers for interstate highways.” For those who don’t care for either academic abstracts or the metric system, the authors then parse their conclusion in pithier terms: “roads cause traffic.” The basis for this confusing reality, write Duranton and Turner, is a three-pronged “fundamental law of highway congestion” that explains why road construction can never keep pace with road congestion:
people drive more when the stock of roads in their city increases; commercial driving and trucking increase with a city’s stock of roads; people migrate to cities which are relatively well provided with roads.
As compelling as the evidence in this study may be, it doesn’t tell perceptive urban planners anything they haven’t known for decades. Duranton and Turner trace the “fundamental law of highway congestion” to a study published by road researcher Anthony Downs in 1962. In an issue of Traffic Quarterly, Downs reported his law of “peak-hour” congestion: “This Law states that on urban commuter expressways, peak-hour traffic congestion rises to meet maximum capacity.”
Before Downs, social critic Lewis Mumford recognized the properties of this law in the behavior of drivers in and around New York City. Mumford was well ahead of his contemporaries when it came to recognizing the perils of road construction; Jane Jacobs is often remembered as the foil of road-builder Robert Moses, but at the time many would have cast Mumford in that role. While researching my history of transportation in the Northeast, I came across an illuminating four-part series Mumford wrote for the New Yorker in 1955 called “The Roaring Traffic’s Boom.” In the second article, Mumford provides a commentary that anticipates Duranton and Turner half a century ago:
[O]ur one-eyed specialists continue to concoct grandiose plans for highway development, as if motor transportation existed in a social vacuum. … Instead of curing congestion, they widen chaos. …
All the current plans for dealing with congestion are based on the assumption that it is a matter of highway engineering, not of comprehensive city and regional planning, and that the private motorcar has priority over every other means of transportation, no matter how expensive it is in comparison with public transportation, or how devastating its by-products.
Prescient as he was, Mumford had little impact on the policies of his day, partly because his suggested solutions — carefully planned communities, population limits, and a fully balanced transportation network — were even more objectionable than curtailing road construction. (And you thought raising the gas tax was politically unpalatable.)
Likewise the findings of Duranton and Turner are the type that should, but probably will not, inform good social policy. This is not a blatant call for more public transportation funding; in fact, the Toronto researchers found that adding transit does nothing to ease highway congestion. When one driver leaves the road, another simply takes his or her place, as Turner explained to Streetsblog DC. (That finding will be misappropriated by road-building enthusiasts to argue for more highways under the fundamental “if nothing helps, then who cares?” error.)
The only solution endorsed by Duranton and Turner is congestion pricing. In April 1955, Mumford predicted that Manhattan would soon have to “banish private wheeled traffic” in midtown during the day. This extreme version of congestion control did not come soon, and when Mayor Bloomberg did introduce a pricing plan recently, it was roundly beaten. “People, it seems, find it hard to believe that the cure for congestion is not more facilities for congestion,” Mumford wrote. Some beliefs are hard to change.
Image: via Air Experts of Michigan
Eric Jaffe is on Twitter
This entry was posted on Monday, June 6th, 2011 at 10:30 am and is filed under Automobiles, U.S. Infrastructure. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.
21 Responses to “Why Building Roads Creates Traffic”
- Ora Martindale Says:
June 6th, 2011 at 10:53 amIn the first sentence after the New Yorker article quote, you have “Prescient as he was, Moses had little impact…” I think you meant “Prescient as he was, Mumford had little impact…”
- Eric F Says:
June 6th, 2011 at 11:05 am“People drive more when there are more roads to drive on, commercial driving and trucking increases with the number of roads, and, to a lesser extent, people migrate to areas with lots of roads. Given that new capacity just increases driving, they find that “a new lane kilometer of roadway diverts little traffic from other roads.”
Shorter: Roads enable people to do stuff, both commercial and social, and when people have the opportunity to do stuff they use it.
It’s a very cramped view of the world that would lead one to take on clamping down on mobility as an avocation.
- Eric Fischer Says:
June 6th, 2011 at 11:18 amYou mean “Mumford had little impact on the policies of his day,” not “Moses had little impact on the policies of his day,” right?
- BrianTH Says:
June 6th, 2011 at 11:37 amAlthough congestion pricing is the correct answer to the problem, transit investment is part of making sure that congestion pricing doesn’t have regressive effects.
- BrianTH Says:
June 6th, 2011 at 11:38 amEric F,
Congestion isn’t good for mobility.
- Omri Says:
June 6th, 2011 at 11:41 am“Shorter: Roads enable people to do stuff, both commercial and social, and when people have the opportunity to do stuff they use it. ”
Which means that if the state wants to widen the road through my neighborhood, telling me it will improve traffic and therefore benefit me, I should take it with a salt lick. Because once they’re done widening it, I now have a major loss to my property values.
- Eric Jaffe Says:
June 6th, 2011 at 12:19 pmThanks for the keen eye, @Ora Martindale and @Eric Fischer. That sentence has been changed.
- Rick Risemberg Says:
June 6th, 2011 at 12:38 pmBuilding effective (instead of perfunctory let-’em-eat-crap) transit provides the mobility necessary for civic and commercial engagement at far less cost to the environmenta, human culture, and municipal treasuries than does supporting cars. Driving nowhere pays its way through gas and car taxes and fees, but is as subsidized as transit, while taking vast areas of land out of productive uses–land that would be providing homes, or services, or business opportunities otherwise–and paying taxes. (see for example http://tinyurl.com/2dmawlr )
Bicycling also provides mobility at low financial and social cost, and stongly supports retail activity. (Example: http://tinyurl.com/3pumoum )
Road building creates congestion because of the perception that it is free of costs to the user–as it partially is, thanks to huge socialization of tis costs!–and so is eternally self-defeating.
Plus it cripples civic culture through the inevitable isolating of citizens in publicly-subsidized arrogation of public space.
- K Nichols Says:
June 6th, 2011 at 12:44 pmAt some point though, the induced demand argument must reach a limit, and further increases in capacity will not induce sufficient demand to “max” it out.
I think a better rephrasing of the induced demand concept would be that *incremental* increases in capacity will produce an equal increase in demand. However, if you go crazy, and turn the Eisenhower expressway in Chicago into a 400 lane monster, it would be hard to conceive of how demand could ever reach that level of capacity.
For a practical example, look at a city like Buffalo. Buffalo’s highway system was conceived at a time when regional planners expected the population to be roughly double what it eventually stabilized at. Today, you’d be hard pressed to find a traffic jam in Buffalo. Clearly the highway system was so excessively overbuilt that it overcame even induced demand.
I know this doesn’t mean much in a world of limited resources, but it’s important to remember that induced demand isn’t an iron-clad law. At some point, you can indeed out build demand.
- BBnet3000 Says:
June 6th, 2011 at 12:58 pm“It’s a very cramped view of the world that would lead one to take on clamping down on mobility as an avocation.”
Im of the opinion that we should encourage mobility for people and goods rather than cars.
- Eric F Says:
June 6th, 2011 at 1:35 pmThey found that when they increased the size of Giants Stadium, it was just as crowded there as when the stadium had less capacity! One could argue that this means that the Giants should play in a 1,000 seat stadium, since simply adding seats will do nothing about the ‘congestion problem’. I’d take it to mean, gosh there are an awful lot of people who want the experience of attending a Giants game, and the larger stadium enabled some of that pent up demand to be met.
- Eric F Says:
June 6th, 2011 at 1:45 pm“Im of the opinion that we should encourage mobility for people and goods rather than cars.”
Because in your world cars are driven by replicants.
- Paulus Magnus Says:
June 6th, 2011 at 2:42 pmBecause in your world cars are driven by replicants.
Only in the off world colonies.
- BrianTH Says:
June 6th, 2011 at 3:27 pmEric F.,
I don’t understand your stadium analogy.
Traffic congestion occurs when too many vehicles are trying to use the same roadway at the same time. It slows everyone down and causes economic loss, increased fuel consumption, increased health problems, and so on. Traffic congestion sucks, and it isn’t good for mobility.
Your analogy to a pro sports stadium is strange because of course a pro sports stadium typically limits the number of people it will admit, and rations those admissions through a price mechanism. To draw an analogy to a highway that was experiencing regular congestion and had no congestion pricing, you would have to imagine a pro sports stadium that was free, and let in everyone who wanted to try to pack into the stadium. That would suck, and no team would think that was a good way to run a stadium.
So why do you think that is a good way to run highways? Just like with a pro sports stadium, during peak demand it will often make sense to only give out as many admissions tickets as there are seats, and to ration those tickets through a price mechanism. As applied to highways, that is called congestion pricing.
- Eric F Says:
June 6th, 2011 at 3:42 pmI agree that it’s hardly a perfect analogy. The point I’m trying to make is a narrow one. Greater capacity allows more access, even if the access route is slow. NJ just added a lane on a highway that is, I think, a pretty important weekend summer route for beach traffic. If the prior configuration could handle 100,000 vehicles (I’m just making up numbers for the example), and the new one can handle 150,000, well even if it’s slow going on Sunday night heading back home, the fact is that 50% greater access to the beach has been created. Even if no one moves a single mph faster, you’ve added 50% to a key recreational travel route. I don’t buy the idea that adding capacity is a futile way of fighting congestion. It’s an imperfect way, but it’s not like U.S. cities have been on a capcity binge for the last 40 years, so it’s a method that needs to be duted off. But even if you accept the idea that adding capacity does not speed traffic flow, the capacity nevertheless expands access to job markets and the rest which has value whether or not the added capacity is free-flowing.
- BrianTH Says:
June 6th, 2011 at 4:02 pmEric F,
You can’t just wish away the congestion problem by refusing to “buy it”. It is predictable that failing to congestion-price highways in high potential-demand corridors will lead to regular congestion, and empirically that is what happens.
All that is actually distinct from the issue of whether additional capacity on a given highway is a good idea. It may or may not be, and answering that question in any given case will likely require a detailed study. But the point here is that solving the congestion problem–which is a real problem–typically can’t be solved with additional capacity. Which means once you are starting to hit regular congestion in a corridor, it is time to start introducing congestion pricing, even if you also conclude additional capacity would be a good idea.
And in fact your stadium analogy was quite useful. No one would think it is feasible to provide free, unlimited admission to popular events in stadiums. And the same goes for roads.
- Eric F Says:
June 6th, 2011 at 4:13 pm“It is predictable that failing to congestion-price highways in high potential-demand corridors will lead to regular congestion, and empirically that is what happens.”
I don’t have a problem with using market mechanisms to price capacity. But from my standpoint, I see a society that during my lifetime has undergone a massive population boom and seen a huge increase in mobility demand. In response, at least in my neck of the woods, the added capacity in response has been zilch. Nothing at all.
I don’t doubt that one can induce travel by adding capacity, but in the current framework induced demand is the least of our problems. Under current conditions, drivers already pay massive tolls, high auto ownership and fuel costs and still sit in multi-hour back ups on weekend night heading home from social plans. Though pricing could knock some of those people off the roads, what is the point of further taxing a grandparent in Syossett trying to see her grandkids in Fairfield? What is the societal benefit in axing that trip compared with the benefit of adding some connections and lane miles to allow the trip to happen?
- BrianTH Says:
June 6th, 2011 at 4:53 pmThe point of congestion pricing, and by extension pricing of the entitlement to use the land occupied by the roadway in question, is to make sure the roadway/land in question is being used as efficiently as possible. Again, whether or not additional land and materials should be devoted to roadways in a given corridor is a different issue from how to make sure the roadways we do have are being used efficiently.
If under those conditions–the roadway/land in question has been priced for efficient use–a particular grandparent cannot afford to see her grandkids, you might or might not want to address that problem by subsidizing that grandparent’s travel costs. But it doesn’t make sense to subject the entire system to inefficient use as a way of addressing such concerns. And it certainly doesn’t make sense to do so in the name of “mobility”–again, congestion is bad for mobility.
- Chad Says:
June 6th, 2011 at 6:57 pm“Shorter: Roads enable people to do stuff, both commercial and social, and when people have the opportunity to do stuff they use it.”
Wrong, Eric. Rather, the new roads make it slightly more favorable to build in the tenth ring of suburbs rather than re-build or build upwards downtown or in the inner rings. This spreads everything out even further, requiring us to drive farther to do the same things we were doing before.
It’s to the point now where things have sprawled so much that you have to drive to get across the street - often as much as a mile so that you can navigate the stoplights, parking lots, and turn-arounds.
- John F Says:
June 6th, 2011 at 7:44 pmMassive tolls? I’ve lived in Chicago, a 10 million person CSA, New Orleans, a 1.5 million person MSA and pensacola, a half million person MSA. Even in the Chicago, commuting 30 miles each way from the suburbs cost me $2.80 round trip. In the New Orleans, it was a $1.00 toll round trip if I had to cross the river. I pay nothing now for a 25 mile round trip commute. Hardly a crushing burden.
In all my life, there is one newly constructed road (southern extension of I-355) I’ve used to get me from one place to another. I used it 4 times a year to get from one interstate to another. It was clearly designed in anticipation of further sprawl and its 6 lanes are usually pretty barren (for now)
Congestion pricing could do two things. First, encourage those who aren’t particularly constrained by time to use more leisurely routes, and second, fund truly faster ways of getting from point A to point B (to include ensuring current routes are properly maintained).
My favorite commute, by far, was living in Chicago proper and commuting to Northwestern every day. I had the option of public transport or driving and I more often than not chose the former due to the extra study time being on a train provided. I rarely experienced congestion or delays with either choice, aside from construction. Why can’t it always be like this?
- Chad Says:
June 6th, 2011 at 10:11 pm“Massive tolls? I’ve lived in Chicago, a 10 million person CSA, New Orleans, a 1.5 million person MSA and pensacola, a half million person MSA. Even in the Chicago, commuting 30 miles each way from the suburbs cost me $2.80 round trip. In the New Orleans, it was a $1.00 toll round trip if I had to cross the river. I pay nothing now for a 25 mile round trip commute. Hardly a crushing burden.”
Which only goes to show you to what lengths things are subsidized in the states. Try living in Japan, where a typical tollway might charge you $24 to drive 30 miles (eg, Kyoto-Osaka on the Hanshin). A one-way express train ticket is $4.50 to $5.00 and would get you there in about the same time.
Oh, and all four train lines between the two cities are owned by profitable private companies.
Post a comment:
Otro estudio mas que demuestra que ampliar la infraestructura vial como un remedio a los congestionamientos, es una inversion inutil y contraproducente.
No hay comentarios:
Publicar un comentario