Study: Building Roads to Cure Congestion Is an Exercise in Futility

We hear it all the time: The road lobby insists that the only way to reduce mind-numbing traffic congestion on the roads they built is to build new roads. Federal funding gives huge blank checks to state DOTs, which tend to prioritize road building over transit, bridge maintenance or anything else. But mounting evidence suggests that building new roads won’t do anything to alleviate congestion.

In a paper to be published soon in the American Economic Review, two University of Toronto professors have added to the body of evidence showing that highway and road expansion increases traffic by increasing demand. On the flip side, they show that transit expansion doesn’t help cure congestion either.

We’ll spare you the calculus in the report. Here’s the upshot: “Roads cause traffic.”

Duranton and Turner: If you build it, you will sit in traffic on it. Photo: ## and the Environment##

Professors Gilles Duranton and Matthew Turner analyzed travel data from hundreds of metro areas in the U.S., resulting in what they call the most comprehensive dataset  ever assembled on the traffic impacts of road construction. They write:

For interstate highways in metropolitan areas we find that VKT [vehicle kilometers traveled] increases one for one with interstate highways, confirming the “fundamental law of highway congestion” suggested by Anthony Downs (1962; 1992). We also uncover suggestive evidence that this law may extend beyond interstate highways to a broad class of major urban roads, a “fundamental law of road congestion”. These results suggest that increased provision of interstate highways and major urban roads is unlikely to relieve congestion of these roads.

Duranton and Turner say building more roads results in more driving for a number of reasons: 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.”

Given the huge amount of time consumed by driving (the average American household spent nearly three hours per day in a car in 2001), the authors note that “the costs of congestion are large.” Considering the economic value of time spent doing anything but sitting in bumper-to-bumper traffic, that becomes an economic problem of the first order.

“Transportation accounts for about one dollar in five that Americans spend,” Turner said in an interview with Streetsblog. “The interstate highway system eats up on the order of two dollars of every $100 of every market transaction in the United States. That’s a huge part of the economy and a huge part of people’s lives. Understanding how that works is really important; you don’t want to make mistakes on something that important. You don’t want to build roads and have them not deliver the effects that you expect them to.”

The implications for this research are significant, especially as Congress considers whether to integrate performance measures into federal transportation spending decisions. These findings make a strong case that Congress should not allocate too many scarce resources to road expansion when that’s not a real solution for congestion.

Duranton and Turner say that metropolitan areas tend to get new roads regardless of whether or not the prevailing level of traffic warrants expansion. They urge the establishment of transportation policies based on their findings and the data they compiled, rather than the “claims of advocacy groups”:

Unfortunately, there is currently little empirical basis for accepting or rejecting the claims by the American Road and Transportation Builders Association that “adding highway capacity is key to helping to reduce traffic congestion”, or of the American Public Transit Association that without new investment in public transit, highways will become so congested that they “will no longer work”. Our results do not support either of these claims.

They didn’t find that transit reduces congestion. But that doesn’t mean that metro areas shouldn’t build transit as a way to maximize the efficiency of their transportation networks, they say. Turner said transit is a good way to get more “person-miles” out of roads. But more buses and trains won’t reduce congestion, he added, because regardless of how many drivers switch to transit, other drivers will fill the vacuum.

“If you think about the result that we’re finding for roads – if you add a little bit of capacity, someone uses it, right?” Turner said. “So there are all these people out there waiting to take trips as soon as there’s space on the roads. So if somebody stays home, or if you add capacity to the road, there’s somebody there waiting to use that space. Well you should expect the same thing to happen if somebody gets out of their car and gets on the bus, it’s bringing up a little bit more room on the roads, and there’s somebody out there waiting to use it.”

Still, Turner says transit plays a vital role in maximizing the value of our transportation networks. “Transportation infrastructure is just so expensive,” he said. It’s important to use it efficiently.

The researchers didn’t discern between light rail, commuter rail, and buses. Turner said he feels that buses allow cities to move just as many people with a much cheaper infrastructure network, but there are passionate arguments on both sides of the bus vs. rail debate, and the authors don’t choose one over the other in their paper. In fact, they only have one significant policy recommendation:

These findings suggest that both road capacity expansions and extensions to public transit are not appropriate policies with which to combat traffic congestion. This leaves congestion pricing as the main candidate tool to curb traffic congestion.

“The menu of policy responses to congestion is not really that long,” Turner said in our interview. “You’ve got building more roads, building more transit, and congestion pricing, and if you’d like you can put smart growth on there. We looked at two of those really carefully and found that they didn’t perform as advertised. So if you’re thinking about these things purely as responses to congestion, it doesn’t look like they work. There is some evidence that congestion taxes work. So if you were going to pick one of these things to go for, that would be it.”

They’re working on research now to investigate the impacts of smart growth on traffic.

  • Cindy

    Thatcher was Reagan in drag.

    “Widening roads to relieve congestion is like loosening your belt to relieve obesity.” –Walter Kulash

  • Nat

    The world is not so simple so as to yield short and sweet conclusions to every malady or imbalance.

    It’s always complicated. However, we have nearly 100 years of land development to draw upon and glean from: the cost and price of fuel, car ownership, public and private land, government priorities, macroeconomic policy, just to name a few.

  • Mplotz

    I used to believe that. Then gas prices rose and congestion dropped.

    Yes, these are complex systems, but it is amazing how manipulating that single variable–fuel prices–caused VMT to drop, and cities/regions to reconsider their land use and transportation planning.

  • Josef Bray-Ali

    “The menu of policy responses to congestion is not really that long,” Turner said in our interview. “You’ve got building more roads, building more transit, and congestion pricing, and if you’d like you can put smart growth on there. We looked at two of those really carefully and found that they didn’t perform as advertised. So if you’re thinking about these things purely as responses to congestion, it doesn’t look like they work. There is some evidence that congestion taxes work. So if you were going to pick one of these things to go for, that would be it.”

    They forgot one key policy response: tear that shit down. Tearing out highways reverses the inducement to drive. Closing off lane miles via road diets to cars does the same thing.  

  • Anonymous


    The spike in gas in 2008 and now in 2011 also coincided with a huge recession, meaning fewer people had a jobs to go to or money to burn at the mall. So it’s not quite so clean. But yes, fuel prices certainly seemed to at least shift people on to transit in a meaningful way.

  • Kent Strumpell

    I believe road congestion is a fact of life in any reasonably prosperous place.  As this study confirms, we’re pretty much stuck with it, barring taxes on driving or severe economic decline.  I think the most viable response is to give people better alternatives so they can opt out of congestion: walkable communities with a rich mix of uses and desirable destinations; frequent, convenient transit service; safe, inviting bikeways, etc.  By all means, stop expanding the problem and, as Josef suggests, let’s begin convert our roads to better uses, as in the proposal for Figueroa.

  • saw

    The “fundamental law of highway congestion” is just an example of the Jevons effect: “technological progress that increases the efficiency with which a resource is used tends to increase (rather than decrease) the rate of consumption of that resource.”  Recycling, energy efficiency, building roads, etc, can only reduce consumption if coupled with user taxes like congestion pricing to counter the subsequent decrease in resource costs.

  • what is the name of the Journal entry please?

  • jackal

    As one of the comments suggested — start building communities so that they are more user friendly.  In California, for example, suburbs have been, and are being built without the basics for food, shopping or entertainment, forcing everybody into cars since the nearest malls are 15 miles away.  One could bike there, but you would have to do that once a day to manage the groceries your car will carry in one trip that would serve you for a week.  Besides, 50%+ of Americans are so obese they couldn’t ride a bike anyway.

  • David

    The mall dynamic is one of many long term solvable issues here. Local street level commerce needs to be encouraged by cities, walking from your house or condominium to shop needs to replace the mall. It must become an expensive shopping option. Free home delivery could also encourage small vehicle or walking access. The common wall life style is becoming sought after…affordable, convenoient.

  • Terry

    While I have not seen the article, the analysis appears flawed because it apparently does not consider land use issues.  Abundant research and public results (see the Arlington Rosslyn-Ballston corridor) shows that congestion can be reduced on urban streets if urban development (ie–transit-oriented development) and transportation policies (lots of public transit, limited and expensive parking, other TDM measures) support it.

  • I’ve been saying this for 14 years.

  • David: transit and smart growth doesn’t reduce congestion on roads. Instead, it significantly multiplies the efficiency of people movement and allows for much higher density of development. However, even if a higher percentage of people live without cars in that area, the population there booms. So congestion still exists. Look at Manhatten.

    That’s not a bad thing. Look at how efficient Manhatten is and how many people they have there. Without the subway, it wouldn’t support that. Nevertheless, it is still congested. 🙂