Does Destroying Highways Solve Urban Traffic Congestion?

Aerial view of collapsed sections of the Cypress viaduct of Interstate Highway 880 in Oakland, California in 1989. (H.G. Wilshire, U.S. Geological Survey)

Strange how the traditional laws of supply and demand go out the window when it comes to traffic. Studies over the last decade (like this one, this one, and this one; plus the book Suburban Nation) have pretty much dismantled the theory that more roads equal less traffic congestion. It turns out that the opposite is often true: building more and wider highways can increase traffic congestion. If only people like Robert Moses and Le Corbusier had known this before their grand urban plans left our cities clogged with traffic, and carved up by ugly, value-destroying highways.

As part of its series on urban transportation, the Mother Nature Network has this recent post, which includes a nice rundown of the evidence against highways as congestion relievers, plus a discussion of the latest idea taking hold in urban traffic management circles: destroy highways to reduce congestion.

A particularly dramatic case in point comes to us from traffic-clogged Seoul, Korea, where a few years ago a handful of “crazy” visionaries in the transport department somehow managed to sell a new mayor on the demolition of an elevated downtown highway. Fast-forward to today: the highway’s gone, a formerly paved-over river has been rehabilitated, the resulting green space is a source of urban pride, and — wait for it — motor vehicle travel times have actually improved in the neighborhood of the old highway.

The MNN piece reminds us that highway tear-downs have had similar results in New York City and San Francisco, but that it took natural disasters for those to happen: New York’s West Side Highway collapsed under the weight of a cement truck in 1973, and San Francisco’s Embarcadero Freeway was removed after suffering damage from the Loma Prieta earthquake in 1989.

One final tidbit on the economic costs of traffic: according to the Texas Transportation Institute, traffic congestion costs us $87 billion a year in wasted fuel. And that’s not even counting all those hours lost (and road rage)!

Til Vater

Bookmarked :-)

Could I use parts of that article for my own Blog?
I would backlink you of course (:

best wishes from Germany!


Seems like other comments have picked this apart effectively. One larger point is that this information just adds to the notion that a "reduction in congestion" just is not the standard by which new highways should be judged. How about "reduction in travel time." Two congested highways could very well dramatically reduce travel time over one. Even better, the extent to which more highways increases quality of life by allowing people greater choice in their living and workplaces. No effort seems to be made to establish the extent to which "urban pride" might have been offset by untold thousands of Seoul residents accepting a less desirable job than they otherwise would have with the transportation. This would be hard to measure, but I'll bet the clever Freakonomics folks could think of a way!


But quality of life is such a subjective thing. And how about the quality of life of those who live near the freeway (and experience noise, fumes, etc) but would seldom if ever drive on it?

Marty Nemko

You ignore the costs of not building freeways, for example, people being coerced to not go where they want. When, for example, the Embarcadero Freeway in San Francisco was torn down and not replaced, countless people, rather than facing the gridlock on the city streets that was caused by the lack of freeway, simply stayed home. You economists are supposed to be expert at considering side effects. Loss of freedom is a serious side effect. Too often, economists' political biases, like so many social scientists', interfere with the fair-minded, not-so-dismal science that economics is.


Yep, roads solve nothing.

Readily available, fast TRANSIT solves a lot.

It's just too bad "shut down roads, traffic goes away" has become a virtual meme in some environmental circles without including the qualifier above.

For example, advocates for tearing out one of only two through highways in Seattle keep pointing to the Embarcadero example in SF, completely neglecting to mention SF's extensive trolley system and BART, neither of which we have in Seattle.


This smells too Keynsean to me. You give no explanation of why this happens. You know, we'd have no traffic congestian at all if we nuked the entire surface of the Earth.


The thing is, if you are an actual resident of a large American city, the highways that cut through it don't benefit you that much. They are mainly for the benefit of people who live in suburbs and commute long distances to work. If highways are torn down and the land is reclaimed by people who actually make the city their home all the better.


The Constructal Law ( has something to say about this sort of thing.

Al V.

Makes sense to me. NYC has closed and narrowed streets in Manhattan, putting in bus and bike lanes, and traffic is no worse. It may be better.


What's the opportunity cost of not being able to drive into the area? Drivers bear the $87 Billion cost in wasted fuel, plus the non-wasted fuel and the marginal time. There's a thesis in there somewhere.


I would expect less traffic on nearby roads. The nearby roads provided access to the main artery that does not exist anymore. Same argument works for rail transportation, and all transportation corridors. Remove the transportation corridor and traffic on its access roads disappears. No mystery here. The existing traffic goes somewhere else.

The whole argument seems empty. Decreasing hospital beds does not decrease the requirements for hospitalization. Digging more graves does not cause people to die.

Generally speaking light cars and trucks increase faster than the population and people have more travel requirements. The research has a serious hole in not addressing the vehicle miles for the entire urban area.


human behavior cannot be quantified so there is no mathematical solution to this phenomenon of traffic congestion.