RODOVIARISMO: Estradas representam entre 20 e 25% de todas as áreas urbanizadas nos EUA

Can Americans Just Stop Building New Highways?

Freeway nation: Traffic on the 110 freeway near downtown Los Angeles, California, US, on Saturday, April 12, 2025.

Photographer: Alex Welsh/Bloomberg

A new book argues that the expansion of the US roadway network has exacted social and environmental costs that far outweigh the benefits.

By David Zipper

“The Interstate Highway Act literally brought Americans closer together,” President Bill Clinton said in 1996, referencing the bill that launched the 47,000-mile federal highway network. “We were connected city-to-city, town-to-town, family-to-family, as we had never been before. That law did more to bring Americans together than any other law this century.”

In his new book, Overbuilt , Erick Guerra, a professor at the University of Pennsylvania Weitzman School of Design, offers a markedly less rosy assessment of the US highway system. By blasting their way through cities, Guerra argues, interstate designers sacrificed urban wealth and quality of life, particularly within low-income neighborhoods.

And the still-expanding network of highways feeding into the interstates has only exacerbated the pain. In a 2024 study , Guerra and his coauthors concluded that roadways now take up a fifth to a quarter of all urbanized land in the US — an amount of real estate as large as West Virginia. He writes:

Most national transportation problems stem from having too much road infrastructure, rather than too little. The desired increased economic activity or reduced congestion from building more highways or expanding existing ones rarely materialize. More than anything, more highways lead to more driving, which produces more pollution, more traffic fatalities, and more auto-centric cities and towns that require more driving to participate in basic civic, social, and economic activities.

CityLab contributor David Zipper spoke with Guerra about highways’ hidden costs, and what might be done to mitigate them. Their conversation has been edited for clarity and concision.

Much of your book concerns the interstate system, which many people believe President Dwight Eisenhower — a World War II general — designed to enhance military logistics. To what extent is that story accurate?

It’s a convenient narrative, but I don’t think that was a primary motivator for the system that was built. If the interstates were built purely for military purposes, they would have circumvented cities instead of going right through them. If you’re going through cities, you have congestion, which makes it harder to move large vehicles.

Credit: Island Press

But the highway engineers responsible for planning the interstate system recognized that that cities were where there was a demand for traffic. Those sections were going to fund the entire system through gas tax revenues.

Most European countries have limited-access highway networks that closely resemble the US interstate system. But in Europe, those highways generally stop at the urban periphery instead of going through cities. Why is that?

Yes, it’s much more common in the US for highways to barrel through urban neighborhoods. Ironically, a lot of the interstates’ design was modeled around the autobahn in Germany, but the autobahn intentionally avoided cities.

In the US, constructing interstates dovetailed with other things happening at the same time, like urban renewal, so you have an argument for wanting to clear urban neighborhoods. Some of that is local responses to the Great Migration, and some of it is about suburban flight.

In your book, you wrote: “The places that seem to be doing the best in terms of economic efficiency, traffic safety, social equity, and environmental sustainability are frequently the places with the fewest highways and major arterials per capita.” Can you offer some examples?

These are going to be regions with relatively dense central cities, high GDP, and few crash deaths per capita. Places like Boston, New York City and San Francisco have big urban highways, but there are relatively few of them on a per capita basis. Another way to say this is that a city like Houston with fewer highways would be a healthier and richer place.

Houston is an interesting example, because the Texas Department of Transportation is currently spending billions to widen I-45 there. Are you saying that project will be counterproductive?

Unequivocally, yes. That’s not to disparage Houston, which is a wonderful place. But it would be a much more wonderful place with fewer highways.

Let’s talk about induced demand . In your book, you explain that highway expansions generally fail to deliver promised congestion relief because the added lane space leads people to take new trips, drive instead of using transit, or travel at peak instead of off-peak times. Fair enough. But a wider highway doesn’t force anyone to change their travel choices. Presumably, those who subsequently adjust their behavior are doing so because it makes them better off. As a result, couldn’t a congested, wider highway still create societal benefits — since people are traveling in ways that are for them superior to their options before the expansion?

I don’t disagree at all. At the end of the day, “induced demand” is really just “demand”: people adjusting their behavior in response to changes in travel time. Those changes are worth something to people — they’re just not worth as much as highway expansions cost, particularly when you factor in the costs of land, public health and the environment.

So yes, there are benefits, but it’s still bad public policy. There are also benefits to getting rid of the liquor tax, but I’m not sure that’s good public policy, either.

Drivers travel under the Grand Parkway offramp into I-45 in Houston in 2023.Photographer: Brett Coomer/Houston Chronicle via Getty Images

Transit advocates often claim that investing in bus and train service will reduce congestion. Is that true?

We don’t have any evidence of that being the case over time or across places. If you think about it, the places that are most congested often have the best transit. If people switch to transit, it’s the exact same thing as induced demand from a highway widening — newly available roadway space leads people to adjust their driving behavior until the roadway fills up.

But I would say that investing in good transit allows people to get around congestion. It allows us to have dense, highly productive commercial centers. Transit has a lot of benefits; I just don’t think that reducing congestion is one of them.

I wonder if transit supporters talk about congestion mitigation to win a political argument. I’m thinking of the famous headline from The Onion that “98% of US Commuters Favor Public Transportation For Others,” and the subsequent research concluding that many people really do support transit because they hope other people will take it, allowing them to drive faster. Perhaps congestion arguments are necessary to obtain the political support that transit needs to survive.

Maybe, but I would much rather we reframed how we evaluate all types of transportation investments. I think we would do a lot better if we focused on things like access — how easy is it for people to get to work — instead of being focused solely on congestion.

Erick Guerra. Credit: Island Press

I like the concept of access too, but we live in a political world, and many voters understand that they want less congestion. Do they understand that they want more access — or is there a way to get them to understand that?

Access is definitely a more complicated idea than congestion. I think people try to get at access with ideas like the 15-minute city, but there has been quite a bit of backlash to that framing. But as a researcher, I don’t think we should be pushing something that’s ultimately untrue. At the end of the day, we’re better off being honest about the limits of mitigating congestion.

To what extent would better transit service reduce Americans’ driving?

There are individual transit investments that make sense, just like there are road investments that make sense. But from a national policy perspective, we will not solve the problem of overbuilding highways by building more transit.

Why is that?

Much of the country is so auto-oriented that even if you just invested billions of dollars in a brand-new light rail system, you’re just not fundamentally affecting accessibility enough. It’s still easier to go from door to door in your car.

So what should the US be doing, if not expanding transit?

The first thing to do is to stop building more infrastructure, especially highways. We’re already overbuilt, so don’t add to the problem. Eventually, we want to decommission roadways. Over time, people will drive shorter distances at lower speeds, which are safer. As things shift, you’ll have more people in cities, and more transit proposals that look good.

The idea of deconstructing highways en masse seems likely to be a political non-starter — and virulently opposed by the highway lobby. I wonder if it’s tilting at windmills.

It’s a difficult sell, but I do think there’s an appetite for stopping new highway building. And there are ways to decommission roads that can add to the tax base by creating land for development, not just shutting down a highway overnight. And some aspects of un-building roads are quite expensive, which could create jobs and business for roadbuilding interests.

In your book, you cited an academic paper by Zach Liscow and Leah Brooks that is quite popular within the abundance movement . Liscow and Brooks concluded that highway construction became significantly more expensive in the 1980s than in the 1960s, and they claim that “the rise of the citizen voice” is a big reason why. My question: If highway construction costs were to fall, how much better off would we be as a society?

Oh, I don’t think it would help at all. With the current incentives in place, if we have more money, we're going to build more highways.

Also, if you remove community engagement, the social and environmental costs of highways are probably going up. People say “community engagement,” but it's tearing people's houses down and giving them a dollar.

What do you think about getting the federal government out of highway funding entirely, and leaving construction and maintenance up to the states?

There is a decent argument supporting more decentralization like that. Some states would respond by raising their gas taxes, but some may switch entirely to a general fund model, which is how a lot of the world does it.

A newly repaved bridge crosses Interstate 66 in Vienna, Virginia, in 2021.Photographer: Andrew Caballero-Reynolds/AFP via Getty Images

You think funding transportation from general revenues would be better?

Yeah. Right now, with the Highway Trust Fund, you have a system that self-perpetuates. We raise money from driving in order to make it easier to drive — and there’s no reason to stop, ever. As a funding mechanism, it’s pretty flawed.

In your book, you wrote that “congestion pricing, gas tax alternatives, vehicle automation, and new transit technologies have a role to play in future urban transportation systems but will do little to nothing to resolve the fundamental problem of an overbuilt transportation system.”

That passage seems to dismiss approaches that many reformers are excited about, such as congestion pricing and Vehicle Miles Traveled taxes. Why?

Things like New York City’s congestion charging are a wonderful way to raise revenues and do things that are good for a city. But as a model for federal transportation policy, it just doesn’t get us there, because it doesn’t limit the amount of new infrastructure.

I do think a variation of congestion pricing is highly effective, if people are charged for the marginal cost of using the system — but that can’t just be in urban zones.

How would that be different from a VMT tax?

If we just generate revenue from the VMT tax and put it into roadways, that increases VMT, and we’re in the exact same cycle that we are now with the gas tax.

Understood, but let’s assume that revenues collected from road users go into the general fund, not a dedicated highway fund. In that case, would we be better off collecting revenue through a VMT tax that varies based on location, time of day, type of vehicle, et cetera?

Yes. If you could do a variable tax, that would be extremely effective. The premise would be that you can charge the most money to people who are making the most expensive trips. Right now we do the opposite of that. We basically collect all the money from everyone, and we spend it to make the most expensive trips easier, such as widening an extremely expensive urban expressway.

What do you think about tolls?

I love tolls. I think we’d be much better off if in 1939 the Bureau of Public Roads had planned to finance a national highway system through tolls, as they were instructed. Their report Toll Roads and Free Roads is a master class in civil servants taking over a political agenda.

Tolls ensure that the people who use a road are the people who are paying for it. They reduce the likelihood that we build things that people simply don't want.

Project 2025 , the Trump administration’s policy playbook, calls for more tolled roads. Does that suggest an opportunity for a grand bargain between Democrats and Republicans?

I would love for that to be the case. I think there is an opportunity to bring many more conservative people onto the side of not wanting to build and widen highways.

But one of the challenges right now is a lot of times when people are talking about tolls, they're talking about them as a way to add capacity. But I think tolls are most valuable as a way to limit capacity. Instead of building a new toll roadway, you toll an existing road so that you don't need to expand.

If you could change one thing about Congress’s next Surface Transportation Reauthorization, what would it be?

I would want some way to decouple the gas tax, or whatever the funding mechanism is, from what we build. That would be my biggest hope.

Fonte: Bloomberg