Keep Austin Houston?

Why “induced demand” is a tough nut to crack — and whether Texas’s capital city is destined to sprawl
Subscriber Only
Sign in or Subscribe Now for audio version

If You Build It, Will They Come?,” the feature essay of our Spring 2024 issue, generated a lively response from readers and public commenters. Editor Ari Schulman spoke with Washington Examiner policy editor Joseph Lawler to learn more about “induced demand” and the choices facing Austin and other booming cities.

Ari Schulman: The induced demand critique has been around for a long time. What was missing from the public conversation about it that you wanted to address in your article?

Joseph Lawler: The conversation is changing and becoming more politically charged because of the rise of the new movement to stop highway expansions and even remove downtown highways. That group of activists has successfully wielded the concept of induced demand as a talking point against highway expansions and brought it onto the radar of the broader public, where it otherwise might have been consigned to academic journals. In response, transportation agencies and other proponents of expanding highways have, at times, been sort of backed into saying that there is no such phenomenon as induced demand, or that the idea is overblown or misapplied.

The country is headed toward a series of battles over highways, and this article, hopefully, will help readers understand one of the important features of the battlefield landscape.

Schulman: For people who do not have economist-brain, what experience can you draw on to make sense of induced demand? Does it mean that there is virtually no limit to how many trips people would take by car, if they could do so without travel times slowing? I’ve also wondered if you could observe the effect in reverse — say, when an interstate bridge collapses, like in Minneapolis in 2007 or Baltimore this year. If people are highly sensitive to travel times, the result should be fewer trips taken, not more congestion, right?

Lawler: One analogy that might be helpful is concert ticket sales. Popular artists usually make fans wait in some sort of virtual line for tickets. They could auction the tickets off to the highest bidder, but that would alienate their most hardcore fans. So they ration tickets via queuing rather than by price.

When huge acts have free shows in easily accessible areas, huge numbers will show up. For example, supposedly 1 million fans attended a Garth Brooks concert in Central Park. Obviously, demand for the show was not unlimited. But probably almost everyone who was a Garth fan in New York City showed up. But if you make fans wait in line for the opportunity to buy tickets for the event, only a small fraction of that number bothers to go through with it.

The cost of the ticket is part of limiting the quantity demanded. But so is making fans jump through hoops by waiting for the tickets to drop online or signing up for the fan club. Congestion can be thought of analogously, as a situation where demand is managed largely through queueing and only rarely by pricing (in the form of tolls).

As to the second part of your question: Certainly, the highway critics are willing to bet that removing certain downtown freeways won’t lead to worse congestion. Some cities have permanently removed highways or parts of them, including Portland and San Francisco.

Yes, in theory, the outcome would be unchanged congestion but fewer trips. Measuring that, though, would be difficult.

Schulman: The studies you cited all found that highway elasticity is around 1, meaning that if you expand highway-lane miles by, say, 10 percent, then highway trips taken will also increase by about 10 percent, entirely consuming the new supply. But there are competing studies. Usually a good place to start with research is a literature review, and there is one from 2018, for example, that found an elasticity of 0.2, although I know that was looking at the effects of expanding entire road networks rather than individual city highways. Why did you pick the studies you did over the other ones available? And how certain is the research generally?

Lawler: In writing this story, I was looking for studies that met two criteria: First, that they examined the question of interest, meaning what happens to congestion in cities when highways are expanded. So we excluded studies that looked at rural roads, for instance, where I don’t think anyone is really worried about congestion. Second, that they had a design that allowed for causal inference — that is, that they attempted to establish not just correlation between highway construction and congestion, but causation. Those two qualifications narrow it down significantly.

The studies we’re left with are impressive, but they’re limited at the same time. They provide only a partial answer to the question of what happens to traffic patterns when highways are expanded. There are a million possibilities that we won’t have the data to explore. For instance, it’s quite plausible that a highway expansion might not reduce peak congestion, yet at the same time might make traffic much lighter, and life much more pleasant, on side roads.

In the real world, that would be huge. But, the existing studies don’t have that granularity of data, because they involve aggregated data about vehicle-miles traveled per mile of highway lanes. Maybe in the future, with ubiquitous GPS and phones, we’ll see an analysis that does get into that level of detail.

Schulman: It’s interesting that just about everyone writing on induced demand treats it as an anti-highway idea, regardless of what side of the debate they’re on. “Expanded roads just fill up with new traffic so why should we bother?,” as highway advocate Tory Gattis summarizes the idea. Does your article echo that equation?

Lawler: Yes, one of the interesting findings from my reporting was that both sides of the fight over the expansion of I-35 in Austin led their arguments with claims about how the project would affect congestion, even though the underlying conflict is about urban form rather than traffic.

Congestion is a real consideration. I’ve talked to people who’ve moved from one city to another because they couldn’t stand the congestion. So I wouldn’t want to dismiss those experiences. But it’s just one of many facets of a city that make it more or less desirable.

What I heard from the highway opponents in Austin is that the real concern is not congestion, but rather that the metropolitan area will continue to sprawl further out into the Hill Country, shunting people into detached single-family homes and car-dependent lifestyles. They want to see a shift toward a more classic urban form and the construction of more walkable neighborhoods.

By the same token, apart from the Texas Department of Transportation itself, there was not much expectation on the part of highway proponents that the expansion would lessen congestion. They saw it more as accommodating the commutes and shipments that they know will be needed, based on the continued growth of the metro area.

For the vast majority of people, though, congestion has more visceral appeal, because they hate being stuck in traffic. So it shouldn’t be a big surprise that that’s where the rhetoric is aimed.

Schulman: But it sounds like you’re also pushing back on this equation. Yes, induced demand is real — but that doesn’t mean that highway expansion is pointless. I’m thinking about the Baltimore bridge collapse again. You could look at the collapse and say “No highway value was really lost, because people’s behavior will eventually adjust such that congestion will go back to where it was.” But that would be nuts. And actually, all the major players you reported on in the I-35 debate agree that it’s inherently valuable when people can take more trips, which highway expansion does provide. So maybe induced demand shouldn’t take so much focus. Right?

Lawler: We have to separate two things. On the one hand, we have the narrow concept of “induced demand” as a movement along the demand curve for highway travel. On the other hand, we have the critique of “induced demand” as meaning the entire pattern of urban development that is created by the model of subsidized highway construction and tightly regulated land use.

It’s confusing that the term has taken on two meanings. I think that understanding both would be helpful to anyone interested in this debate.

We don’t want to get spun around to the extent that we think bridges and highways are bad. But nor would it be good to have the debate collapse to the extent that congestion is the only consideration.

Schulman: What’s the problem with urban sprawl, exactly? Joel Kotkin points out, in a response to your piece in this issue, that dense downtowns just aren’t as relevant to how Americans actually work and live now as they were many decades ago. This is a common refrain of urbanism’s critics: sprawl actually means space to raise families, recreation, newly built town centers that are easier to get to, and ever more jobs. In a word: Why not Houston?

Lawler: Right, as we explored in the essay, sprawl is partly a function of better transportation options. Some amount of sprawl is just a reflection of the fact that we have cars, and thus is a marker of progress, in a sense. If we get flying cars or hyperloops, we can expect to see metropolitan areas getting even more spread out.

What I’ve heard from urbanists, though, is that the phenomenon Kotkin identifies is the exact assumption they’re trying to challenge. They don’t see it as a purely market-driven development, but rather the result of policy choices, namely subsidized highways combined with restrictive land use policies that steer people into living further out from downtowns. Again, not all of this is the result of government policies. But to the extent it is, they want to push back.

I would also be remiss if I didn’t mention that much of this critique is an environmentalist one. In general, dense cities are thought to be greener in a number of ways. The premise that city living is and always will be better for the climate and more sustainable is a contestable one, but certainly it is one you’ll encounter quickly if you engage with highway critics.

Schulman: One difficulty you must face in writing about this issue is that readers will take you as tacitly advocating for one of the sides you’re reporting on. Well, let’s be more candid: urbanism produces a lot of suspicion, and so does reporting on debates about urbanism. I know some readers responded to your article this way. How do you think about that problem?

Lawler: Does urbanism produce suspicion? I’m not sure about that. I suspect it’s off the radar for most people. If urbanism is about building more places like Georgetown in Washington, D.C., or the North End in Boston, it seems like the vast majority of people love those places and would be open to it.

Anti-car sentiment, though, definitely arouses suspicion. Relatedly, we’ve seen clearly for the past few years that the public is extraordinarily sensitive to gas prices. Most people are alarmed at the idea of their mobility being limited.

I was glad of the assignment because it gave me an opportunity to talk to a wide range of people with a variety of opinions on highways, urbanism, land use, and much more. It’s freeing to review all the facts and arguments and then try to present them as clearly as possible to readers. Yes, there’s a culture-war aspect to this, but this publication has covered much hotter cultural battles than this for years.

Schulman: Tell me more about the Seattle path for Austin. If this means developing more places like the delightful, walkable South Congress neighborhood, it sounds great. I grew up there, next door to immigrants who brought us fish they’d caught from the river. But instead of Weird Austin favorite Lucy in Disguise, SoCo now has a Hermès. I wonder if neighborhoods built for working families inside the city limits are just a thing of the past. And car infrastructure is already so baked into the city that it’s hard to imagine that most people won’t still need, or demand, a car, even with more density and transit. What is a realistic vision for this change?

Lawler: Seattle was invoked as a model for Austin independently by several highway critics I spoke to. They see Seattle as an example of a city that is doing some things at the margins — at the right margins — to shift the metropolitan area away from further sprawl. Seattle seems to have succeeded in getting some share of workers to commute via transit rather than by car, and to have densified the city a bit.

Getting to a level of density that would be required to get people to give up their cars en masse would not be a marginal change. That would be an absolute sea change.

Yet it is realistic to expect that Austin and its inner suburbs will get denser in the years ahead. In fact, it’s been happening already — keep in mind that Austin has done very well, relatively speaking, in terms of growing in population and remaining somewhat affordable. The government is looking to go further in that direction.

It was very interesting to me that Paige Ellis, the former mayor pro tem I quoted in the story, said that the city was 10–20 years behind in terms of coming up with the tools that would be needed to accommodate the expected growth of Austin in an urbanist way.

Realistically, Austin is not going to provide for hundreds of thousands of additional daily commutes by foot, bicycle, and transit in the next decade. In theory, it could — it’s been done before, in other cities. But in reality, it would require a total overhaul of land use policies and permitting, the construction of many new railways, bus lanes, bike lanes, bridges, etc. We’ve lost the ability to allow things to be built that fast in the U.S.

But it is easy to imagine that some share of those commutes could take place within the urban core, rather than via highway from the Hill Country.

Schulman: A question we often push New Atlantis writers to answer is “If you’re right, then who should do what differently?” That question feels unfair when we have someone writing about a deeply stuck policy debate — but that’s also when it’s most needed. One of your conclusions is that the “Houston or Seattle?” choice is actually not one that anyone has the governance power to decide for Austin right now. If that’s so, then who should do what differently?

Lawler: One thing we’re seeing is the increasing organization of highway opponents. As we’ve seen in Austin, they are not yet at a point where they can stop projects, but they’re increasingly driving serious resistance to them and gaining allies at different levels of government.

So far, they tend to come across or be derided as left-wing, although they regret that perception. It will be very interesting to watch whether, in the years ahead, their set of proposals becomes a partisan proposition at the national level. That could help them achieve legislative and regulatory victories via Democratic Party successes. But it could also engender a partisan backlash from Republicans.

Yet on balance, my guess is that it is helpful to have people thinking about these policies and for the public to force officeholders to engage. The set of institutions that shape our cities — starting with the Department of Housing and Urban Development and going on down to the New York Metropolitan Transit Authority and I-35 itself — were largely put in place in the 1960s. There are new problems to be addressed, but our institutions, policies, and norms have not evolved. Instead, they’ve become ossified. They’re unresponsive — not because the politicians and planners of yesteryear got it right, but because those of today don’t care.

Keep reading our Summer 2024 issue

The Amish on AI  •  Why UFOs  •  Facts vs. us  •  EA as self-help  •  Subscribe

Subscribe today for as low as $24/year

Subscribers receive new issues weeks to months before articles are posted online.


See Subscription Options