It's Not Always a Bad Idea to Expand a Highway

Critics of adding road capacity ignore its benefits while proposing solutions that won't fix traffic congestion.


The much-discussed bipartisan infrastructure bill might get a vote in the House of Representatives later this week, paving the way for $500 billion in new spending on a broad array of infrastructure projects and priorities, including $110 billion for roads and bridges.

The prospect of more federal highway spending is proving controversial with environmentalists and public transportation advocates, however. They argue that expanding existing roadways will only create more pollution while failing to address traffic congestion.

"Road expansion projects have failed to deliver the promised benefits. In fact, the evidence shows that they actually make traffic and pollution worse," said Ben Holland, the manager of the Urban Transformation Program at the Colorado-based environmental non-profit RMI, last Thursday to Reuters.

The phenomenon he's describing is called "induced demand." It is the idea that adding new lane miles only encourages people to drive more, thereby erasing any congestion relief the new asphalt was supposed to provide.

Last week, RMI, alongside environmentalist group the National Resource Defense Council and the public transportation advocacy outfit Transportation for America, released a calculator that lets users estimate how much induced demand would be created by adding lanes of roadway in a particular metro area.

This calculator is only the latest broadside against the concept of highway expansion as a means of relieving congestion. In September, Bloomberg CityLab published a long piece by David Zipper, a fellow at Harvard University's Kennedy School, criticizing the false promise of road widening in general and an ongoing project to widen I-35 in downtown Austin, Texas, in particular. Zipper argued that the realities of induced demand and requirements that (some) states spend all gas tax dollars on highways incentivize state departments of transportation to build endless freeways, despite the minimal effects that such building has on congestion.

This line of thinking is starting to get a hearing in Washington. A transportation bill pushed by House Democrats earlier this year, for example, would have made it harder for states to spend federal transportation dollars on roads. It also would have given cities, which are more transit-friendly, more say over how transportation dollars are spent.

Meanwhile, the Biden administration, while pushing for more spending on roads, has used some novel means to delay particular highway projects.

Back in April, the Federal Highway Administration asked that the Texas Department of Transportation pause soliciting contracts on a downtown highway widening in Houston because of civil rights concerns raised by local activists and politicians.

To be sure, the idea that building more roadways results in more car trips is not generally a controversial idea. There is, however, some heated disagreement about whether that extra auto travel is necessarily a bad thing. There is also debate about whether induced demand is really an inescapable "iron law" that makes every highway widening futile.

"Much of what happens when a freeway is widened is explained by latent demand—trips that were previously made on other roads or by other modes that shift to the expanded freeway when it becomes available because the freeway is now a better alternative," wrote Robert Poole, director of transportation policy at the Reason Foundation (the nonprofit that publishes this website), earlier this month.

Those extra highway trips, Poole has argued, shouldn't be considered an evil given that they involve people voluntarily changing their route from a nearby arterial road (or bus line) to the highway. There is a reason those people are making the switch and it is probably not because traveling on the now-expanded highway makes their commute worse.

Poole argued in a 2019 newsletter that much of the evidence for the futility of highway widenings comes from projects that marginally expanded already congested roads. More aggressive highway widenings, he pointed out, have resulted in considerable reductions in traffic congestion.

It's worth noting that one of the solutions that the critics of highway widenings offer as an alternative—expanding public transit—would also fail at reducing congestion. People switching from being motorists to being transit riders would also open up spare capacity on roadways that would, presumably, be gobbled up by other motorists thanks to that same induced demand phenomenon.

That doesn't mean more transit spending is a bad idea, but it also doesn't make more spending an obvious fix for urban gridlock.

A more effective solution to traffic congestion—one that both Poole and BMI have offered—is congestion pricing. This is where motorists are charged a variable toll to enter certain areas of a city or to travel on tolled express lanes.

By rising and falling depending on traffic demand, congestion tolls give people the option of paying a money price to travel during peak hours if they really need to. Those who can be more flexible with the timing of their trips can pay less or nothing at all to travel during off-peak hours.

Furthermore, because there is a money price being charged to drivers, new capacity isn't automatically filled by another motorist.

The D.C. metro area features congestion-priced lanes on sections of I-66. New York City is currently in the long, slow process of implementing congestion pricing for drivers entering lower Manhattan. International cities including London and Singapore have successfully used congestion pricing to reduce traffic in their downtowns.

One possible takeaway is that congestion pricing obviates the need to expand road capacity. Instead, higher tolls could be applied to existing lanes to manage demand.

But that could get expensive fast for drivers, particularly those of low or modest incomes. Expanding highways along with congestion pricing, alternatively, could lower the value of the tolls necessary to ensure freer flowing traffic.

Regrettably, the infrastructure bill that might soon pass Congress does very little to move the needle on the adoption of congestion pricing, which is already very difficult because of a general federal prohibition on tolling interstates.

Worse still, the $100 billion for new roads and bridges included in the bill would come from the general fund, effectively making it one big subsidy. That approach moves highway funding further away from the "user pays, user benefits" approach that prevailed when federal gas taxes (paid by road users) covered all federal highway spending.

Highway expansions can be a worthwhile endeavor. But those expansions should be funded by the drivers who will benefit from them—through gas taxes or, better yet, tolls and mileage fees. The latest congressional infrastructure bill recognizes the value of expanding road capacity, even if it doesn't go about paying for it in the right way.