But late last year, Virginia added another option. The state started letting single-occupancy cars use the road, but at a price. The exact price would depend on how many people were driving on it at a given moment, bringing the concept of supply and demand to the daily commute. The greater the congestion on the tolled road, the higher the charge. The idea was to keep traffic moving quickly along the highway, even if that meant discouraging people from using it. The exemption for electric vehicles was eliminated, and the rush hour period was extended from two and a half hours each way to four. The state promised to use up to $10 million of the toll money to pay for more bus routes, bike-share stations and other options that would give commuters choices other than clogged highways.
Those details, though, were quickly overshadowed by how high the tolls went. At one point in the first two months, the posted price along the 10-mile stretch hit $47.25, and it broke the $40 mark several other times. True, very few people paid that toll -- statistics showed that for the first 594,381 trips taken under the new rules, only 461 drivers paid a toll of $40 or more. And anyway, that was the point, to clear traffic off the road. But it didn’t stop drivers, local newscasters and Northern Virginia politicians from denouncing the “highway robbery.”
Managers at the Virginia Department of Transportation, on the other hand, viewed the project as a success. The fact that some drivers, even if just a few, were willing to spend $40 for a quick trip in or out of Washington, D.C., showed VDOT the value of adding a tolled option. If not for the changes, drivers traveling alone who really needed to get somewhere fast during rush hour were literally stuck with no choices. Now, at least, they had one, even if it meant paying an exorbitant price.
The sky-high tolls drew national attention, but similar systems, sometimes known as “congestion pricing” or “demand pricing,” are being installed throughout the country. There are roughly 40 of them now in 15 metropolitan areas, the vast majority of which opened in the last decade. The concept of paying more when traffic is heaviest is expanding in Atlanta, Los Angeles, Seattle and Tampa. New toll charges are being debated in Maryland and in Portland, Ore. And New York City is considering a related idea: charging motorists a fee for driving into lower Manhattan at busy times. “Political leaders and people say they want to solve the congestion problem,” says Jennifer Dill, the director of the Transportation Research and Education Center at Portland State University. “Right now, people are paying for congestion in time. Transportation planners and academics like me say, if you want to solve the congestion problem, you have to charge the right price.”
It’s the appeal of solving two problems at the same time -- managing traffic and bringing in revenue -- that largely explains why the idea of congestion pricing is getting so much attention these days. “It is a very powerful tool for managing congestion,” says Patrick Jones, the executive director and CEO of the International Bridge, Tunnel and Turnpike Association. “We’ve seen that with I-66. There’s been quite a bit of hue and cry about it, but nobody is compelled to pay that toll. The express lanes are doing what they set out to do.”
The prices and peculiarities of the I-66 toll system are new to Virginia, but the concept of congestion pricing has a long history in the state. Well before Virginia added its recent tolls, it granted Transurban, an Australian company, the right to add high-occupancy toll lanes (HOT lanes) to two of the region’s biggest expressways. The company started using demand-based pricing in 2012, when it opened toll lanes on a rebuilt 14-mile stretch of the Capital Beltway. A little later, Transurban added fluctuating tolls to the reversible lanes along a 29-mile segment of Interstate 95 through the outer Washington suburbs.
Motorists can check the shifting rates by looking at roadside signs or mobile apps. But the process that Transurban actually uses to set the tolls is nearly invisible to most users. Roadside radar devices track how many cars are traveling down certain segments of the tolled lanes at any given time, and cameras cover the entire length of the Transurban highway. The data and video feeds are sent back to a suburban office park where Transurban runs its U.S. operations out of a two-story building with an angular glass face.
Inside, Transurban employees follow the traffic on 18 flat-screen televisions covering the front wall of a sunlit room that feels like a modern college lecture hall. They also each watch a bank of computer monitors, looking for anything that could impede the flow of traffic. If a car breaks down and pulls over, they’ll send a maintenance crew. If there’s a crash backing up traffic, they can post warnings on highway message boards so that motorists can avoid the area, or at least get out of the toll lanes to avoid paying money to sit in bumper-to-bumper frustration. Operators can change the speed limit if needed.
The process of calculating tolls, though, takes place in a back room, somewhere out of sight, where servers crunch real-time data on traffic conditions. The goal is to spit out toll rates that will keep traffic cruising in the toll lanes at a breezy 60 mph.
Like the I-66 tolls, the rates on Transurban’s lanes are not capped, which theoretically means they could go just as high as they have on the neighboring interstate. But Transurban says its average tolls have been much lower: $5.77 last year on the Beltway, and $8.46 for I-95. The system saves rush hour Beltway commuters an average of 23 minutes per trip, according to the company. Even motorists using the free lanes save time, Transurban says, because other drivers are in the toll lanes.
The goal of keeping traffic moving at 60 mph in the toll lanes is a federal requirement, but it’s also crucial to satisfying customers who want the maximum benefit from the system, says Transurban spokesman Michael McGurk. “Folks that are paying the tolls want to be going at highway speeds,” he says. “They want to go 60 mph if they’re paying tolls or picking up carpoolers. They don’t want to go 40 mph.”
The arrangement seems to be working for Transurban. The company reported that in the last fiscal year, the number of average daily trips increased by 12.8 percent over the previous year. Some of that is because carpooling and bus trips have been increasing. But demand for the toll lanes is going up, too, and that, in turn, is pushing up toll prices and bringing in more money. Transurban says its toll revenue in the Washington area grew by 23.7 percent in the year ending in June 2017, to $157 million. The average price to use the toll lanes increased by 21 percent on the Beltway and 19 percent on I-95.
In other words, more drivers are using the toll lanes, and they’re paying more to do it. It’s no wonder Virginia and other jurisdictions are planning to build more of them.
While many state and local officials are still getting comfortable with congestion pricing, the idea has been in the works for decades.
The so-called “father of congestion pricing” was William Vickrey, who first became fascinated by the link between prices and traffic when he took a largely empty train from New Haven to New York as an undergraduate in the 1930s. He started thinking about how wasteful it was that so many of the seats went unfilled when many of his fellow classmates would have loved to spend some time in the city if the train fare were lower.
In the early 1950s, Vickrey tried to convince New York City to charge subway riders more if they wanted to ride during rush hour. He suggested that all vehicles should have transponders installed, so that city officials could charge drivers for entering crowded areas, and charge them more for entering during high-traffic times. Neither idea moved forward.
Transurban workers monitor traffic conditions to make sure vehicles can travel swiftly in its Virginia toll lanes. (David Kidd)
In 1963, Vickrey, by then a professor at Columbia University, laid out the case for congestion pricing on New York’s bridges. He questioned why New York officials let drivers cross older East River bridges for free because the bridges were “paid for,” but charged tolls on newer bridges. That approach promoted congestion on the older bridges. “The delusion still persists,” Vickrey wrote, “that the primary role of pricing should always be that of financing the service rather than that of promoting economy in its use.”
Over time, his ideas gained more acceptance. Many transit systems today charge passengers more for traveling during peak hours. Everything from airline fares to utility rates to baseball tickets is sold using some form of congestion pricing. Vickrey, the champion of all these once-obscure ideas, won the Nobel Prize in Economics in 1996, just days before his death.
The use of congestion pricing on toll roads in the United States began slowly. The first to make use of it was State Route 91, south of Los Angeles in Orange County, Calif., which opened in 1995. The project added four tolled lanes next to eight free lanes along a 10-mile stretch of the highway. Although a private consortium built and originally ran the toll road, Orange County took it over in 2003 and kept the congestion pricing. The project gave transportation planners a real-life example of how the abstract idea of congestion pricing might work for roadways.
It came at a time when much of the technology that makes demand-based pricing practical was becoming more widespread. Electronic transponders, like the E-ZPass now used in much of the Midwest and on the East Coast, were first rolled out in the early 1990s. The transponders allowed for the widespread introduction over the next decade of open road tolling, in which drivers don’t have to slow down to pay their fares at a toll booth. That saves them a lot of time and makes the toll itself seem less onerous.
The new experiments also benefited from changes in federal transportation policy. In 1991, Congress passed a transportation funding law that for the first time let states use congestion pricing to set tolls on federally backed highways. Congress has long prohibited states from adding tolls to existing interstate lanes, but it did start allowing states to convert carpool lanes into toll lanes. It also let states toll newly added lanes along those interstates. The administrations of both George W. Bush and Barack Obama encouraged states to take advantage of those exceptions, and states found many private companies like Transurban eager to get into the tollway business.
But despite the fancy algorithms and larger societal goals, the fact remains that congestion pricing still takes money out of drivers’ pockets. In many quarters, “toll” is still a four-letter word.
For most traditional opponents, the pricing structure doesn’t fundamentally change the argument against tolls. Truckers are one example. The American Trucking Associations has long favored raising fuel taxes to pay for better roads rather than imposing new tolls. The trucking industry says it can live with congestion pricing, if it allows truckers to avoid tolls by using general purpose lanes. But any attempt to use variable rate tolls to discourage truckers from traveling at peak times will likely fail, says Darrin Roth, the group’s vice president of highway policy, who says truckers rarely opt to pay for faster lanes when given the choice. “The goal of congestion pricing,“ he says, “is to get people onto alternative modes of transportation or change the time of day they travel. But changing modes is not an option for truckers. And pickup and delivery times are determined by customers.”
Most trucking companies don’t have any way to assign the costs of particular tolls to individual customers; instead, the costs of tolls are lumped together as an overhead expense that’s borne equally by everyone who buys the service. Charging a customer for getting a delivery during rush hour could be even more complicated under congestion pricing, Roth says, because a company would not know what toll rate to charge the customer until the truck was on its way.
Likewise, congestion pricing doesn’t change the calculus for opponents of adding tolls to existing interstates, something that the Trump administration, like the Obama administration before it, has encouraged.
Stephanie Kane, a spokeswoman for the Alliance for Toll-Free Interstates, says toll roads can overcrowd surrounding communities when drivers get off the interstate to look for a free route on smaller roads nearby. And toll roads are relatively inefficient ways to raise money for transportation projects, at least compared to fuel taxes. “Running it is complicated,” she says. “The contracts are complicated. Enforcement is complicated. They’re so inefficient in so many ways, that tolls should be removed from the conversation about transportation funding altogether.”
Managed lanes, whether they have fixed or variable toll rates, have also long been disparaged as “Lexus lanes,” catering to rich motorists. Toll operators insist that the tolled lanes are not there to provide an everyday shortcut for wealthy drivers, but to make it easier for ordinary motorists to get where they are going when they cannot be late, whether it’s catching a flight, meeting with a boss or picking up children from school. Many researchers have concluded that this is, in fact, how most drivers use the lanes -- as an option in emergencies -- although richer motorists do use them more often.
But among the voices urging caution about congestion pricing are experts in the tolling industry itself. They warn that such a system works only in certain circumstances, and that it needs to be tailored to the specific needs of a region.
Matthew Click, an expert on managed lanes for the construction engineering company HNTB, has reviewed dozens of congestion pricing proposals. Most times, he says, the projects aren’t feasible. They fall short for a number of reasons. From a practical standpoint, an area has to have enough of a traffic problem -- measured by distance, density and duration -- to make congestion pricing workable. There needs to be a way to solve that problem with new lanes. Even more important, local leaders must first agree on what the problem is. “The very first thing public officials have to ask themselves is why they want to institute this type of policy in their area,” Click says. “Is it to improve transit? Better air quality? Mobility benefits? A lot of people rush to ask how we’re going to do it, but I would say they should start with the why.”
Robert Poole, the director of transportation policy at the libertarian Reason Foundation and one of the proponents of the original Orange County experiment in the 1990s, agrees that some of the difficulties of demand pricing projects trace back to differing visions of what they should accomplish. Some proponents see the tolled lanes as a way to encourage people to share rides. On the other hand, Poole puts a priority on providing a congestion-free route for buses, emergency vehicles and drivers who are making high-value trips.
It’s not just an academic distinction. One reason the tolls on I-66 are so high, Poole says, is that, for the time being, vehicles carrying only two people ride free there. With so many two-person carpools, there’s little room for toll payers. Virginia plans eventually to raise the threshold for free rides on I-66 to three passengers per vehicle, which would bring it in line with the other toll facilities in the area. But Poole would abolish the carpool exception completely, something that toll roads in the Austin and Baltimore areas have already done.
Atlanta is moving in that direction, too, as the region plans to expand its network of tolled lanes aggressively over the next decade, says Chris Tomlinson, the executive director of Georgia’s State Road and Tollway Authority. The state will continue to allow a free ride on Interstate 85’s existing carpool lanes, but in its other projects, where the state will be adding new lanes, the only vehicles that will likely get out of paying tolls are transit buses. That policy is simpler, Tomlinson says. “We are all in on the public policy that these toll roads should provide a benefit for those who pay the tolls or are using public transit operating in these corridors.”
Another factor that can slow traffic in managed lanes is an artificially low cap on toll prices. Low tolls are politically popular, but if they are too popular, they can defeat the purpose of the toll lanes, which is to keep the flow of traffic at an acceptable pace. That’s part of the problem in the Salt Lake City area, where Utah officials hoped to keep traffic in the Interstate 15 toll lanes at 55 mph. Tolls there have been capped at $1 a ride, and average speeds at peak times have dropped to 31 mph. State leaders are considering doubling -- or even tripling -- the maximum toll.
Another problem in some areas is cheating: People will set their transponders to carpool mode even if they have no other passengers in the car. On the toll lanes of Los Angeles’ 110 freeway, for example, it is estimated that 25 to 30 percent of drivers are in the lanes illegally. The county is working on technology that can determine how many people are in a car, so it could automatically send tickets to freeloaders. But Poole says the tolling industry has been trying for years to develop that type of tool, with no success. “The only enforcement has been when an officer in a patrol car looks in and sees that there’s not two or three people riding in the car,” he says. “It’s kind of a joke, frankly. And a lot of people are troubled by it.”
Perhaps the biggest threat to congestion pricing on toll roads is the potential for a political backlash, like the one in Virginia early this year. But in Virginia and other states that are building a network of toll lanes, changing course could be difficult.
Several Virginia lawmakers from the D.C. area tried to convince their colleagues to change the tolling scheme for I-66. Some of their ideas would not have affected the top rate, but merely scaled back the hours when the tolls are imposed. So far, though, those changes haven’t found much traction in the legislature. One reason is that the state already signed a deal to add new toll lanes using congestion pricing for I-66 outside the Beltway. As part of the deal, the private operators promised to pay $500 million to the state up front. But some or all of that money could be in jeopardy if Virginia changed the tolling rules for its segment of the highway. Meanwhile, the state has reached deals with Transurban to extend its congestion-priced toll lanes on I-95 north toward Washington and south toward Fredericksburg. Despite the recent outcry, it looks as if Virginians will see a lot more congestion pricing, not less, for years to come.