It's been 10 years since New York City proposed congestion pricing as part of its PlaNYC sustainability initiative. That effort didn't end the way many mobility professionals working for city government at the time --- myself included --- would have preferred. The unfortunate result is that metro area commuters still spend far too much of their lives sitting in traffic, and that the transit system remains in search of steady funding sources.
The idea behind congestion pricing was remarkably simple: when people see that driving costs more, they drive less. This basic economic lesson has been offered in response to recent tech-driven hopes that tunnel networks or self-driving vehicles can solve traffic on their own. "The bottom line is, when you give away something valuable for free, you create insatiable demand," writes UCLA researcher Herbie Huff in the L.A. Times. "Traffic is the result."
Driving, of course, is not free. The average household pays roughly $9,000 a year in costs related to car ownership such as gas, tolls, insurance, maintenance, and parking fees. But driving often seems free because most of these costs are hidden on a per-trip basis. The car insurance bill comes monthly, and might be auto-paid. That big repair bill stings when it happens but is hard to factor into daily behavior. No one sits down to disaggregate all these small costs before running out for a gallon of milk.
Digital tools and data feeds enable us to think about this problem in a new way. If you could use navigation apps to reveal the full cost of a car trip, you might sway people's decisions about when and how to travel, or even whether they need to make a trip at all. This type of "true cost" comparison could also help inform policymakers, local planners, and companies hoping to address congestion.
At the moment, navigation apps tend to keep driving costs largely hidden. Most apps rank routes and modes based on travel time: the fastest option wins. The implicit assumption here is that the monetary costs among options balance out. But if you did see all the monetary costs of a trip choice before you left the house --- in addition to the time costs --- you might make a different travel decision.
Take a hypothetical commute into Manhattan from New Jersey on a Saturday morning. The time comparison alone would strongly favor driving, since traffic is lighter on weekends and public transportation runs less frequently. But adding up the true cost of the drive changes the mental calculus: $15 for the bridge toll, $35 for parking, and maybe $5 for other aggregate car costs amounts to several times more than the fare for transit --- compelling someone to think twice about driving, even if it's a faster trip.
A true cost tracker could support policy innovation, too. The most direct policy lever for congestion is road pricing, whether that's the type of cordon proposed in New York City or the mileage-based fees being tested in Oregonand California. The most common forms of pricing right now are managed or express lanes on highways or bridges, where prices vary based on traffic conditions. By showing people the cost of their trip ahead of time, commuters could potentially reserve a certain lane rate and know immediately what benefit they were getting in terms of time savings. And with advanced notice of someone requesting travel in the express lane, transportation agencies would learn more about the coming demand and be able to plan accordingly, or to send alerts or incentives to encourage a change in behavior.
Of course, road pricing raises important equity issues. But there is evidence that drivers of all income groups appreciate the option to get where they need to go faster, and tolls and road-pricing schemes arguably have a less negative impact on lower-income households than many existing revenue sources, such as a sales tax that funds transportation initiatives. Tolls and congestion zones can even improve alternatives to driving and car-ownership by redirecting money into public transportation, just as London's congestion charge does with local buses (and New York's would have, had it passed).
There are plenty of technical challenges to revealing costs up front in trip navigation tools. The biggest is data standards. Toll agencies often release data in fragmented datasets or in non-standard formats --- if they release them at all. Some agencies have recognized that problem; the Santa Clara Valley Transportation Authority in San Jose, California, for instance, has started to publish APIs for its data on Github. But coordinating all this information is a considerable challenge for developers.
There are some behavioral challenges, too. For most people, the daily commute is such an ingrained part of their routine that it's not a decision so much as a cognitive reflex. But there are certain times when more information or incentives can have an outsized impact on travel behavior, such as after moving homes or changing jobs.
And just as commuters have come to rely on tools like Waze to get them through traffic, they can learn to adjust their behavior when exposed to new streams of information. Since traffic is non-linear, even getting a small percentage of drivers to shift trips outside of rush hour could dramatically improve congestion. One Boston area study found that a 1 percent decrease in traffic from the most congested neighborhoods would reduce all metro commute times 18 percent.
Drivers are already paying for every trip they take --- the immediate cost just isn't as transparent as it is for transit riders. The result is often that driving appears much cheaper than it truly is, which encourages more people to drive, and makes traffic much worse. Technology alone won't solve the congestion problems facing high-demand cities, and policy innovations such as cordon pricing (which is back in the news!) face uphill political battles. But by revealing the true cost of travel in both time and money at the start of a trip, digital tools can help cities and commuters better understand the choices in front of them.
You can read and comment on this post at Medium.