Phantom Toll Booth

On most days, my wife and I commute together by car.  And since my kids started a new, out-of-the-way school, our commute has gone from a fairly straightforward 15 minute trip—mostly in the carpool lanes—to a congested daily slog that, depending on traffic, can last over 45 minutes.

We definitely pay for our longer commute, through higher bills for gas and repairs.  But we don’t have to pay for the road space—we drive on the “freeway,” after all. 

But stuck in rush hour gridlock, among all the other drivers parked on I-5, it strikes me that the term “freeway”  is a misnomer. In fact, we do pay a toll to drive on the freeway, especially during rush hour.  It’s just that we pay with our time, not our money. 

Perhaps it doesn’t have to be that way.  Last Friday, the Seattle Timesreported on a high-tech experiment, conducted in the Puget Sound region, designed to test what would happen if people had to pay tolls to drive on the region’s roads.  In theory, tolls can ease congestion by reducing the number of trips on a crowded highway; tolls, in effect, let drivers exchange money for time.

That’s the theory, at least.  And the experiment bore it out.  Charging drivers variable tolls, based on the type of road and the time of day, did encourage people to drive less—especially during rush hour.

But, perhaps, not as much less as I might have hoped.

Here, from the Seattle Times article, is how the trial worked.

Researchers established virtual tolls ranging from a nickel to 50 cents a mile. They gave participants pre-paid accounts of between $600 and $3,000, and told them they could keep whatever the tolls didn’t eat up…

Participants…had devices mounted on their dashboards…that tracked their travel and transmitted the information to a central computer.  Tolls, which varied by road and time of day, were deducted electronically… using a global-positioning system and cellular technology.

In effect, the devices created a comprehensive system of roadway tolls.  All roads were tolled, but the tolls were calibrated to ease congestion on the busiest roads at the busiest times of day. And it worked: participants’ total mileage fell by 2.5 percent, while mileage during morning rush hour fell by 4 percent, and evening rush hour by 10 percent.

Reductions of that scale are nothing to sneeze at. A 10 percent decline in afternoon rush hour travel could have a huge impact on congestion.  Plowing toll revenue back into transit could further ease congestion, while making transportation more affordable for people who don’t, or can’t afford, to drive.  (That wouldn’t solve all of the equity issues with road tolls, but it would help.)

But at the same time, it came as a bit of a surprise to me how little the tolls affected driving overall.  I would have thought that people would leap at the chance to make some money, by combining trips or finding ways to walk or take the bus when possible.  But the 2.5 percent reduction in mileage seemed pretty small—it suggests to me that drivers avoided tolls by adjusting when they travelled, but not how far.

Of course, there’s reason to believe that a full-fledged tolling system—one that applied to everyone—could be more effective over the long term.  Gradually, tolls would influence decisions about where people work and live, and those decisions can have a far greater impact on travel habits.  Tolls might give a source of revenue for other forms of travel, such as transit and vanpools—which could further reduce demand for driving. Plus, recent work in psychology and economics shows that people tend to pay more attention to economic losses (such as tolls) than to potential gains (as were set up by the experiment).  If tolls were actual, out-of-pocket costs, I imagine that people would go to greater lengths to avoid them.

But maybe not.  Right now, congestion is pretty much the only force that regulates demand for road space.  Free-flowing highways would certainly attract some new drivers—even if they had to pay for the privilege.  So it could be that rush hour tolls would have to be even steeper than in the experiment to make such a large dent in rush hour travel.

In the end, it could be that the biggest effects of tolling would be psychological.  People would no longer think of driving as free; tolls would make it clear that mobility has a price.  That could help dampen the demand for new roads—which is  a major determinant of how much we drive.  And it might make people think a little differently about transportation more generally.  Perhaps mileage-based tolls might help us all realize that the point of a transportation system is to provide convenient access—to stores, jobs, and other amenities—rather than simply promote motion for motion’s sake.