Why You'll Love Paying for Roads That Used to Be Free: A Guest Post


Eric A. Morris is a researcher at U.C.L.A.’s Institute of Transportation Studies, concentrating on a variety of transportation issues including history, economics, and management. He weighed in here earlier on the gas tax. Here is his first of two posts on road tolls.

Why You’ll Love Paying for Roads That Used to Be Free
By Eric A. Morris
A Guest Post

To end the scourge of traffic congestion, Julius Caesar banned most carts from the streets of Rome during daylight hours. It didn’t work — traffic jams just shifted to dusk. Two thousand years later, we have put a man on the moon and developed garments infinitely more practical than the toga, but we seem little nearer to solving the congestion problem.

If you live in a city, particularly a large one, you probably need little convincing that traffic congestion is frustrating and wasteful. According to the Texas Transportation Institute, the average American urban traveler lost 38 hours, nearly one full work week, to congestion in 2005. And congestion is getting worse, not better; urban travelers in 1982 were delayed only 14 hours that year.

Americans want action, but unfortunately there aren’t too many great ideas about what that action might be. As Anthony Downs‘s excellent book Still Stuck in Traffic: Coping With Peak-Hour Traffic Congestion chronicles, most of the proposed solutions are too difficult to implement, won’t work, or both.

Fortunately, there is one remedy which is both doable and largely guaranteed to succeed. In the space of a year or two we could have you zipping along the 405 or the LIE at the height of rush hour at a comfortable 55 miles per hour.

There’s just one small problem with this silver bullet for congestion: many people seem to prefer the werewolf. Despite its merits, this policy, which is known as “congestion pricing,” “value pricing,” or “variable tolling,” is not an easy political sell.

For decades, economists and other transportation thinkers have advocated imposing tolls that vary with congestion levels on roadways. Simply put, the more congestion, the higher the toll, until the congestion goes away.

To many people, this sounds like a scheme by mustache-twirling bureaucrats and their academic apologists to fleece drivers out of their hard-earned cash. Why should drivers have to pay to use roads their tax dollars have already paid for? Won’t the remaining free roads be swamped as drivers are forced off the tolled roads? Won’t the working-class and poor be the victims here, as the tolled routes turn into “Lexus lanes”?

And besides, adopting this policy would mean listening to economists, and who wants to do that?

There’s a real problem with this logic, which is that, on its own terms, it makes perfect sense (except for the listening to economists part). Opponents of tolls are certainly not stupid, and their arguments deserve serious consideration. But in the end, their concerns are largely overblown, and the benefits of tolling swamp the potential costs.

Unfortunately, it can be hard to convey this because the theory behind tolling is somewhat complex and counterintuitive. This is too bad, because variable tolling is an excellent public policy. Here’s why: the basic economic theory is that when you give out something valuable — in this case, road space — for less than its true value, shortages result.

Ultimately, there’s no free lunch; instead of paying with money, you pay with the effort and time needed to acquire the good. Think of Soviet shoppers spending their lives in endless queues to purchase artificially low-priced but exceedingly scarce goods. Then think of Americans who can fulfill nearly any consumerist fantasy quickly but at a monetary cost. Free but congested roads have left us shivering on the streets of Moscow.

To consider it another way, delay is an externality imposed by drivers on their peers. By driving onto a busy road and contributing to congestion, drivers slow the speeds of others — but they never have to pay for it, at least not directly. In the end, of course, everybody pays, because as we impose congestion on others, others impose it on us. This degenerates into a game that nobody can win.

Markets work best when externalities are internalized: i.e., you pay for the hassle you inflict on others. I’ve blogged on the externalities inherent in driving on this blog, and Levitt and Dubner wrote about it here.

Using tolls to help internalize the congestion externality would somewhat reduce the number of trips made on the most congested roads at the peak usage periods; some trips would be moved to less congested times and routes, and others would be foregone entirely. This way we would cut down on the congestion costs we impose on each other.

Granted, tolls cannot fully cope with accidents and other incidents, which are major causes of delay. But pricing can largely eliminate chronic, recurring congestion. No matter how high the demand for a road, there is a level of toll that will keep it flowing freely.

To make tolling truly effective, the price must be right. Too high a price drives away too many cars and the road does not function at its capacity. Too low a price and congestion isn’t licked.

The best solution is to vary the tolls in real time based on an analysis of current traffic conditions. Pilot toll projects on roads (like the I-394 in Minnesota and the I-15 in Southern California) use sensors embedded in the pavement to monitor the number and speeds of vehicles on the facility.

A simple computer program then determines the number of cars that should be allowed in. The computer then calculates the level of toll that will attract that number of cars — and no more. Prices are then updated every few minutes on electronic message signs. Hi-tech transponders and antenna arrays make waiting at toll booths a thing of the past.

The bottom line is that speeds are kept high (over 45 m.p.h.) so that throughput is higher than when vehicles are allowed to crowd all at once onto roadways at rush hour, slowing traffic to a crawl (more on how this works in a minute).

To maximize efficiency, economists would like to price all travel, starting with the freeways. But given that elected officials have no burning desire to lose their jobs, a more realistic option, for now, is to toll just some freeway lanes that are either new capacity or underused carpool lanes. The other lanes would be left free — and congested. Drivers will then have a choice: wait or pay. Granted, neither is ideal. But right now drivers have no choice at all.

What’s the bottom line here? The state of Washington recently opened congestion-priced lanes on its State Route 167. The peak toll in the first month of operation (reached on the evening of Wednesday, May 21) was $5.75. I know, I know, you would never pay such an exorbitant amount when America has taught you that free roads are your birthright. But that money bought Washington drivers a 27-minute time savings. Is a half hour of your time worth $6?

I think I already know the answer, and it is “it depends.” Most people’s value of time varies widely depending on their activities on any given day. Late for picking the kids up from daycare? Paying $6 to save a half hour is an incredible bargain. Have to clean the house? The longer your trip home takes, the better. Tolling will introduce a new level of flexibility and freedom into your life, giving you the power to tailor your travel costs to fit your schedule.

But is this fair? Haven’t we already paid for these roads with our tax dollars? Won’t the wealthy buy their way out of congestion, while the rest of us are stuck? And won’t conditions in the free lanes deteriorate as they are packed with all the drivers who have fled the toll lanes?

These are perfectly legitimate and important concerns. But I should warn you: toll advocates are a pretty dogged bunch, and they have an answer to everything. I’ll let you know their (quite convincing) responses to these questions in my next post.


I'm reading Tom Vanderbilt's Traffic and this is essentially his same argument. It is counter-intuitive, but once you evaluate the rationale is becomes clear that this would work. We've already paid for the roads, but now we'd be paying for the efficient usage of them. Might not be "fair" but it's reality.


One point to make is that this has been done elsewhere in the world (http://www.edf.org/page.cfm?tagID=6241), and its worked. It is amazing how stubborn people can be when they don't understand something and rush to judgments. I can guarantee that this post will soon be flooded with comments of people irrationally berating the merits of this idea.


Regressive tax of the worst kind.


I admit, I'm interested to hear the rationale for requiring people to pay yet again for the roads that they use. After all, we already pay for these roads all the time, through taxes that pay for construction and repairs. The problem I see with Morris' logic is that this seems to be a purely economic treatment of the toll question, which ignores the fact that the roads are not an economic resource - the equity angle that Laura brushes off at the end is a huge part of the discussion, and glossing over or ignoring it is ignoring a big part of the debate.


It seems to me that there needs to be better traffic engineering and city planning rather than just using the default economist answer of "make people put their money on it and see how much they really value it". Sometimes I wonder if that's all an economist ever has to say.


Also, yot is right - this amounts to a regressive tax on poorer drivers, and I have yet to see any evidence that a regressive tax is a good solution for anyone other than the wealthier income brackets.


Alan -

The ends justify the means is not a well formed argument in favor of anything.


Get rid of the highway taxes on gas and I'd go along with this.


The "tax" is actually very progressive. By moving those who can afford pay to the "pay lanes", it clears up the "free" lanes for those who can't once the pricing is in balance. Better for everyone and only those who can afford to pay do. What could be more progressive?


Shamelessly stealing from another blog comment author, why not index the toll to the street value of the vehicle itself? Or the income of its owner?


My question is do these data suggest that privatization would be a good solution? I'm not a free market zealot, but if I were I'd probably try to make that argument. Would government computers do a better job of maximizing traffic flow than a private business? If we are going to pay for roads each time we use them, is a public model the best one?

a student of economics

This is a great idea and I hope it becomes more widespread. An additional point in favor of congestion pricing is that it generates revenue. That money can be used to cut taxes on work (e.g. payroll taxes) or simply rebated to everyone equally. In other words, you can make the whole scheme revenue neutral. On average, we pay no more than we do now.

All we have to lose is the time spent stuck in maddening traffic delays.

Caliban Darklock

I am a firm believer in tolls.

Many years ago, in Virginia, I used to take the Dulles toll road to work. By doing so, I shaved about 45 minutes off my daily commute for a grand total of 70 cents. Every time someone commented on how far away I lived, I would mention that it's not so bad when you take the toll road.

I remain completely mystified by the visceral opposition people had toward the idea of paying less than a dollar to use a road.

I do have a similar opposition, myself: carpool lanes. But my opposition to carpool lanes comes from the basic reality that I pay for them, yet cannot use them, and when you graph the expected result of carpool lane adoption... well, you find that in almost every single case, it makes traffic WORSE. I can scarcely imagine a worse investment.


this idea sounds about as ancient as the Romans- ever hear of public transportation infrastructure? or how to stave off a depression the feds should pump billions into infrastructure development?- it really is hard for me to believe that when the time and conditions have finally come for the US to catch up to Europe and build public transportation, we get the cadre of experts instead lobbying for more expensive single car/gas transit


The number of miles of new roads has not kept up with new cars since the late 1960s. Thus we have traffic jams. Indeed in the 1970s and 1980s new roads projects were vigorously opposed by many local pressure groups. Now we are reaping the consequences.

The solution? Raise the federal and state gas tax such that it could pay for the building of those unbuilt roads and pay for the building of subway lines in major cities. I envision raising the federal gas tax from 18.4 cents to 70 cents and raising the average state and local tax of 30.0 cents per gallon to 60 cents. This would not bring the US to European levels of taxation. Yet it would enable us to successfully tackle the problem.
(source for figures http://www.api.org/statistics/fueltaxes/upload/October_2008_gasoline_and_diesel_summary_pages.pdf)


This could lead to other activites' desirability being revealed:

"Why were you late for work today?"
"I just don't like this place enough to pay the toll."

Many other scenarios come to mind.


In Chicagoland, the Toll Roads are already variably priced depending on fixed hours, more during "rush hours" and less in the middle of the night.

I don't know how much this has gone to change driving behaviors as there is no "cheaper" option other than side roads.

I DO know that I have altered my driving habits over the last 10 years due to congestion, I only drive into and through Chicago (free highway for now) during the day between 10:30am and 2pm
If I can't get what I need done within that window of time, I don't do it, the crush of cars and drivers is just too maddening.


I'm curious to hear your opinion about alternatives, specifically the way Vancouver deals with congestion heading into downtown. Since downtown Vancouver is an island, it's understandably challenging to get in and out. There's a lot of public transit available with buses and high-speed ferries (I believe just for people, with park-and-rides at the non-downtown terminal), but businesses also have flexible start and finish times. People who work in offices start as early as 6:00 or as late as 10:00, so the afternoon rush is staggered from 3:00 to 7:00.

I used to drive I-394 every day (thankfully against traffic), and I always thought that staggering start and finish times in downtown would have had a major effect. In the afternoon I had to get through downtown to make class at 5:45. If I left at 4:45, I'd get to the U at 5:10. If I left at 5:00, I'd get to the U at 6:00 because of everyone rushing out of the office at 5:00 on the dot. and the pay to use the carpool lane system? I haven't seen the revenue, but there weren't nearly enough cops to keep people from gaming the system. They couldn't just ticket cars without transponders because it's free if there are two+ people in the car.


Nick Schoch

Ideally, these toll revenues would be used to offset the costs associated with future road improvements, reducing the future tax burden. That or the funds could be reallocated towards cheap mass transit. This would help offset the "regressive" impact of the tolls.


Sure, it works, if by "works" you mean "stops those darn poor folks from slowing down traffic for economists."

People who don't have a lot of disposable income deserve public services too. Laughing about $6 for a half hour of your time is a lot less funny if you make $10/hour.