Monday, April 25, 2011

On Roads, Tolls, and the Beauty of Dallas

It's not every day that you write a letter to a respected policy blogger and have him openly admit that the letter changed his viewpoint radically.  I now direct you to my letter featured on Matthew Yglesias' digs.  The first thing you will notice is that my letter in no way caused a re-think on Yglesias' part, nor did it even contain a point, per se.  

The Dallas Morning News reported on a construction project about to be undertaken by a Spanish firm along the 635 corridor in North Dallas--my stomping grounds.  This five year operation will dig beneath the current highway to build new toll lanes.  No additional free lanes will be built, only toll lanes.  Most importantly, the price of the toll will fluctuate based on current traffic conditions: during rush-hour (when 635 is unbearable), the toll might be several dollars.  

Unfortunately, I was in such a rush to get the email off to Yglesias that he (understandably) mistook my question about whether or not such an operation is likely to "work" as a question only about the likelihood of the toll lanes to decrease congestion and not about the likelihood of the presence of the toll lanes to decrease congestion on the free lanes.  

In Tom Vanderbilt's "Traffic: Why We Drive the Way We Do", the author states forcefully that, based on all available studies, building more lanes does not decrease congestion.  Rather, more lanes simply invite more traffic.  This happens because humans are unusually adept at planning for, negotiating, and avoiding traffic.  Therefore, when new lanes appear, it won't be long before commuters recognize the additional lanes and adjust their travel patterns accordingly (i.e. instead of taking I-90 to 35 to work, they will take 635 to 35 thanks to the additional roadway).  At first traffic will be noticeably smoother, but we shouldn't expect this to last long as additional cars (containing humans making rational decisions) will fill up the new lanes recreating the congestion the additional lanes were intended to fix.  

The question Yglesias took me to be asking concerned the toll lanes.  His response: "Naturally, everyone wants to take the free road. But as 'everyone' tries to crowd onto it traffic moves slowly and some people will want to exchange money for time by taking the toll road. And at any given time of day, there’s got to be some price at which the tolled road will be uncongested."  So, yes, the toll roads will "work."  

But the question I am really concerned with is the free lanes that the average commuter uses on a daily basis.  My initial thought was that the more people that opt for the toll lanes, the more space there will be in the free lanes for the traffic that does not want to pay tolls.  But one of the commenters to Yglesias' post says this is incorrect, and instead uses the logic of Vanderbilt's book to argue that the free lanes may experience less congestion for a short period of time, but before long, more traffic will come to fill the openings left by those cars willing to pay the toll.

A couple commenters disagree with the basic idea (c/o Vanderbilt) that it's impossible to build enough roadway to counteract the increased amount of new traffic.  They argue that people don't drive around aimlessly, so there will not always be a constant stream of new traffic to occupy all available lanes (in other words, drivers will not notice available roadway and drive on it just because they can--remember, drivers are surprisingly strategic).

Another interesting point presented by a commenter is that the more roadway constructed, the less available space for city-living.  I would be inclined to agree with this if not for the extreme driving culture that we see in Dallas/Ft. Worth.  The infrastructure in Dallas is phenomenal; from the High-Five to the Mix-Master to the DFW-Connecter, our roads are--while certainly headache inducing--beautiful in their utility and vision.  The Ancient Romans would, once they understood how we managed to survive without slave labor, marvel at our infrastructure accomplishments.  Thus, in Dallas, I believe, road construction could continue on indefinitely (through double-deckers, under-grounders, etc.) without any loss to the spirit that is Dallas.  

Whether or not all of this is a good thing for the environment and human prosperity is a different question entirely.  Suffice it to say, I am--as I always am--intrigued by the lastest attempt to efficiently move capital from the suburbs and exurbs into the heart of the Metroplex.  Forget the impending apocalypse, the Dallas Megalopolis is stunning.    



  1. Ok, so I tried posting this once and it failed, so now I'm ticked. I've put about 10 minutes of thought into this, so just hear me out. I think this Vanderbilt logic does not apply to congestion priced roads. Let's use the example of roller coaster lines and varying price fast passes. As the number of people in line increase, the cost of the fast pass increase. Now your argument is that every person that buys a fast pass is replaced by another schlup who sees that the regular line is a bit shorter and therefore strikes the iron while its still hot. This ignores the demand driven nature of the pricing scheme. When the regular line is long, some people decide that the fast pass is a good idea, even though its quite expensive. However, now that the regular line is getting shorter, this means the opportunity cost of paying money to get the fast pass is reduced because the regular line is now shorter. Instead of flocking to get the fast pass, people flock to the regular line. You see there is a constant ying and yang of people flocking between the fast pass and regular line. Vanderbilt's example assume that a massive band-wagoning effect occurs, whereby people see empty space and make the rational assessment to take the quickest route. However, the congestion pricing scheme means that there is differing incentives for the roadways at different times. When the congestion is heavy on the normal road, people are incentivized to go to the fast route. However, when congestion is light, no one would choose to fork over money when they can travel quickly on the normal road. This is the second time I've written this, and I feel my explanation last time was crisper.

  2. But doesn't this assume that there is a closed system operating? In other words, there are only so many cars at any given time that will use the roadway--some will opt for the tolled road, others will stick with the free lanes. Or in the rollercoaster example: out of the 60,000 people that decided to spend their Saturday at Six Flags, 5,000 choose the "fast pass", while the remaining 55,000 choose to wait in line longer.

    On the contrary, I'm saying that during rush hour there are a certain amount of people that would never be caught dead on a given road due to unbearable congestion and instead take an alternate route to and from work; but, once they discover that additional lanes have been added and traffic flow has improved, they will change their route accordingly.

    I don't think the Six Flags example is the same though: there is only one (new-and-improved) Texas Giant; the fast pass simply breaks the line into two, with a small percentage of Six Flag-goers paying extra to jump the line. Those who refuse to pay for the pass will wait the same amount of time as if there were no fast pass.

    I think this all goes back to my contention that people are rather strategic when it comes to such headaches as waiting.