Building More Roads Can Increase Travel Time
https://www.citylab.com/transportation/2018/09/citylab-university-induced-demand/569455/
The Katy Freeway in the Houston metro is up to 26 lanes wide at it’s widest point. It was built over the course of 3 years, at an expense of $2.8 billion, with the primary purpose of alleviating severe traffic congestion. However, counter intuitively, the new roadway had the exact opposite effect. Data from Houston’s traffic monitoring agency reported a 30 percent increase of the typical travelers morning commute time, and an increase of 55 percent during the evening. This was caused by an effect commonly referred to as induced demand. Essentially, induced demand is the idea that increasing roadway capacity only encourages more drivers to take that roadway, thus failing to alleviate congestion. These “new” drivers, that seemingly arise from nowhere, have presumably taken alternative modes of transport, or perhaps traveled at unpopular hours, before the new option was added. These drivers were given some perception of faster travel times, that encouraged them to undergo some behavioral changes. For example, the article proposes that some might “choose to take the new highway to a more distant grocery store that has cheaper prices” or may begin to consider that “trips that may have been accomplished by bike or public transportation might now be more attractive by car.” When people are given better options, they are likely to take advantage of them.
To me, this discussion was very reminiscent of our discussions of Nash Equilibrium on road networks. If we were to imagine, like we usually did in class, a road network between two cities with some roads between them, we can envision how this scenario might play out. There is some number of drivers to spread across the routes, that naturally form an Equilibrium that minimizes their own travel time. Then, a new, fast, perhaps even 26 lane, road is built between the cities. Clearly, the Equilibrium will change, and if this was an example in class, we would expect the travel times to go down. However, according to this article, in the real world this isn’t usually the case, because the number of drivers seeking a travel route won’t be static. In this scenario, many more drivers will join the network after this new road is built, because they believe that it won’t be as congested as their previous travel plans, or they may even create new travel plans after learning of the road’s efficiency. And evidently, due to the nature of this induced demand, it’s likely for the travel times in this new Nash Equilibrium to be higher than the travel times in the old one.
It is a truly bizarre, but empirically accurate, relationship. And, although we didn’t consider it in class, it would make sense in the context we thought about road networks and Nash equilibria, if we only considered a non-constant number of travelers.