The Braess Paradox and Induced Demand
https://www.wired.com/2014/06/wuwt-traffic-induced-demand/
The article linked above discusses the counter-intuitive phenomenon that adding new roads, wider highways, and more public transit does not actually decrease traffic congestion– it just increases driving demand. The concept of new roads and other increased transportation options structures actually materializing more travelers, because of the higher incentive to travel they bring about, is referred to as induced demand. New roads allow more people to move around, and with more people driving on the roads, there is no improved road congestion. Say, for example, that Boston, one of the most congested cities in the US, was to add a new road connecting Fenway Park to the Boston Harbor. There would be a new path, and thus more space, for cars driving from one of these places to the other. However, people who wouldn’t want to drive between Fenway and the Harbor before the new infrastructure would now want to because of the new road, and businesses that rely on roads, such as trucking, would swoop in and take advantage of the new travel opportunity. Therefore, overall traffic congestion and flow remain the same as before. Even if public transit is added or improved, some people will switch over to this mode of transportation but again, new drivers will replace the ones that left. The reverse is true as well, in that taking away roads or decreasing the number of lanes on highways does not actually increase traffic congestion– traffic simply readjusts and more people walk or use other methods of transportation since they do not want to sit in the traffic that they believe will increase due to the new downsized roads. The article explains that implementing tolls is a method that will actually decrease traffic congestion, because putting a price on driving these roads at certain times, specifically rush hour, will incentive people to travel at different times or use public transit.
The Braess Paradox, as discussed in class, takes this same general notion and applies it to the creation of a new road on an existing road network, stating that each person’s travel time would actually be increased by the new structure. Looking at the graph below, we found in class that by creating a minimally time-consuming road from C to D, where before the only traffic flow options were ACB and ADB, the nash equilibrium of the new road network would be for every driver to take A→C and then D→B. That is because with this new road, the only dominant strategy is to take ACDB, as this is the only way no driver will have an incentive to deviate as deviating will only increase their traffic time. If there are 200 drivers on this road, the total travel time of ACDB if everyone takes this route is 40 minutes, as 200/10+200/10=40. No driver would want to switch from this path because if a driver decides to deviate and take the path ACB, their travel time would end up being higher (65 minutes, as 200/10+45=65). Taking the new road and going the ACDB route is the dominant strategy, but implementing this new road actually increased the travel time for everybody, as before the new road, the dominant strategy was for 100 drivers to take ACB and 100 drivers to take ADB (which results in a travel time of 55, as 100/10+45=55). A travel time of 55 is less than the nash equilibrium travel time of 65 that results when the CD road was added. So the new road increased travel time for everybody, but taking CD is still the dominant strategy. Additionally, the routes which aren’t dependant on amount of drivers (CB and AD, as they always will be 45 minutes no matter how many drivers take them) can be seen as public transportation such as trains, as the amount of traveller that take trains does not really affect the travel time of these modes of transportation. Even if more train routes are added (in this case, another route from A to B that has a travel time not dependant on amount of drivers, such as 45 minutes), and more people start utilizing the trains, other drivers will take advantage of the cleared up space on the x/10 and y/10 routes and fill up the driving roads to just as they were before. Although public transit does do good, as it allows more people to move around and is good for the environment, it usually shouldn’t be looked at as a method of traffic decongestant, just as mentioned in the article.
Overall, the Braess Paradox supports the real-life notion talked about in the article that adding new roads (such as CD in this example), or adding new or improved public transit (such as adding another 45 minutes train route) does not actually improve travel time for drivers. The paradox concept discussed in class thus applies directly to the real-life traffic and road addition phenomenon discussed in the article.
