Why new roads do not alleviate congestion

[Jonno]

http://economicsintelligence.com/2011/10/24/why-new-roads-do-not-alleviate-congestion/

Great article from Economics Intelligence that unemotionally explains why we have never reduced congestions.  Not a tree hugged in sight just pure economics.

Each year, motorists in Germany collectively spend 234 million hours in traffic jams. According to estimates by the German highway authority, this congestion causes an economic loss of about 3.5 billion Euro per year.

At first sight, there's an easy solution to this problem: just build more roads, and the congestion will vanish, won't it? Automobile lobbyists are pushing this view very strongly, and they are at least partially successful.

Even in cash-strappedBritain, which is currently enduring the most severe austerity programme since the second world war, the government spends 897 million pounds on new roads, reports George Monbiot.

However, more roads will not bring an end to congestion. Instead, they will bring more traffic. The American economist Anthony Downs put forward this hypothesis more than four decades ago.  He called this idea the "fundamental law of highway congestion".

According to an amazing paper by two Canadian economists, Downs really had a point.


Gilles Duranton and Matthew Turner, both affiliated to the UniversityofToronto, empirically analysed the relationship between road building and traffic volumes and came to a very sobering conclusion:

"increased provision of interstate highways and major urban roads is unlikely to relieve congestion of these roads."

The paper is entitled "The Fundamental Law of Road Congestion: Evidence from US Cities" and has been published in the latest issue of the "American Economic Review", one of the globally most renowned and demanding economic journals.  (Here's a free working paper version of the paper.)

The meticulous study rests on highly detailed regional data about public expenditures for roads and the development of traffic volumes between 1983 and 2003. Duranton and Turner looked at these  numbers from different perspectives. On the one hand, they analysed how the road network and the traffic volumes in a particular region have evolved over the years. On the other hand, they investigated differences in infrastructure expenditures and traffic flows across regions in a given  year.

Both approaches produce very similar results: A one percent increase in roads leads to a one percent increase in traffic within less than ten years. In other words, if you double the road infrastructure, car traffic also doubles very quickly.

Duranton's and Turner's study is highly relevant for transport policy and for urban planning. They came to the conclusion that the "fundamental law of highway congestion" does not only hold on Highways. Apparently, it is also valid within cities. For a "a broad class of major roads" within urbanised parts of America, a 1 percent increase in road kilometres triggers a 0.67 to 0.89 percent increase in vehicle-kilometres travelled.

Amazingly, the quality of the local public transport network does not affect this pattern. According to the paper, increased "provision of roads or public transit is unlikely to relieve congestion."

However, do the economists really observe a causal relationship? A mere correlation does not explain anything about cause and effect. In general, it is possible that more streets are built in regions where traffic volumes are on the rise because of other reasons. However, the economists make sure that their results are not driven by regional economic activity, population growth or other socio-economic factors.

Additionally, they use a method that is common in econometrics for disentangling cause and effect: They employ an instrumental variable approach. As they explain in their paper:

"To identify the causal effect of roads on traffic, we examine both time series and cross-sectional variation in our data and exploit three instrumental variables to predict the incidence of roads in metropolitan statistical areas (MSAs). These instruments are based on the routes of major expeditions of exploration between 1835 and 1850, major rail routes in 1898, and the proposed routes of interstate highways in a preliminary plan of the network. Our results strongly support the hypothesis that roads cause traffic."

What exactly causes the additional traffic? Duranton and Turner addressed this question, too. The most important factor is that people change their driving behaviour. A better road network induces people to drive more. That areas with a better roads attract new residents turns out to be less important.

Hence, when it comes to traffic, Say's law really seems to hold: new supply creates its own demand.