That a road doesn't generate traffic can be seen by the consideration of a piece of road which is built in the middle of the countryside. It doesn't connect to any existing roads at either end. How much traffic is generated by this road?
The answer, clearly, is none. Roads do not generate traffic. Roads are a strip of tarmac, and do not have the capacity to generate anything.
An alternative point of view is that roads induce traffic - that is to say, that providing an extra connection in the highway network will induce car drivers to drive more. Eventually, the new road will be a full, and so will the old road.
The Highways Agency contracted Atkins to write an assessment [large PDF] of the operation of the bypass. Despite the claims of high levels of induced traffic, Atkins identified the following causes of the increase in traffic growth along the bypass. The induced traffic component is the smallest one on the list.
The report indicates that the main area where there has been extra traffic growth is in Newbury, due to a major reduction in congestion - this is the entirety of the induced traffic. The majority of the traffic on the bypass is reassigned, in other words, it comes from other less suitable roads. So, the bypass is doing what was intended.
The report summarises the advice given by the government about induced traffic:
The report goes on to say that the bypass hasn't sustantially changed the amount of time taken to get around, and so very little induced traffic has occured. However, much of the remaining traffic increase can be explained by the large amount of development which has occured in the area.
The effects of development can be mitigated. Before the highway network improvements, it is difficult for a developer to show that the network can accomodate the extra traffic from the proposed development. After the improvements, it is easier to demonstrate this. The local council therefore has to assess the development against the pre-improvement network. Of course, this lies in opposition to the interests of the local economy, and so the application has to be considered on more than just traffic effects.
Where congestion is relieved in an urban area, it is likely that traffic will be induced. However, as this can be reasonably foreseen, steps can be taken in the urban area to switch the residents away from cars.
