Widening a road reduces congestion for a period and then does not.
The traffic that fills the new is called , and it comes from three sources.
Some drivers switch from other routes, some switch from other times of day, and some make journeys they previously did not make.
Only the third group represents new travel, and it is the group that takes longest to appear.
conducted a year after opening therefore record a success that the five-year figure removes.
The mechanism is not mysterious and is the same one that operates in any market where a price falls.
The price of a journey is mostly time, and widening a road reduces it.
Nobody is surprised when cheaper bread is eaten in larger quantities.
The finding is nevertheless resisted, because it appears to prove that nothing can be done.
It proves something narrower: that alone does not reduce congestion in the long run.
Congestion falls where the price of a journey rises or where an alternative becomes cheaper in time.
Both are politically harder than building, which is why building continues.
There is also a version of the effect that runs in the other direction and is less well known.
Removing road does not produce the that is always predicted.
Studies of some seventy found that a substantial share of the traffic simply disappeared.
It did not reappear elsewhere; the journeys were , combined or not made.
Traffic is treated in public argument as a fixed quantity that must go somewhere.
It behaves instead like , which is the single most useful thing anybody can know about it.