Congestion and Price Elasticity of Demand

London, England, charges a fee for driving in the central area. The effect of the fee is twofold: it reduces congestion in the central core, and it raises revenue for the city.
[h/t to BF for these stories]

London's congestion charge is to rise to £8 this summer, a 60 per cent increase on the £5 daily rate faced by drivers in the centre of the capital....

The charge, which was introduced in February 2003, has been credited with reducing congestion in central London by 30 per cent, or by 70,000 vehicles. It raised £79.8m in its first year and was projected to raise about £100m annually in subsequent years. ...

Mr Livingstone [London Mayor] said: "Congestion charging has achieved its key objective of reducing congestion and has also provided an additional stream of revenue to help the funding of other transport measures within my transport strategy. The charge increase will maintain the benefits currently witnessed in the zone and build upon its success, cutting congestion even further and raising more revenue to be invested in London's transport system."

There is no need to estimate the price elasticity of demand here; the mayor knows it is less than one; otherwise, raising the fee would not lead to increased revenues.

This story illustrates how the goal of reducing congestion might be in partial conflict with the goal of raising revenue: If demand is price inelastic, raising the fee will have less of an impact on deterring driving in the city core but will raise revenue for the city; but if demand is price elastic, raising the fee will actually raise less total revenue, but will have a much bigger impact on reducing congestion. In the long run, the elasticity is likely to be greater as people find alternatives for driving in the central zone. Eventually, businesses will relocate, as will traffic, to areas outside the core. And so, if revenue generation is a goal, more tolls will be needed on additional roads outside the core.

well, son of a gun....

In a later, but related piece, we learn that the authorities are also considering congestion tolls that would vary with the time of day on other roads.

The blueprint would cover all roads within the M25 and see drivers charged different rates on every road according to the distance they travelled and the time of day. For example, motorists would pay up to £1.30 a mile for the busiest roads such as the North Circular at peak times, while quieter routes or times of day would be free or carry a lower levy.

I was pleased to see that the planners have taken into consideration the effect that tolls on some roads might have on the traffic on non-toll roads.

The CfIT has not yet finalised how to deal with all the details of the scheme, for example how drivers would be informed of what charges they were paying. In addition, there would be uncertain knock-on effects. For example, drivers may divert off onto smaller “rat runs” that would then in turn become congested.

The primary detail to be worked out involves monitoring and transaction costs. It appears they will rely on some combination of satellite tracking and electronic tags.