What is induced demand?

Induced demand is the effect that adding capacity, such as a new road lane or a faster service, attracts extra users who were not using it before. This means congestion often returns to its earlier level after an expansion. It matters because it can make a large investment disappoint, and it shows why demand cannot be treated as fixed.

How it works

Users decide whether to use a service by comparing its cost, including time, with the alternatives. When capacity grows, delays fall, and the lower cost draws in people who changed routes, times or modes, and new users who were previously deterred. The extra users then fill the space and delays rise again until the cost is again close to its earlier level.

The loop is: more capacity lowers delay, lower delay raises demand, higher demand raises delay. The loop settles where the delay is low enough to be attractive but high enough to hold back further growth. How much of the new capacity is filled depends on how sensitive users are to delay.

Example

Say a highway carries 10,000 vehicles an hour at peak with an average trip of 40 minutes. Adding a lane raises capacity by 20% and cuts the trip to 30 minutes at first. If each 10 minutes saved attracts 1,500 extra vehicles an hour, traffic rises to 11,500 and delays increase, until the trip is back near 36 minutes. The lane gave a 4 minute lasting gain rather than the 10 minutes first seen, and a much larger number of vehicles use the road. These numbers are an illustration.

How to model it in Machinations

Use a Pool for traffic on the road and a Register for the delay, which rises with the load relative to capacity. A state connection from the delay controls the Source of new users, raising it when delay falls. Make capacity and user sensitivity Levers, then Play and compare delay before and after a capacity increase.