We analyzed how the presence of traffic lights, as a cause of time loss, is taken into account by drivers in planning their route through urban areas. Our hypothesis was that routes with fewer traffic lights are preferred even if the probability of having to stop at those lights is high and the waiting time at the red light is long. We carried out a questionnaire-based study in which car drivers (n = 194) chose the route they preferred from pairs of hypothetical itineraries. The binary dependent variable was the type of route chosen: either a route containing fewer lights at which being forced to stop was highly probable, or a route containing more lights at which being forced to stop was far less probable. We found that the number of traffic lights was the preferred criterion, and that this preference could sometimes induce non-optimal route choices. Red- and green-light durations were also used as choice criteria. However, manipulation checks showed that participants did not estimate the probability of being forced to wait vs. being able to go through the light. We concluded that they estimated a threshold of acceptable waiting time at red lights depending on the number of traffic lights along the itinerary.
