Understanding the impact of fare complexity on rail demand


Date: January 2019

Background
The UK Rail Delivery Group (RDG) wished to examine the potential effects on passenger rail demand of a range of innovations to the fares regime.   The rail fare structure in the UK was widely considered at the time to be complex for train users. While it offered flexibility by including many different types of tickets, it was also perceived as causing confusion, leading some to a negative view towards rail travel.  The concern amongst RDG and its members was that the perceived complexity might be inhibiting demand, and that rail demand may therefore potentially be higher given a simpler fare regime.

The core objective for the study was to test this hypothesis by quantifying the impact of alternative fares structures on demand, including removal of Advance tickets, and pricing all advance tickets the same for a given journey. In addition, an increased complexity offer was also to be tested which would involve adding new Flexible Advance tickets (valid on the services immediately before or after the chosen service) to the fares regime.

Approach
The study used an innovative stated preference survey based around a choice experiment that was designed to mirror almost exactly the actual booking experience when buying tickets online, in all its complexity. The following gives an example of one of the choice screens shown to survey participants.  The buttons were all fully functional, while the times, costs and ticket types shown were all drawn from an experimental design created for the study.

Participants could choose from up to 531 different ticket type combinations for the outward and return legs of a trip and from up to 25 possible train services for each leg. The key design attribute was complexity, defined as the range of different Advance tickets on offer.

Following a successful pilot, the survey was applied to a sample of 1027 users and 179 non-users of the rail network on the London-Leeds route.   The core analysis consisted of a nested mixed logit model linking ticket type and train service choice.  The parameter estimates from this model were entered into a simulator tool calibrated using industry data on real-world ticket prices and sales levels.

Results
The results from the modelling suggested that, all else equal, reducing complexity by removing Advance tickets would lead to a substantial reduction of demand (11 to 45%, depending on route segment). Equalizing the price of Advance tickets for all train services was predicted to cause a smaller reduction (3–6%). By contrast, increasing complexity by adding new Flexible Advance tickets (valid on the services immediately before or after the chosen service) would increase demand by 4–15%. These findings ran counter to the hypothesis that simplifying the fare structure would lead to increases in demand for rail travel.

Academic paper