The Impact of Tariff Differentiation on Time of Day Choice and Railway Demand in The Netherlands



The Impact of Tariff Differentiation on Time of Day Choice and Railway Demand in The Netherlands

Authors

M Kroes, A Mitrani, Steer Davies Gleave, UK; , L Weesie, NS Reizigers, NL; F Hofker, ProRail, NL

Description

Steer Davies Gleave and NS have developed a business planning model to evaluate the potential demand and revenue impact of tariff differentiation by time and location on the Dutch railway network, based on extensive SP research with rail users.

Abstract

NS Reizigers (NS) is responsible for the operation of the majority of rail services in The Netherlands. Because of poor peak revenue/cost ratios, NS is seeking ways to improve yields and/or reduce costs in the peak.

Whilst capacity varies very little during the day, at 35% averaged across the network, levels of peak demand are very high. Peak demand determines the need for rolling stock and personnel. This results in costs exceeding revenues as resources needed in the peak are not fully utilized at other times.

The current fares structure is predominantly based on distance travelled and applies between any combination of origin and destination stations. There are only two forms of tariff differentiation: 1st class travel at a premium of 70% and discounted travel outside the morning peak periods: 40% discount for holders of a VDU discount card (? 55 per year). At the moment any changes to prices of important ticket types are bound by regulations between NS and the government, however, as a result of problems with crowding, the lack of possibilities to increase capacity and the need for profitability, it is not inconceivable that future railway operators will have greater freedom to determine fares, and thus better manage demand and revenue.

In 2006 NS commissioned Steer Davies Gleave (SDG) to conduct research into potential customer responses to tariff differentiation. The key area of interest was differentiation by time and location, as this would potentially offer the greatest scope for crowding relief at peak periods, and the opportunity for exploitation of off-peak capacity.

The introduction of tariff differentiation is likely to have implications in terms of which product (season ticket or discount card), if any, a passenger may buy. Given ownership of a specific product, it is also likely to affect how and when they choose to make specific trips. Because of the complexity of the research the work was carried out in a number of phases. Firstly focus group research was conducted to gauge possible reactions towards potential changes in the fare structure. To further explore the issues identified and to understand their relative importance a large internet based market research survey amongst rail users was undertaken. The third and key component of research consisted of undertaking Stated Preference (SP) research to provide choice models for product and ticket choice.

Internet based SP research with rail customers was undertaken in June 2006 and over 4500 completed interviews analysed. Season ticket holders were presented with choices between an unrestricted season ticket at greater cost, and a restricted season ticket at a reduced price. Discount Card holders and non-product holders were presented with a product choice and a ticket choice SP exercise. The former offered respondents the opportunity to buy alternative discount cards compared to the current product available, whereas the ticket choice SP offered people the opportunity to choose between alternative tickets (given that they had or not had a discount card) offering the choice between a more expensive but fully flexible ticket and a cheaper ticket with restricted validity periods.

As a final step a tariff structure evaluation model was developed, an ?equilibrium? model forecasting the take-up of different products and likely decisions on when to travel for a given tariff structure. It forecasts demand in terms of annual revenue, journeys and passenger km, as well as changes in demand for different time periods and days of week. The model is structured in two layers:

? A ?product? choice model, forecasting the choices made by individuals about whether or not to purchase a season ticket, VDU card or any alternative product where the decision is made on the basis of an assessment of future rail journey requirements over a year; and
? A ?ticket? choice model, forecasting the choices made for individual trips, given the forecast ownership of season tickets, VDU cards and new products among different segments of the population.

Each customer?s product choices and ticket choices are forecast using choice models developed from the Stated Preference research. Ticket choices are forecast conditional on products held, considering customers? ideal departure times, and evaluating the trade-off between paying the appropriate price for travelling at that time or possibly paying less to travel at an alternative time. The inconvenience of having to make changes in departure times is evaluated (again drawing on the Stated Preference research) and included with the fare paid in a measure of generalised cost.

Publisher

Association for European Transport