Demand forecasting for new local rail stations and services

Passenger rail services in Britain are undergoing something of a revival. Since 1970 about 200 new stations have been opened on publicly owned railways in Britain, and British Rail (BR) have developed a number of new services. Before a new service or a new station can be evaluated a key question has to be posed: how many people will use it? An immediate problem is that BR's usual approach to demand forecasting is not suitable for predicting the effects of non-marginal changes in levels of service at a local level. This is discussed in section 2. So the Institute for Transport Studies at Leeds University has been working since 1982 on the development of suitable forecasting approaches. In the course of this work a variety of approaches have been developed and tested. These include a simple Trip Rate Model (TRM), a direct demand model that we have entitled an Aggregate Simultaneous Model (ASM), and a disaggregate Mode Choice Model (MCM). These models are all based on actual behaviour/choices (that is, Revealed Preference (RP) data) ; they are discussed in section 3. An alternative approach is based on asking people directly how often they would use a new facility. This is termed the Stated Inten tions (SI) approach. It is well known that this, if applied without adjustment, is likely to lead to overestimates of usage. A Stated Preference (SP) experiment has been devised in order to correct for this bias. This SI/SP approach is discussed in section 4.