Estimating the Demand Impacts of New Rolling Stock
R Sheldon, C Heywood, Accent, UK; A Meaney, N Robins, Oxera, UK; M Wardman, ITS, University of Leeds, UK
To determine the impact of new rolling stock on passenger demand an innovative study was undertaken for the PDFC using two research methods: ?event analysis? and market research.
The members of the Passenger Demand Forecasting Council wanted to develop an improved understanding of the effects of new and refurbished rolling stock on rail demand and recommendations on the values for application to the PDF Handbook and the Rail Industry Forecasting Framework (RIFF).
Two methods were used: ?event analysis? and market research. The objective of the event analysis was to determine the demand impact associated with new or refurbished rolling stock based on revealed preference evidence from ticket sales data. The main objective of the market research was
to collect direct evidence of any impact that new rolling stock has had on passenger demand. The research was undertaken on flows that represented London suburban, InterCity and regional.
For the ?event analysis? demand models were estimated before the change in the rolling stock Data for the explanatory variables after the change in stock, and the parameters from the demand models, have been used to suggest what demand may have been in the absence of any change in the rolling stock (?the predictions from the demand model?). The predictions have been compared with actual demand, and the difference between actual and predicted demand may be attributable to the change in rolling stock, controlling for changes in real fares, GJT, delay and real GVA.
The market research was undertaken on routes also covered by the event analysis. The method was the random distribution and collection of self-completion questionnaires on-train. The questionnaires asked respondents to state why they were making the journey and to say whether the new rolling stock has had an impact on their use of the train. 1,335 questionnaires were administered, split between three routes.
The market research explored whether passengers had noticed the new rolling stock, and if they had, their attitudes towards different aspects of the stock.
Respondents were asked whether they had changed class of travel as a result of the new trains and whether the new trains had led them to change the number of journeys they made by rail on the route. The data on whether respondents had travelled more because of the new trains was then converted
into the impact on demand.
The ?event analysis? found that the change in rolling stock may lead to a small uplift in demand, relative to the level of demand that would have occurred in the absence of any change in rolling stock
The market research provided sensible looking results that compared well with the revealed preference models. This type of market research exercise could collect appropriate data very cost efficiently allowing demand impacts to be examined across a broader array of market segments.
Association for European Transport