UK DfT Rail Passenger Demand Forecasting Survey
A Mason, J Segal, MVA Consultancy, UK; N Fleming, Department for Transport (Rail Division), UK
Research was undertaken on the behalf of the UK Department for Transport to understand better the determinants of rail passenger demand in Britain. The conclusions of the study will be used to develop a thorough and consistent forecasting framework.
The current methodology for forecasting rail passenger demand in GB is provided by PDFH. The PDFH approach underpins the forecasts produced by the Network Modelling Framework which has been developed to assess options for the High Level Output Specification. Research was undertaken for the British Department for Transport to better understand the determinants of rail passenger demand in Great Britain, reviewing the PDFH methodology and comparing it with multi-model forecasts. This work will culminate in the development of a consistent demand forecasting modelling framework to meet the various needs of the Department for Transport. These needs include short, medium and long term forecasts at potentially a significant level of disaggregation and cover time periods up to 30 years (or more) for long term strategic planning and major projects. The framework must be able to address global changes such as the economy & fares and local effects which might include new developments, timetable changes and new stations.
Accurate demand forecasts are the principal foundation for the decisions that the DfT has to make. These cover long-term infrastructure investment; investment in new rolling stock; platform lengthening; franchise specification and evaluation; and other schemes. Because of the wide range of uses made of the forecasts, it is essential to have an understanding of the strengths and weaknesses of the various forms of model and the specific models themselves.
The study had three main workstreams:
?Þ Review of existing rail forecasting mainly based on PDFH which had been regularly updated since it was first prepared by British Rail in 1986;
?Þ Specification of how a detailed demand dataset should be compiled, incorporating rail data from a variety of sources: ticket sales, surveys, local multi-modal data and how this would be combined with demographic and economic data to enable robust econometric analysis in a later study;
?Þ Comparison of rail forecasting methods with multi-modal methods, both on theoretical grounds and of the forecasts from four multi-modal models with those of the current rail forecasting. Advice is provided on how to reconcile these.
The final element was the preparation of scope of further research to be undertaken in phase 2 of the study.
Association for European Transport