Placing a Value on Overcrowding and Other Rail Service Quality Factors



Placing a Value on Overcrowding and Other Rail Service Quality Factors

Authors

J Baker, N Myers, Mott MacDonald, UK; P Murphy, Faber Maunsell, UK

Description

This paper looks at a Stated Preference (SP) study undertaken to assess how West Midlands? rail passengers (rail users and non rail users) value and trade off service reliability, frequency and travel in less crowded conditions.

Abstract

The West Midlands Passenger Transport Executive (CENTRO) are responsible for improving passenger services and accommodating growth, in line with their Business Plan and wider Centro objectives, within the constraints of Network Rail views on available capacity.

This paper looks at a Stated Preference (SP) study undertaken by Mott MacDonald and Faber Maunsell to assess how West Midlands? passengers (rail users and non rail users) value and trade off rail service reliability, frequency and travel in less crowded conditions.

A comprehensive assessment was conducted which involved a review of other research carried out looking at these service attributes including evidence from sources such as Passenger Demand Forecasting Handbook. As well as providing evidence of current valuations of these attributes, the review also provided an insight into how these attributes have been represented in other SP studies.

Existing data sources were also reviewed to explore whether any robust data, either cross section or time series, on overcrowding, reliability and frequency exists. As part of this review, an internal survey with local Birmingham-based Faber Maunsell and Mott MacDonald office staff was carried out. This was mainly to collect information on which parts of the network those issues are important and to collect people?s experiences and attributes towards these issues and their use of rail. The results of the survey showed that overcrowding on trains happens more regularly than unreliability or poor frequency and that overcrowding has got much worse over the last 12 months when compared to unreliability and poor frequency. Interestingly however, the findings highlighted that change in behaviour and mode of transport (transfer to car/bus) was more likely to occur if unreliability and poor frequency increase than if overcrowding increases. This was further reflected by respondents? current behaviour where the majority said hey had not used other modes of transport as a result of overcrowding.

Qualitative research using focus groups (with users and non users) were also carried out to explore how travellers think about, and trade off, overcrowding, reliability and service frequency and to test different methods for exploring these issues in SP experiments. The results of this survey gave some valuable insights which were subsequently fed into the SP survey. As with the internal survey, overcrowding was ranked as the most important issue among rail users. For all, overcrowding comes down to one thing ? not having a seat. However, user participants did not always see standing as overcrowding but they defined different levels of standing which defined overcrowding: people standing in centre aisles, people standing in vestibule, people standing so that access to doors/aisles is blocked and, standing without being able to move as a result of no room space. For most participants, overcrowding is mainly as a result of unreliability combined with no extra carriages added as a result and unsurprisingly, unreliability was where they felt it was most important to make improvements.

A full Stated Preference survey was carried out with around 800 rail users to establish the penalties for, and trade offs between, crowding, reliability and frequency. The survey also collected detailed information on rail user journeys, attitudes and demographic information. Self-complete post back questionnaires were distributed at 6 stations in the West Midlands. The resulting models were well defined and presented plausible values which compare well with other research carried out in this area. The SP models provided values of four different crowding levels and a value of expected lateness which are very similar to that currently recommended by the Passenger Demand Forecasting Handbook.

The paper will present the results from this study as well as comments on the lessons learnt from the development and application of the chosen methodology.

Publisher

Association for European Transport