Abstract
This paper investigates the role that enhanced service quality introduced into a deregulated market has in improving the experience of bus travel by a sample of passengers in the Tyne and Wear area of England. A generalised ordered choice (GOC) model that accounts for preference heterogeneity through random parameters, as well as heteroscedasticity in unobserved variance, and random parameterisation of thresholds, is implemented to identify sources of influence on the overall experience of bus travel in the presence and absence of the quality-enhanced treatment of service. The GOC model is contrasted with a standard ordered logit model, and the marginal effects associated with the preferred GOC model are derived for each influencing attribute, taking into account the various ways in which each influence contributes to the utility associated with each level of bus experience. The paper supports a view that the introduction of quality improvements, via a Quality Bus Partnership, does contribute non-marginally to an increase in a positive bus experience, and signals a way forward through cooperative intervention, to grow patronage. Knowing which attributes successfully deliver a more positive experience (and those that do not) means that resources are effectively targeted at the aspect of service provision which will increase patronage and therefore revenues, satisfying the objectives of both the bus operator and the local authority partner.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Statistics derived from Transport Statistics 2008, available from the Department for Transport, UK.
We recognise that there is an extensive literature on service quality including the developments by Hensher (e.g. 2002) briefly mentioned in the next paragraph; however in the current application it is sufficient to recognise the popular use of a 5-point scale such as the one herein. Readers interested in the broader literature on service quality might refer to the classic text by Zeithaml et al. (1990).
For example, it is not valid to assume that the level of satisfaction associated with the difference between very satisfied (coded as 5) and satisfied (coded as 4) is the same on a preference scale as the difference between unsatisfied (coded as 2) and very unsatisfied (coded as 1).
Several normalisations are needed to identify the model parameters. First, given the continuity assumption, in order to preserve the positive signs of the probabilities, we require μ j > μj-1. Second, if the support is to be the entire real line, then μ-1 = −∞ and μ J = + ∞. Finally, assuming that x i contains a constant term, we will require μ0 = 0. With a constant term present, if this normalisation is not imposed, then adding any nonzero constant to μ0 and the same constant to the intercept term in β will leave the probability unchanged. Given the assumption of an overall constant, only J − 1 threshold parameters are needed to partition the real line into the J + 1 distinct intervals. The identification issues associated with unordered choice models in selecting the relevant base alternative-specific constant (as so eloquently shown in Joan Walker’s research—see Chiou and Walker 2007), is not an issue in ordered choice models.
The ‘Superoute’ network was the result of a comprehensive network review. Although there is no way of knowing whether each sampled person had ever used both a ‘Superoute’ and ‘non-Superoute’ services, the current study sought responses on the service at the time of survey.
Self-selection bias occurs if there was more than one alternative available and the empirical analysis was undertaken only on specific aspects on the chosen alternative (a classic example is when there is a choice of automobile type and the evidence available is only on the chosen type). The analysis herein is conditional on a respondent using a bus and thus the potential self selection bias associated with non-bus users does not apply in this case.
New code was developed in Nlogit 5 to estimate the generalised ordered choice model.
Available on request from the authors.
This holds for continuous variables only. For dummy (1,0) variables, the marginal effects are the derivatives of the probabilities given a change in the level of the dummy variable.
References
Bhat, C.R., Pulugurta, V.: A comparison of two alternative behavioral mechanisms for car ownership decisions. Transp. Res. Part B 32(1), 61–75 (1998)
Chiou, L., Walker, J.L.: Masking identification of discrete choice models under simulation methods. J. Econom. 141, 683–703 (2007)
Eluru, N., Bhat, C.R., Hensher, D.A.: A mixed generalized ordered response model for examining pedestrian and bicyclist injury severity level in traffic crashes. Accid. Anal. Prev. 40(3), 1033–1054 (2008)
Fairhurst, M., Edwards, D.: Bus travel trends in the UK. Transp. Rev. 16(2), 157–167 (1996)
Greene, W.H.: Econometric analysis, 6th edn. Prentice Hall, NJ (2008)
Greene, W.H., Hensher, D.A.: Modeling ordered choices: a primer and recent developments. Cambridge University Press, Cambridge (2009)
Hensher, D.A., Prioni, P.: A service quality index for area-wide contract performance assessment. J. Transp. Econ. Policy 36, 93–113 (2002)
Hensher, D.A., Stopher, P.R., Bullock, P.: Benchmarking and service quality at a market segment level. Transp. Res. A 37, 499–517 (2003)
Kuby, M., Barranda, A., Upchurch, C.: Factors influencing light-rail station boardings in the United States. Transp. Res. Part A Policy Pract. 38(3), 223–247 (2004)
LEK/Commission for Integrated Transport (CfIT) (2002) Obtaining best value for public subsidy for the bus industry, London
Mackie, P.J.: Principles of public policy for bus services. In: Grayling, T. (ed.) Any more fares?—delivering better bus services 19–39 2001. IPPR, London (2001)
Matthews, B., Bristow, A.L., Nash, C.A. (2001) Competitive tendering and deregulation—a comparison of bus markets in London and the British Metropolitan areas Seventh Conference on Competition and Ownership in Land Passenger Transport, Molde
Reeves, C.A., Bednar, D.A.: Defining quality: alternatives and implications. Acad. Manage. Rev. 19(3), 419–445 (1994)
Rosenbloom, S., Clifton, K.: The puzzle of income, race, and density: preliminary evidence on transit use from the 1991 American housing survey. J. Public Transp. 1, 87–102 (1996)
Stradling, S.G., Anable, J., Carreno, M.: Performance, importance and user disgruntlement: a six step method for measuring satisfaction with travel modes. Transp. Res. Part A 41, 98–106 (2007)
TRRL (2004) The demand for public transport: a practical guide. TRL Report 593
Zavoina, R., McElvey, W.: A statistical model for the analysis of ordinal level dependent variables. J Math Sociol 4, 103–120 (1975)
Zeithaml, V., Parasuraman, A., Berry, L.: Delivering quality service: balancing customer perceptions and expectations. Free Press, New York (1990)
Acknowledgments
The comments of three referees have materially improved this paper.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Hensher, D.A., Mulley, C. & Yahya, N. Passenger experience with quality-enhanced bus service: the tyne and wear ‘superoute’ services. Transportation 37, 239–256 (2010). https://doi.org/10.1007/s11116-009-9240-x
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11116-009-9240-x