Valuing crowding in public transport: Implications for cost-benefit analysis

https://doi.org/10.1016/j.tra.2016.06.025Get rights and content

Abstract

This paper investigates the valuation of crowding in public transport trips and its implications in demand estimation and cost-benefit analysis. We use a choice-based stated preference survey where crowding levels are represented by means of specially designed pictures, and use these data to estimate flexible discrete choice models. We assume that the disutility associated with travelling under crowded conditions is proportional to travel time. Our results are consistent with and extend previous findings in the literature: passenger density has a significant effect on the utility of travelling by public transport; in fact, the marginal disutility of travel time in a crowded vehicle (6 standing-passengers/m2) is 2.5 times higher than in a vehicle with available seats. We also compare the effects of different policies for improving bus operations, and the effect of adding crowding valuation in cost-benefit analysis. In doing that, we endogenise the crowding level as the result of the equilibrium between demand and supplied bus capacity. Our results indicate that important benefits may be accrued from policies designed to reduce crowding, and that ignoring crowding effects significantly overestimate the bus travel demand the benefits associated with pure travel time reductions.

Introduction

Fast growing transport needs are a common concern for urban areas in both the developed and developing worlds. To address this issue, many cities have already implemented improved high-capacity transit systems (BRT, tramway or Metro). Often these systems are designed using an engineering standard of six (and sometimes more) passengers/m2 for the average supplied capacity. This design standard is an average across all vehicles of a service during the peak period, which is exceeded in a significant fraction of the operating buses/trains in some route segments. For instance, in Santiago, Chile, the average density across all trains in the most loaded segment during the morning peak hour exceeds 6 passengers/m2. As many individuals are not willing to use the system under such crowded conditions, they choose travelling by car or shift to car as soon as it becomes available. This prevents public transport modal shares from growing, increasing congestion and emissions. Moreover, although passenger density below a maximum design threshold of say 6 passenger/m2 may not be considered problematic at the design level, it is relevant because crowding may influence users’ preferences even for low levels of passenger density.

Crowdedness is usually left aside in most public transport demand models used for strategic planning. When planners evaluate transit network improvements, such as bus exclusive lanes, their goal is to increase the demand for public transport. Usually, this approach neglects the negative effects of crowding caused by the new induced demand. Notwithstanding, Tirachini et al. (2013) discussed its effects on operating speed, waiting time, travel time reliability, route and bus choice, and optimal levels of frequency, vehicle size and fare. The need for a more detailed understanding of crowding on travel decisions and its impact on project evaluation or cost-benefit analysis (CBA) is becoming an urgent priority. In this paper we focus on the valuation of crowding and its effect on mode-choice modelling. We estimate bus travel demand using a mode choice model that includes crowding effects and analyse the impact of using a wrong model, without crowding effects, in estimating demand and users benefits.

The general objectives of this paper are two: (i) to measure the willingness-to-pay (WTP) for crowding reductions in existing transit systems, and (ii) to study empirically the implications for CBA of making the demand for public transport endogenous with respect to the crowding level. The study is based on data from Santiago, Chile.

Most work addressing the valuation of crowding in public transport systems has used choice-based stated preference (SP) methods (e.g. Li and Hensher, 2011). But Guerra and Bocarejo, 2013, Haywood and Koning, 2015 applied contingent valuation to find the willingness to pay (WTP) for reducing overcrowding in the Bogota bus system and in the Paris Metro system, respectively. Li and Hensher (2011) reviewed public transport crowding valuation research, focusing on studies conducted in the UK, USA, Australia and Israel. Most studies have used logit models with SP data from commuters, and focused mainly on in-vehicle congestion costs. Nevertheless, Douglas and Karpouzis (2005) also estimated crowding costs at the platform (related to waiting time) and in the access-way/entrance to train stations (related to walking time). The way crowding is represented in SP experiments is highly relevant. Wardman and Whelan (2011) suggest that passenger density is a better indicator of in-vehicle congestion, given that a same load factor may have different levels of crowding across different types of vehicles/wagons with varying seat composition.

Our study was performed within a mode choice SP framework. In the choice experiment respondents had to choose between two transport modes, which could be bus, Metro or car. Each alternative was described by a number of attributes (e.g. cost, travel time, waiting time), and one of them was related to crowding. Specifically, pictures depicting passenger densities on board of vehicles served to represent the level of crowding. Valuations were derived from the estimation of mixed logit (ML) models (Train, 2009) using these data.

To explore the effects of crowding valuation in CBA results, three transport policies - typically proposed for improving bus corridor operations - were modelled: increasing bus frequency, increasing vehicle capacity, and building exclusive bus lanes. We used the estimated modal choice model to solve the equilibrium problem for bus demand (induced by the dependence of the bus utility function on crowding levels that, in turn, depend on bus demand). By doing so, we identified the pure effect of each policy and the effect of endogenous crowding levels in CBA. We found, for instance, that increasing bus travel times overestimated demand and user benefits if the endogenous effect of crowding was not taken into account.

The rest of the paper is organised as follows. Section 2 presents the SP survey experimental design and the information collected. Section 3 discusses our discrete choice modelling approach and presents the main estimation results. Section 4 discusses the effect of including crowding on the cost-benefit analysis of three common measures to improve the performance of a bus corridor. Some final comments are given in Section 5.

Section snippets

Survey design

Prior to the experimental design, we conducted focus groups that served to define which attributes would be most important to consider and which could be their levels of variation. Alternatives were finally described by six attributes: transport mode, travel time, travel cost, average waiting time, waiting time variability (coefficient of variation), and crowding level inside the vehicle (bus or train).

The experimental design in SP surveys is represented by a matrix that summarizes the choice

Model specification

The framework for our model specification is random utility theory. In the context of the choice of transport mode, the theory can be summarized in the following assumptions about individual behaviour (Ortúzar and Willumsen, 2011, Chapter 7).

  • There is a (finite) set of transport alternatives, mutually exclusive, for the individual’s trip.

  • Individual preferences for the alternatives can be represented by a utility function that depends on attributes of the alternatives and individual’s

Implications for cost-benefit analysis

In this section the impact of including the effect of overcrowding in the CBA of a project to improve public transport operations is discussed. For this, we consider the case of a bus corridor operated by a single bus line. Initially, buses operate on a street with mixed traffic (cars and buses), can accommodate 100 passengers and have a frequency of 15 buses/h. Total demand in the corridor is 5000 passengers/h, and the travel alternatives are bus and car. The analysed measures for improving the

Final comments

This paper values the effect of crowding in public transport using data from a stated preference survey. The level of crowding was measured as in-vehicle standing passenger density and presented to respondents by means of appropriately designed pictures. We used flexible discrete choice models to value crowding and specified modal utility functions where passenger density increased the effect of travel time on utility. Thus, we assumed interactions between passenger density and travel time. The

Acknowledgments

We wish to acknowledge the support of the Institute in Complex Engineering Systems (ICM: P05-004F; FONDECYT: FB016), the All Latitudes and Cultures BRT Centre of Excellence funded by the Volvo Research and Educational Foundations, the Centre for Sustainable Urban Development, CEDEUS (CONICYT/FONDAP/15110020), and FONDECYT Project N° 3140327. The paper benefited greatly from the constructive criticism of two outstanding referees; we are really thankful for their insightful comments.

References (36)

  • Batarce, M., Galilea, P., 2013. Cost and fare estimation for the urban bus transit system of Santiago. In: 92nd TRB...
  • M. Batarce et al.

    Use of mixed stated and revealed preference data for crowding valuation on public transport in Santiago, Chile

    Transp. Res. Rec.

    (2015)
  • Ben-Akiva, M., Morikawa, T., 1990. Estimation of travel demand models from multiple data sources. In: Proceedings 11th...
  • M. Ben-Akiva et al.

    Combining revealed and stated preferences data

    Mark. Lett.

    (1994)
  • BRT Centre of Excellence, EMBARQ, IEA and SIBRT, 2015. Global BRT Data, Version 3.4. Available at:...
  • J. De Cea et al.

    Transit assignment for congested public transport systems: an equilibrium model

    Transp. Sci.

    (1993)
  • A.C. DeSerpa

    A theory of the economics of time

    Econ. J.

    (1971)
  • Douglas, N., Karpouzis, G., 2005. Estimating the passenger cost of station crowding. In: Proceedings of the 28th...
  • Cited by (89)

    • Satisfaction measurement in the sustainable public transport of Budapest

      2024, Transportation Research Interdisciplinary Perspectives
    View all citing articles on Scopus
    View full text