Household demand and willingness to pay for clean vehicles

https://doi.org/10.1016/j.trd.2007.03.001Get rights and content

Abstract

This paper examines the factors and incentives that are most likely to influence households’ choice for cleaner vehicles in the metropolitan area of Hamilton, Canada. Data collection is based on experimental design and stated choice methods through an Internet survey. Choice alternatives included a conventional gasoline, a hybrid and an alternative fuelled vehicle. Each option is described by a varying set of vehicle attributes and economic incentives, customized per respondent. Controlling for individual, household and dwelling-location characteristics, parameters of a nested logit model indicates that reduced monetary costs, purchase tax relieves and low emissions rates would encourage households to adopt a cleaner vehicle. On the other hand, incentives such as free parking and permission to drive on high occupancy vehicle lanes with one person in the car were not significant. Furthermore, limited fuel availability is a concern when households considered the adoption of an alternative fuelled vehicle. Finally, willingness-to-pay extra for a cleaner vehicle is computed based on the estimated parameters.

Introduction

The automobile, the dominant mode of urban transport, generates one sixth of the world’s greenhouse gas emissions. It is also a major contributor of local air pollutants, such as particulate matter (PM), carbon monoxide (CO), nitrogen oxides (NOx) and volatile organic compounds (VOC) (Fenger, 1999). In Canada, for example, transportation accounts for approximately 52% of all NOx emissions, 40% of CO, 20% of VOCs and 5% of PM (Transport Canada, 2004). Alternative fuelled vehicles (AFVs) are expected to produce approximately 40% lower emissions compared to gasoline vehicles, while maintaining the flexibility and performance of private motorized travel. Under special conditions, future long-term technologies such as electric, hybrid-electric and fuel cell vehicles are expected to achieve emission reductions up to 80% of present day vehicles (Wang, 1999).

Nevertheless, to achieve substantial penetration of AFVs in the market requires large investments in infrastructure for vehicle and fuel production, and an expansion of the network of refuelling facilities (MacLean et al., 2004). Alongside the development of infrastructure it is important to evaluate the willingness of consumers to adopt cleaner vehicles. Hence, it is important that AFVs represent attractive options with performance characteristics comparable to those of conventional vehicles. Building an understanding on the factors and incentives that would likely encourage households to adopt AFVs would help to inform industry stakeholders and to develop policy interventions.

Several studies on the potential demand for AFVs have been conducted in California, while experience from other places is limited (Dagsvik et al., 2002). The majority of the studies were based on stated choices (SC) methods and experimental design techniques, taking the household or the individual as the unit of analysis (Louviere et al., 2000). A study by Adler et al. (2003) looked at three hypothetical vehicle options: a conventional gasoline, a hybrid-electric and a diesel vehicle. Analyzing responses from approximately 3,300 households, estimations of a nested multinomial logit (NMNL) model showed that fuel cost savings and incentives, such as tax-free purchases and free parking, would encourage Californian households to adopt a cleaner vehicle. In the Canadian context Ewing and Sarigollu (1998) developed a stated-choices experiment to examine the factors influencing the preferences of suburban commuters in Montreal when considering lower and zero-emission vehicles. In addition to vehicle attributes, they tested the importance of economic instruments, such as commuting costs and time, which were also significant. Estimates of using a multinomial logit (MNL) model revealed that if price and performance characteristics are comparable, there could be a high demand for clean vehicles.

This paper offers empirical results for the Canadian urban context. In contrast to the US examples that focused at the national and state levels, we concentrate at the urban scale, the metropolitan area of Hamilton in Ontario. In addition to vehicle attributes and household socio-demographic and economic characteristics, we introduce neighbourhood characteristics at the place of residence (e.g. urban form measures) as covariates of a household’s vehicle-type choices. Until recently, neighbourhood characteristics have been neglected in vehicle-type choice studies with the exception of Cao et al. (2006) who introduced neighbourhood characteristics in their analysis of conventional gasoline vehicle-type choice behaviour.

We look at whether the vehicle attributes and household characteristics that likely to influence preferences for clean-vehicle technologies in a Canadian urban area, whether there are there any urban form characteristics that may affect these preference, which policy initiatives are likely to encourage the adoption of cleaner vehicles and how willing consumers are to pay extra for such vehicles.

Section snippets

Experimental design

The focus is on the demand for short-term clean vehicle technologies, such as hybrid electric and a generic type of an alternative fuelled vehicle (e.g., natural gas). We use stated choices methods to considered three vehicle options: conventional gasoline, hybrid electric and alternative fuelled. Each option was labelled based on seven class/size categories of vehicles: subcompact, compact, mid-size, large, pick-up, van and sport utility vehicle (SUV).

The most significant attributes of

Data collection

Data collection through the Internet is a relatively new method, which can be implemented with low fixed-costs independently from the number of collected observations. Response times are considerably shorter than conventional methods such as mail or telephone, while it is not required to hire and train interviewing personnel (Dillman, 2000). With the completion of the Internet-survey, the collected data are immediately available for analysis because digital storage, management and coding occur

Discrete choice modelling: nested logit

Estimation of parameters was conducted using the Nested Logit (NMNL) model in order to avoid potential violations of the IIA property among vehicle options. The NMNL accommodates differential degrees of interdependence between subsets of alternatives in a choice set (Ben-Akiva and Lerman, 1985). Hensher and Greene (2002) offer two specifications of the NMNL model, namely RU1 and RU2, that can be estimated through LIMDEP Ver. 3.0. RU1 normalizes the lower level and RU2 the higher-level scale

Conclusions

Individual preferences play a key role in stimulating demand for clean vehicles. This paper has examined the factors most likely to influence the demand for Hybrid and AFVs using experimental design and stated choices methods through an Internet survey. The Internet-survey used offers a number of advantages compared to conventional methods of data collection, with reduced costs and time being the most important. While the full cost of the Internet survey was fixed at $6,000, a conventional

Acknowledgements

This research was funded by the City of Hamilton and the Natural Sciences and Engineering Research Council. The first author acknowledges the support of Alexander S. Onassis Benefit Foundation and the second is grateful to the Social Sciences and Humanities Research Council Canada Research Chairs program. We appreciate the comments and suggestions of Thomas Adler, Warren Kuhfelt and John Rose.

References (24)

  • M. Ben-Akiva et al.

    Discrete Choice Analysis: Theory and Application to Travel Demand

    (1985)
  • R.T. Carson et al.

    Experimental analysis of choice

    Marketing Letters

    (1994)
  • Cited by (346)

    • Dynamic consumer preferences for electric vehicles in China: A longitudinal approach

      2023, Transportation Research Part A: Policy and Practice
    View all citing articles on Scopus
    View full text