Modeling the Future Demand for Electric Cars: Assessing the Influence of Information and Awareness of Survey Respondents on the Model Results
J Stark, C Link, Institute for Transport Studies, BOKU, AT
An overestimation of the market penetration of electric cars can be laid back to methodological weaknesses of the research design applied. The paper is focused on methodological precautions of the research design to achieve more reliable estimates.
Recent studies in Europe treat the topic of the actual and future market potential of electric cars. They mainly show a high acceptance of electric cars - what is not reflected by current sales figures. An overestimation of the market penetration can be laid back to methodological weaknesses of the research design applied: A major source of misjudgment could be the respondent?s lack of awareness of travel behavior restrictions caused by the use of electric cars. From the technological point of view electric cars offer restraints especially in terms of the driving range and recharging time. In advance of a purchase decision people would inform about these restrictions and could reject this type of car technology. That means that there is a discrepancy between the supposed behavior of the respondents in the survey and their real behavior. This can finally lead to an overestimation of the market potential. Therefore, an important task in preparing surveys on electric car purchase is to find a way to present the restrictions to the respondents without influencing their answers in an inadmissible way.
The paper focuses methodological precautions of the research design to achieve more reliable estimates. It is based on a stated preference survey that was conducted to assess the future market penetration of electric cars. The survey was conducted within the research project "Smart Electric Mobility" (SEM) in Austria in the year 2011 by the Institute of Transport Studies (University of Natural Resources and Life Sciences Vienna). Car purchase experiments were carried out with 230 persons. The experiments were based on the actual car stock of the household. Respondents were asked to choose between the existing, an electric, a hybrid car or none of these. The alternatives differ in up to seven attributes. If an electric car was chosen the respondents were faced with the resulting restrictions of their purchase decision; this was made with the help of "mode choice experiments": Based on recent long-distance car trips (>150 km) necessary adaptations in travel behavior caused by the recharging of the batteries were shown. Following the mode choice experiments, people had the possibility to revise their purchase decisions. Such feedback loops are not included in other studies.
Within the research project SEM utility functions for the different car technologies were calculated using a multinomial logit approach. They feed into a market model. Assessing different technological developments and framework conditions, the market potential of electric cars can be predicted. For an average technological development a market share of electric cars on the Austrians new cars market of 5.7 percent is calculated for the year 2025. This prediction is based on the revised choice of the respondents after an electric car was chosen and they were made aware of the following restrictions within the mode choice experiment. It could be analysed that about 40 percent of the respondents chose an electric car as their first option. It is assumed, that an analysis of the "first answer" reveals a significant higher market share prediction.
The paper is focused on the methodological effect if the future market potential of electric cars is based on the first purchase decisions or on the revised purchase decisions of the respondents; therefore, the results of original and revised answers will be compared. Further important methodological features of the survey approach, like measures to avoid order effects within the questionnaire, are presented.
Association for European Transport