Electric vehicles adoption: Environmental enthusiast bias in discrete choice models

A Stated Choice (SC) survey, employing a Best-Worst choice design, was administered to 440 households in Perth, Australia as part of a major investigation into consumer preferences and attitudes towards electric vehicles. It was noted that 48 (10.9%) respondents chose EV as their best/most preferred option across all six choice replications. We hypothesise that for most of these respondents their choices reflected their desire to present themselves in a favourable light, with social desirability biasness manifested in non trading behaviour. There were also 24 (5.5%) respondents who chose EV as their worst/least preferred option. We hypothesise that for these respondents lack of interest Or confidence in the new technology and inertia may have driven their decisions. The paper offers demographic and psychographic profiles of non-traders facilitated by additional items being included in the experiment. While there was little difference between the demographic profiles, the attitudinal scores of the non-traders were significantly higher than for traders, which may indicate social desirability. Non-traders (Best) scored significantly higher on environmental concerns and subjective norms and were more likely to rate their intention to purchase and use an EV higher. Conversely, non-traders (Worst) had the lowest environmental concerns and subjective norms. From a choice modelling perspective, keeping non-traders in the estimation biases the taste parameters and therefore the willingness-to-pay (WTP) measures. However, when incorporating the worst alternatives into the choice models, the 'social desirability' non-traders do appear to be making decisions based on the attributes, which is consistent with the rest of the sample. (C) 2017 Elsevier Ltd. All rights reserved.