Key Factors Influencing the Decision to Buy an Electric Vehicle in Emerging Markets: A Comparison Between University Students’ Behavior in Italy and Argentina

Abstract

The transition from Internal Combustion Engine Vehicles (ICEVs) to electric vehicles (EVs) is a tricky issue, but particularly challenging in those markets that are characterized by high ICEV_S ownership rates and that may be identified as emerging markets due to the small diffusion of the technology and of the related infrastructures. This study delves into modeling the willingness to buy an EV by young adults within the emerging Italian market, and to draw a comparison with the findings of a previous study carried out on the Argentinian market. To this aim a stated preferences survey was carried out at the University of Salerno (Italy), and then compared with the same survey designed and carried out at the University of Cordoba (Argentina). The Random Utility Paradigm was employed to model the preferences of the respondents, by focusing on the influence of instrumental attributes and psycho-attitudinal factors through a Mixed Binomial Logit formulation and Hybrid Choice model with Latent variables, respectively. The findings suggest that attitudes and perceptions play a substantial and similar role in emerging markets, with only slight differences across the two markets.

Appendix: Experimental Set-Up - Survey Description

This paper presents the findings of a study conducted Argentina and Italy, investigating the factors influencing the adoption of electric vehicles (EVs).

In light of the ongoing growth of the electric vehicle market in Italy and Argentina, a forward-looking study based on data from a revealed preference survey appeared premature. As a result, the willingness to purchase an EV was investigated through a stated preference (SP) experiment. The survey was conducted online in 2019 through a web-based questionnaire that took no more than fifteen minutes to complete and that was administered to students at University of Cordoba in Argentina and to the University of Salerno in Italy.

The study employed a specially designed SP questionnaire consisting of five sections and included an initial filter section to assess participants’ intention to purchase a car within the next 18 months. Those who did not express an interest were directed to the exit section of the survey.

The first two sections of the survey aimed to collect observable variables that could be relevant in characterizing potential future purchasers of electric vehicles, in terms of their socioeconomic backgrounds (gender, age, household size, and number of cars owned), as well as their usual travel habits (fuel type of the owned car, annual kilometers driven, and primary use of the vehicle, i.e., urban or extra-urban). This information could be included in both the specification of the utility function and in the specification of the latent variables.

The third section (Fig. 1) explored the influence of two attitudes (towards technical features and the environment) on respondents’ willingness to purchase an electric vehicle EV, using a five-point Likert [33] scale (from strongly disagree to strongly agree).

Fig. 2.

Third survey section about attitudes toward the environment and technical features

In the fourth section, participants were presented with different choice scenario for purchasing a new vehicle, with a comparison between an electric vehicle (EV) and a conventional internal combustion engine vehicle (ICEV). This section incorporated an educational module aimed at clarifying the basic principles of electric vehicle technology. The aim was to ensure that the possible propensity to purchase an electric vehicle was based on information that was both reliable and equally accessible to all survey participants. The main features discussed were: zero tailpipe emissions, quiet operation, possible charging locations (private or public), charging times (ranging from 30 min to 8 h), range (minimum of 200km for standard cars, enough for a week of urban trips) reduced maintenance costs (due to fewer moving parts), motor information (same power as combustion engines, but lighter). Then, each respondent was faced with choices.

Each respondent was presented with two cars: a Renault Clio (alternative A, ICEV) and a Renault Zoe (alternative B, EV). The cars were from the same manufacturer, similar in size, color and main features to minimize the impact of these characteristics on the choice decision (Fig. 2.a). It was hypothesized that the interviewee had enough budget to purchase either car, and a travel scenario was set (urban and 40 km/day).

The monthly cost of the conventional vehicle was described (see Fig. 2. b), and then respondents were presented with different scenarios with varying monthly costs of the electric vehicle relative to the conventional one. The five scenarios were equal monthly cost, + 10%, + 20%, + 30%, and + 40%.

More details on the survey description are provided in [3].

Fig. 3.

Fourth survey section: (a) Comparison between Renault Clio and Renault Zoe; (b) Example of a typical choice scenario

In the fifth section, we gathered insights from users regarding their perception of advantages and disadvantages of electric vehicles[3, 34] which could influence their inclination to buy such vehicles. Similarly to the third section, using a five-point Likert scale, respondents were asked to express their level of agreement or disagreement with some indirect statements (Fig. 3).

Fig. 4.

Fifth survey section: (a) Investigation of perception of EV advantages; (b) Investigation of perception of EV drawbacks.

Around 1,000 participants from two universities took part in the study, and each participant encountered a varying number (ranging from three to five) of scenarios presented in a randomized order. This approach was employed to prevent any serial correlations and ensure the independence of all choices made by participants. The survey yielded 467 valid responses from the University of Cordoba and 292 from the University of Salerno, resulting in a total of 2,148 observations for the Argentinian survey and 1,460 for the Italian survey.