Assessment of Selected Determinants Affecting the Acceptance of the Development of Electromobility by the Private and Business Sectors—A Case Study in Portugal

Abstract

The energy transition requires widespread electrification of the transport sector. To promote the penetration of electric vehicles (EVs), it is essential to understand consumers’ perceptions and behavior, particularly regarding the main determinants of EV purchase and the acceptance of electric mobility (EM). With this aim, we focused on an industrialized city in Portugal, addressing the differences between the effective ownership of an EV and the acceptability of EM and between the domestic sector (DS) and the business sector (BS) through questionnaires. Our results indicate that sociodemographic variables are the main determinants of the purchase of EVs and the acceptance of EM in the DS. Men and higher income individuals are more likely to own an EV. On the other hand, younger generations are more likely to have high EM acceptance. Individuals who already own an EV are the ones that have the desire and economic means to do so, regardless of any incentives. Still, widespread market penetration of EVs requires incentives for individuals who desire to own one of these vehicles but do not have the economic power to do so. Additionally, the DS and the BS behave differently; hence, specially designed policies are needed.

1. Introduction

Climate change and its impact on the planet are issues of utmost importance for both society and policy makers. In view of the rapid evolution of these changes, various large-scale impacts are anticipated, such as changes in the disposition of communities, the deterioration of the health and safety of current and future generations, the rise in the average seawater level [1], deforestation [2], and catastrophic events [3,4,5].

Over the years, several Synthesis Reports of the Intergovernmental Panel on Climate Change (IPCC) have indicated that, in order to ensure an increase in global temperature below 2 °C compared to the pre-industrial period, a departure from the business-as-usual scenario (BAU) [6] is urgently needed, with major changes in current business models, the restructuring of different sectors of the economy, and the implementation of the concepts of innovation and sustainability. Additionally, the later the intervention, the higher the associated costs and technological, economic, social, and institutional challenges [7].

The energy sector accounts for about two-thirds of total greenhouse gas (GHG) emissions at the global level [8], to which fossil fuels are the main contributors. It is, therefore, necessary to restructure this sector, focusing mainly on the production of electricity through renewable energy sources (RES) [9]. In turn, the transport sector accounts for one-quarter of energy-related GHG emissions, of which 70% are related to road transport [10,11,12]. Passenger cars account for about half of the transport sector’s global energy use [8]. In Europe, these account for 83.4% of domestic transport and contribute to about two-thirds of total road emissions. On the other hand, commercial, light, and heavy vehicles represent about 13% of the total vehicles of the European continent and account for one-third of the total emissions of the transport sector [13]. Thus, the substitution of fossil fuels by greener and more efficient alternatives in the transport sector is essential, particularly through electrical vehicles (EVs) [14].

The technological advance of electric mobility (EM), its applicability, energy management, and energy storage systems represent crucial factors for the development of the energy transition [13,15], contributing to reductions in GHG emissions and energy consumption and improvements in countries’ energy security [8]. In addition, public health will benefit from the reduction in air pollution [15]. Several countries are planning to discontinue sales of traditional vehicles between 2025 and 2050 [16]. According to Hawkins [17], the combination of EVs with a European energy mix more based on RES will provide a 10–20% reduction in global warming potential. For the transport of goods, if electrification is possible, it could have a major environmental impact. In addition, companies will be able to benefit from greater economic benefits (offsetting the high initial costs with lower operating costs) and image improvements [13]. Still, currently, electric vehicles have higher purchase costs than traditional ones essentially due to the high costs of batteries and the electric drive train [18]. Despite this factor, EVs are increasingly popular among consumers, which is likely related to other factors besides the cost.

Considering the data provided and studied by Associação de Utilizadores de Veículos Elétricos (www.uve.pt), it is possible to simulate the operating cost of driving 100 Km for different types of cars. In

Table 1. Operational costs of traveling 100 Km for different types of vehicles.

Type of Vehicle	Average Consumption	Average Price	Average Cost for 100 Km
Diesel vehicle	7l/100 Km	1.953 €	13.67 €
Gasoline Vehicle	6l/100 Km	1.808 €	10.85 €
EV (domestic charge)	16 kWh/100 Km	0.400 €	6.40 €
EV (FCS 50 kW)	16 kWh/100 Km	0.220 €	3.52 €

, we present some cost simulations of traveling 100 Km for different types of vehicles.

Regarding the price of gasoline and diesel, the average prices of the different types of fossil fuels were used and their average was calculated. For the price of kWh for electric mobility, the cost of charging in the public charging network in Portugal (50 kW fast charging station (FCS)) was used. For the price of kWh for domestic electricity, the regulated market price was used (Last Resource Supplier—SU Eletricidade).

It is possible to observe in Table 1 that the operational cost of traveling 100 Km is substantially lower for the EV. However, it is important to note that the calculated values may change due to several variables, such as:

• Electricity supplier tariff;
• The chosen charging station, and the available power;
• The battery charge level at the time of charging;
• The temperature and the battery itself.

What we intended to show is that for normal and current use, considering the average fuel prices and the reference prices indicated by electricity suppliers, it will always be more economical to use an electric vehicle.

The European Commission has set a series of targets which include the elimination of atmospheric emissions from passenger transport by 2050 and from urban freight transport by 2030. Several European legislations have been put in place to promote the desired transition in the transport sector. For example, the Act on electromobility and alternative fuels, or the directive of 22 October 2014 (2014/94/EU) which imposed on European countries the obligation to transform the field of fuels [19]. Other relevant legislation includes the Euro 7 exhaust gas standards which will apply to all motor vehicles, imposing limits on the pollution arising from engine combustion. Hence, EM plays a key role in this transition, in particular through the introduction of 100% electric cars [13]. To achieve this goal, it is important to ensure public access to charging stations for EVs with an appropriate infrastructure network. Through several incentives, the use of EVs has shown rapid progress since their introduction. In 2016, about 1% of total car sales revenue corresponded to EVs [8]. In 2018, the global stock exceeded 5 million units, a 63% increase over the previous year. In 2021, 45% of electric passenger cars were in China, 24% in Europe, and 22% in the USA [20]. In addition, the number of two- and three-wheel EVs has also been increasing progressively [8]. Still, the spread of EVs in different international markets remains relatively low with very little uniformity across the globe.

The acceptability of EM is fundamental to a proper energy transition in the transport sector. This acceptability may be affected by several factors. These factors can be a combination of technological, regulatory, institutional, economic, cultural, and behavioral aspects. Consumers’ behavior is key to the penetration of EVs in the market. For example, apart from differences in costs, there are specific factors, such as environmental concerns, preference for the latest technologies, and status, among others, that may lead individuals to buy EVs instead of conventional ones. Fiscal and economic incentives may also play an important role. Additionally, it is possible that acceptance factors differ among sectors, for example, for the domestic and the business sectors.

Hence, a central part of this process will be to understand the social perceptions relating to EVs [21] and the responses of consumers, their preferences, motivations, and sociodemographic characteristics. This understanding allows, for example, the design of appropriate policy measures and the possibility of finding efficient policy solutions and exploring business innovation strategies [22]. According to Sovacool et al. [21], some aspects, such as occupation and household size, related to the role of consumers in the acceptance of EVs have been ignored in the literature.

Thus, this article aims to contribute to the understanding of consumers’ perception and acceptance of EVs and to the identification of factors that significantly influence their development. For a deeper analysis, we analyze and compare both the business sector (BS) and the domestic sector (DS) in Portugal in a case study conducted in the municipality of Felgueiras. We chose this municipality because it is a highly industrialized one, where it is possible to obtain a combination of responses from the DS and BS; furthermore, it has a relatively high number of EVs.

Our research questions cover the following aspects: Do sociodemographic characteristics affect the preference for EVs? What are the main incentives and hindering factors behind the preference for EVs? Do the domestic and business sectors display different behaviors?

The main expected differences between the DS and BS relate to individual preferences. The DS, composed of single individuals, may have a stronger component of personal preferences contributing to the acceptance of EM. Often, automobiles are seen as a sign of status or an extension of an individual’s personality. Economic and fiscal incentives are expected to be important for both cases. Other studies (e.g., Nogueira et al. [23]) have focused on the Portuguese case but covered distinct topics. Our contribution to the literature is twofold. Firstly, we compare the results regarding the effective purchase of an EV and the acceptability of EM. This can lead to a deeper analysis of the factors explaining the fact that some individuals accept EM but have not purchased an EV yet. Secondly, we perform a comparative analysis of the domestic and business sectors. As far as we know, this is the first attempt to perform this comparison.

The structure of the article is as follows. After this introduction, Section 2 presents some important literature on the topic divided by the most common factors affecting EM. Section 3 provides details on the Portuguese case. Section 4 describes the methodology and data. Section 5 describes the results of the research. Finally, Section 6 concludes the paper with a discussion of the main results of the research, indicating their limitations and practical applications, and of the future directions of research in this field.

2. Literature Review

The development of EM can be influenced by several factors, from political and financial incentives to consumer perceptions of it. Several studies have already been devoted to the analysis of these factors. For Long and Axsen [24], anticipating demand for emerging or non-existent technologies will be one of the main challenges associated with the development of EM, as it is difficult to predict future dynamics in consumer preferences, given their instability and uncertainty. In a study for the Nordic region, Sovacool et al. [21] stated that the adoption of EVs is, in a way, similar to that of quitting smoking or the regular practice of physical exercise, as it requires the breaking and substitution of behavioral patterns, so the choice of EVs by consumers is not just a preference of the consumer but also requires adaptation to factors such as low autonomy and limited availability of charging stations. The first group of factors that may have an influence on the acceptance of EM is composed of sociodemographic ones. These also include factors directly related to the perceptions, motivations, and intentions of individuals. For example, Biresselioglu et al. [13] indicated that demographic, personal, and lifestyle factors seemed to be essential in the preference for EVs. Neves et al. [25] showed that in Europe higher levels of employability and education contributed to a higher share of EVs. Using a broader approach, Novotny et al. [16] showed the importance of cultural differences for the acceptance of EV using a sample of 21 countries. For Portugal, de Jesus et al. [26] used lessons from the adoption of vehicles powered by gas, essentially propane and butane, or natural gas, to anticipate the evolution of the EV market. The authors highlighted environmental concerns and education levels as important factors to increase the intention to purchase EVs. Hensher et al. [27] found that urbanism, ecological awareness, technophilia, and experience in car sharing are generally factors that increase the acceptance of EVs. Greater acceptance of these vehicles could be achieved by targeting individuals who move frequently and live in urban areas, especially by emphasizing the experience of sharing EVs rather than conventional vehicles. Tu and Chun [28] highlighted the importance of the energy conservation and environmental protection of vehicles for the preference of Chinese consumers regarding EVs. Using an online questionnaire with a sample of Canadian individuals over the age of 19, Long and Axsen [24] found that, generally, the use of new mobility is associated with younger ages, higher education levels, higher incomes, and males. Other factors have also shown relevance, such as travel patterns, environmental awareness, and technology-oriented lifestyles. Qian and Yin [29] identified other important factors for Chinese consumers, such as a perceived increase in self-esteem rising from the alignment between electric mobility and personal values and beliefs. Hence, public initiatives to promote EVs should take into consideration the cultural values of individuals. Policy makers can use education, media, and advertising to improve the ME strand, even facilitating communication between community members with similar principles aimed at the consumption of environmentally friendly products. Aligned with the previous studies, we investigated the following hypotheses:

Hypothesis 1 (H1).

Sociodemographic factors affect the probability of individuals owning an EV.

Hypothesis 2 (H2).

Sociodemographic factors affect the acceptability of EM.

Another important set of determinants of the growth in EM is the environmental, economic, and fiscal incentives. Leurent and Windisch [15] identified factors that contribute positively to the economic accessibility of this technology, such as tax incentives for the acquisition of EVs and lower use costs. The authors indicated that these types of incentives apply in the transition phase, until the technology achieves economies of scale. In this context, the role played by governments has also been identified as a key factor in the development of EM in several countries. In China, Wang et al. [30] showed that subsidies granted by central and local governments, and large non-monetary incentives such as exemption from vehicle stamp duty, represented crucial factors. However, continued sales growth was threatened by China’s persistent regional protectionism, the unsustainability of large subsidies, as well as several cases of fraud by car manufacturers. Geronikolos and Potoglou [31] stated that in Greece the governmental financial incentives were an important first step to promote EVs. The authors also indicated that the incentives applied should consider the different socioeconomic situations, avoiding inequalities. Lorentzen [32] studied the case of Norway, one of the most advanced markets for EVs in the world. The author emphasized the decisive intervention of the Norwegian government, through successive incentive policies, starting in 1990 with the exclusion of taxes on the purchase of EVs. According to the author, due to these incentives, EVs have become able to compete in price in the market and can reach even better values than conventional equivalent vehicles in terms of performance and/or capacity. Zhang et al. [33] found out, through a questionnaire, that environmental benefits perceived by the population are important to sustain the adoption of EVs in the post-subsidizing period. Nevertheless, in order to enable the mass adoption of these vehicles, it will be essential to increase economic benefits through advances in technology. For South Korea, Kim et al. [34] implemented a questionnaire on the perceived value and purchase intention of EVs, concluding that respondents considered the economic benefit associated with savings in the operating cost of vehicles to be important. Environmental benefits and conduction pleasure were also factors with a positive contribution to the adoption of EVs. De Jesus et al. [26] identified low gas emissions and green energy as critical factors for the adoption of EVs for Portugal. Among the economic benefits, an additional point is raised by Wroblewski and Lewicki [19] on the residual value of vehicles in a study in Poland. If this residual value is higher for EVs, this can be seen as an advantage of these types of vehicles. Other incentives, such as free or privileged parking spaces and priority circulation with access to roads for buses and taxis, have also been pointed out in the literature, for example, by Leurent and Windisch [15] and Lorentzen [32] in Norway. Hence, we tested the following hypotheses:

Hypothesis 3 (H3).

Some incentives (such as environmental, economic, fiscal, and parking and circulation advantages) increase the probability of purchasing an EV.

Hypothesis 4 (H4).

Some incentives (such as environmental, economic, fiscal, and parking and circulation advantages) increase the acceptance of EM.

On the other hand, some factors may hinder the penetration of EVs in the market. Leurent and Windisch [15] highlighted uncertainties regarding future costs, future public policies, and market development, especially in view of the evolution of oil and electricity prices. Biresselioglu et al. [13] identified barriers to EM, such as lack of familiarity with the eco-friendly product market, lack of reliability, non-competitive price, lack of motivation, low availability, model limitations, and technological uncertainty. Kim et al. [34] and de Jesus et al. [26] also highlighted the difficulties arising from the high cost of buying and replacing batteries. One of the most studied factors among the possible difficulties for EM penetration is the extensive existence of charging infrastructures. For example, in a study in Europe, Neves et al. [25] found that the development of batteries and the number of charging stations available are very significant drivers. This result was confirmed by Zhang et al. [33] who indicated that to reduce risk the installation of more charging stations is necessary. Geronikolos and Potoglou [31] emphasized the need for greater allocation of resources to the public charging infrastructure with national coverage in Greece to promote EV further. Desai et al. [18] also pointed out the relevance of the existence of charging stations and infrastructures, while Tu and Chun [28] indicated that vehicle charging is the biggest concern for consumers, and Kim et al. [34] pointed out that the risk associated with charging the vehicles was a significantly negative factor in the perception of their value. Sendek-Matysiak et al. [35] showed vehicle prices and operation costs as being the most relevant impediments to EV penetration. In particular, the authors showed the importance of charging conditions for the total cost of ownership of the vehicles. EVs charged at home tended to achieve cost parity sooner than the ones charged in public stations. Regarding this topic, we tested the following hypotheses:

Hypothesis 5 (H5).

Some barriers (such as higher prices and costs, uncertainty, and technical restrictions) decrease the probability of purchasing an EV.

Hypothesis 6 (H6).

Some barriers (such as higher prices and costs, uncertainty, and technical restrictions) decrease the acceptance of EM.

Another topic that was put forward in the literature as a relevant factor promoting EVs was the type of energy sources available. Neves et al. [25] and de Jesus et al. [26] defend that the increase in the share of EVs is promoted by increasing renewable energy generation.

Regarding the comparison between the DS and the BS, the literature is scarce. Neves et al. [25] showed that industries demonstrate a great potential for the adoption of EVs. However, Sendek-Matysiak et al. [35] pointed out that the progress of electromobility in the commercial vehicle sector has been slower than in the domestic sector, but, at the same time, these vehicles will allow firms to implement corporate social responsibility. Regarding this topic, we have the following research hypothesis:

Hypothesis 7 (H7).

The DS and the BS have similar responses when it comes to the incentives for and barriers to the ownership of EVs or the acceptability of EM.

3. The Portuguese Case

In Portugal, the first ME Program was approved in 2009, with the aim of making the country one of the pioneers in this area [36]. In this plan, a strategy was defined to create a pilot infrastructure of high-powered public charging stations and promote EVs mainly through financial and tax incentives but also other benefits in circulation and parking [36].

Table 2. Incentives and tax benefits associated with 100% electric vehicles in force in Portugal [37].

Particular	Enterprises
Incentive of EUR 4000 for the acquisition or leasing of electric light passenger vehicle, whose value may not exceed EUR 62,500, including VAT, up to the limit of 1300 vehicles or EUR 5,200,000, through the Environmental Fund.	Incentive of EUR 6000 for the acquisition or leasing of light goods vehicles, up to the limit of 150 vehicles or EUR 900,000, through the Environmental Fund.
Incentive of EUR 6000 for the acquisition or leasing of light goods vehicles, up to the limit of 150 vehicles or EUR 900,000, through the Environmental Fund.	Exemption from Autonomous Taxation (Article 88(3) of the IRC Code).
Exemption from payment of ISV (Vehicle Tax) (point (a) of Article 2(2) of Annex I to the Vehicle Tax Code).	Exemption from payment of ISV (point (a) of Article 2(2) of Annex I to the Vehicle Tax Code).
Exemption from payment of the IUC (Single Movement Tax) (point (e) of Article 5(1) of Annex II to the Vehicle Tax Code). Allocation of an incentive to install EV chargers in condominiums at 80% of the purchase value of a charger, up to a maximum of EUR 800 per post, and 80% of the value of the electrical installation up to a maximum of EUR 1000 per parking place, allowing the installation of up to 10 chargers per condominium, connected to the EM network through the Environmental Fund.	Exemption from payment of IUC (point (e) of Article 5(1) of Annex II to the Vehicle Tax Code). Deduction of all VAT relating to the costs of acquisition, manufacture or import, leasing, and processing into plug-in electric or hybrid vehicles of light passenger or mixed electric or hybrid plug-in vehicles; when considering tourist vehicles, the cost of purchase cannot exceed EUR 62,500 (point (f) of Article 21(2) of the VAT Code, with the value defined by Article 1(4) of Ordinance No. 467/2010, of July 7, as amended by Law No. 82-D/2014 of December 31).
	Deduction of all VAT associated with expenditure on electricity used in electric or hybrid plug-in vehicles (point (h) of Article 21(2) of the VAT Code).
	Deduction of all VAT associated with expenditure on electricity used in electric or hybrid plug-in vehicles (point (h) of Article 21(2) of the VAT Code).
	Depreciation of passenger or mixed vehicles is accepted as expenses in the part corresponding to the cost of acquisition or revaluation value up to the amount of EUR 62,500 (point (e) of Article 34(1) of the IRC Code, with the value defined by Article 1(4) of Ordinance No. 467/2010, 7, amended by Law No. 82-D/2014 of December 31).

Source: Own elaboration, using the information at MOBI.E.

presents the financial incentives in force in Portugal.

In Portugal, the EV share has been increasing continuously and is currently around 20% of vehicle sales according to the UVE (Association of Drivers of Electric Vehicles). Our study focuses on the municipality of Felgueiras, which is located in the northern region of Portugal in the district of Porto. It has a total area of 115.74 km² and a population of 58,065 people (population density of 501.7 inhabitants/km²). Currently, despite some rurality in the region, it is highly specialized in the footwear industry.

4. Materials and Methods

4.1. Data

The data were obtained through a quantitative cross-sectional study, based on a non-probabilistic population sample obtained by convenience, between inhabitants and companies in the municipality of Felgueiras. The following parameters were considered as inclusion criteria: being a volunteer adult, being a driver’s license holder (more than 18 years old), being an inhabitant of the municipality of Felgueiras, and, in the case of the BS, that the company in question be based in the municipality. We had no additional constraints in our sample and there were no output scenarios, since we were dealing with revealed preferences. The questionnaire was implemented online through the Google Forms platform for the domestic sector and face-to-face interviews for the business sector, preserving confidentiality through data protection. The dissemination of the questionnaire was carried out through specific networks to guarantee that respondents lived in Felgueiras. The final sample obtained was 256 individuals (DS) and 56 companies (BS). In the BS, the respondent was the owner or executive director of the firm.

Both questionnaires included two sections. In the first section, questions concerning sociodemographic data were asked. In the case of the DS, the gender, age group, educational qualifications, constitution and economic situation of the household, and habits of using cars on a daily basis of the participants were identified. In the BS, the size of the firm reflected by the number of employees and the constitution of the fleet were inquired. Section 2 focused on the respondents’ opinions regarding the incentives for and barriers to the adoption of EVs, and we also inquired whether respondents owned an EV and whether they considered EM a good option.

4.2. Methodology

To test the research hypotheses, we estimated two models. These two models allowed us to assess which factors influence the actual acquisition of EVs and also which factors contribute to the acceptability of EM. These are two related questions, but they do not necessarily overlap. In the first case, we dealt with revealed preferences. In order to buy an EV, respondents need to have the means to buy a vehicle and also need to meet the monetary conditions to buy one. On the other hand, in the second case, we dealt with stated preferences. The acceptance of EM is a broader concept. Respondents may consider buying an EV in the future but have not done so yet, either because their current vehicle does not need to be replaced yet or because they may not meet the monetary conditions. Thus, we performed 2 binary logit regressions (M1 and M2) using the Stata software [38]:

$$P\left(Y=1|X\right)=G\left({\beta }_0+X\beta \right),$$

(1)

where G is the function taking on values strictly between zero and one, for all real numbers Z; and in the logit model, G is the logistic function:

$$G\left(Z\right)=\frac{\mathrm{e}\mathrm{x}\mathrm{p}\left(Z\right)}{\left[1-\mathrm{e}\mathrm{x}\mathrm{p}\left(Z\right)\right]}=\mathsf{\Lambda }\left(Z\right).$$

(2)

Before performing logit regression, we ensured that all the assumptions of the logit model were verified, according to Wooldridge [38]:

• Dependent variable is categorical;
• Data are independent, which means that there is no relationship between observations;
• Data must not show multicollinearity;
• Linear relationship between any continuous independent variable and the logit transformation of the dependent variable;
• There are no extreme outliers.

All the assumptions were verified, and the Hosmer and Lemeshow tests confirmed that the model selected was appropriate, that is, fits the data well. In any case, in order to safeguard the robustness of the estimation, the models were estimated using the robust matrix for the standard errors.

In the first regression (M1), the dependent variable used was the answer to the question “Do you own an electric car?” and took the value 1 if the answer was positive and 0 otherwise, and in the second regression (M2) the dependent variable used was the answer to the question “Do you consider electric cars a good alternative to conventional vehicles?” and took the value 1 if the answer was positive and 0 otherwise.

In generic terms, and according to the model in question, where $Y_i$ will vary according to the model used, we can express these models as follows:

$${Logit(Y}_i)={\beta }_0+{\sum }_{j=1}^n{{\beta }_{j}X}_{ij}+e_i,$$

(3)

where $Y_i$ represents EV ownership in the case of M1 or EM acceptability in the case of M2 for each individual i in the sample. ${\beta }_0$ is the constant term; ${\beta }_j$ represents the natural logarithm of the odds ratio of the j variable; $X_{ij}$ represents the j explanatory variable of the I individual, and $e_i$ is the error term of the equation.

The explanatory variables can be divided into three groups. In the first group, we included sociodemographic and general characteristics. These corresponded to the domestic sector: gender, age, education level, type of employment, income level, and driving time. For the business sector, we considered the size of the firm and number of vehicles in the fleet. The second group of explanatory variables included, for both sectors, the importance given to incentives for EM, namely, fiscal incentives (such as tax reductions or incentives), economic incentives (such as lower operational costs), environmental benefits (lower emissions), free parking and priority circulation, the use of a new technology/modernity, and charging flexibility (the possibility to charge the vehicle at home or in urban centers). The third group of explanatory variables included, for both sectors, the importance given to potential barriers to EM, namely, non-competitive prices, higher costs (for example, maintenance of batteries), uncertainty, technical restrictions (such as low battery duration or high charging times), and unsafety.

Table 3. Description of the variables used in the models.

Variable	Description	Category
Gen	Gender	0 = Male; 1 = Female
Age	Age	1 = [18–30]; 2 = [31–40]; and 3 = 41 or +
Educ	Education	0 = Undergraduate; 1 = Higher Education and/or Postgraduate
Sit_prof	Professional Situation	0 = Unemployed, Student, or Retired; 1 = Full-time employed, Part-time, or Self-employed
Income	Household Income	1 = [EUR 500–999]; 2 = [EUR 1000–1999]; 3 = [EUR 2000–3999]; and 4 = EUR 4000 or +
Driving	Driving Time	1 = less than 30 min.; 2 = [30 min.–1 h]; 3 = [1 h–2 h]; and 4 = 2 h or +
Incentive factors for the acquisition of EVs
Fiscal	Fiscal Incentives	0 = No; 1 = Yes
Economic	Economic Incentives	0 = No; 1 = Yes
Environmental	Environmental Incentives	0 = No; 1 = Yes
Parking	Free Parking and Priority Circulation	0 = No; 1 = Yes
Tech	New Technology/Modernity	0 = No; 1 = Yes
Charging	Charging at Home and Urban Centers	0 = No; 1 = Yes
Barriers to the acquisition of EVs
Price	Importance given to Price	0 = No; 1 = Yes
Cost	Importance given to Cost, Durability, and Maintenance	0 = No; 1 = Yes
Uncertainty	Importance given to Uncertainty/Lack of Information and Infrastructure	0 = No; 1 = Yes
Technical	Importance given to Technical Restrictions	0 = No; 1 = Yes
Unsafety	Importance given to Unsafety	0 = No; 1 = Yes

describes the variables used in the model.

5. Results

5.1. Descriptive Analysis

In this section, we describe the responses obtained to each question concerning their frequency and number of respondents both for the DS and the BS.

Table 4. Descriptive analysis of respondents to the questionnaire related to the domestic sector.

Variable	Description	Frequency
		n	%
Gender	Male	121	47.6
	Female	133	52.4
Age Group	18–30 years	169	66.5
	31–40 years	55	21.7
	>41 years	30	11.8
Education	Undergraduate	105	41.3
	Higher Education and/or Postgraduate	149	58.7
Professional Situation	Student/Unemployed	36	14.2
	Employee	175	68.9
	Self-employed	43	16.9
Net Monthly Income	EUR 500–999	80	31.5
	EUR 1.000–1.999	88	34.6
	EUR 2.000–3.999	74	29.1
	EUR >4.000	12	4.72
Daily Driving Time	<30 min	99	39.0
	30 min–1 h	96	37.8
	1–2 h	36	14.2
	>2 h	23	9.1
Incentive factors for the acquisition of EVs
Fiscal Incentives	Yes	166	65.4
	No	88	34.6
Economic Benefits	Yes	188	74.0
	No	66	26.0
Environmental Benefits	Yes	187	73.6
	No	67	26.4
Free Parking and Priority Circulation	Yes	64	25.2
	No	190	75.8
New Technology/Modernity	Yes	63	24.8
	No	191	75.2
Charging at Home and Urban Centers	Yes	148	58.3
	No	106	41.7
Barriers to the acquisition of EVs
Non-competitive price	Yes	137	53.9
	No	117	46.1
Cost, Durability, and Maintenance	Yes	161	63.4
	No	93	36.6
Uncertainty/Lack of Information and Infrastructure	Yes	163	64.2
	No	91	35.8
Technical Restrictions	Yes	140	55.1
	No	114	44.9
Lack of Security	Yes	29	11.4
	No	225	88.6

shows the sociodemographic characteristics of the respondents of the domestic sector. The majority of the participants were female (52.6%), were between 21 and 30 years old (64.3%), and had higher education (58.6%). Most participants were employees (68.9%) and had a net monthly income between EUR 1000 and 2000. About 40% had a driving time of less than 30 min per day.

Regarding the BS,

Table 5. Descriptive analysis of respondents to the questionnaire related to the BS.

		Frequency
		n	%
N° of employers	1–49	43	76.8
	>50	13	23.2
Number of Fleet Vehicles	1–5	34	60.7
	>6	22	39.3

shows the characteristics of the sample considered in the study. Most companies belong to the footwear business (42.9%), with a number of employees of less than 50 (76.8%). In terms of the formation of the business fleet, most companies contain between 1 and 5 vehicles (65.4%).

5.2. Binary Logit Regression

This section presents the results of our estimations using a binary logistic model.

Table 6. Binary Logit Model Results for the Domestic Sector.

Included Observations: 254 Coefficient Covariance Computed Using Observed Hessian
M1—Dependent Variable: Have EV				M2—Dependent Variable: Good Option
Variable	Coefficient	Std. Err.	t-Value	Coefficient	Std. Err.	t-Value
Gen	−1.19 **	0.48	−2.47	0.26	0.33	0.81
Age_31_40	−0.04	0.48	−0.09	−0.90 **	0.37	−2.45
Age_41+	−0.83	−1.00	0.32	0.38	0.52	0.74
Educ	0.37	0.51	0.72	0.28	0.34	0.79
Employed	−0.01	0.69	−0.02	−0.15	0.48	−0.31
Income_500_999	−2.56 ***	0.87	−2.95	1.29 *	0.73	1.87
Income_1000_1999	−2.38 ***	0.82	−2.90	1.24 *	0.71	1.83
Income_2000_3999	−1.94 **	0.79	−2.45	1.68 **	0.72	2.42
Driving_30 min_1 h	0.83	0.53	1.58	−0.74 **	0.34	−2.16
Driving_1 h_2 h	−0.30	0.62	−0.47	−0.08	0.48	−0.17
Driving_2+	0.99	0.75	1.31	0.18	0.58	0.28
Incentive factors for the acquisition of EVs
Fiscal	0.48	0.47	1.01	0.47	0.32	1.48
Economic	−0.91	0.48	−1.89	−0.14	0.36	−0.40
Environmental	0.50	0.50	1.00	0.08	0.35	0.20
Parking	−0.32	0.67	−0.48	1.23 ***	0.42	2.93
Tech	−0.86	0.63	−1.36	−0.16	0.36	−0.44
Charging	0.49	0.46	1.08	−0.51	0.31	−1.65
Barriers to the acquisition of EVs
Price	0.01	0.50	0.02	−0.14	0.32	−0.47
Cost	−1.00 **	0.47	−2.15	−0.28	0.32	−0.85
Uncertainty	0.81	0.51	1.58	0.07	0.33	0.21
Technical	0.38	0.46	0.83	0.24	0.32	0.74
Unsafety	−1.78 *	1.05	−1.70	−2.09 ***	0.44	−4.79
Constant	−0.13	1.39	−0.10	−0.44	0.98	−0.45
Log-likelihood	−74.396			−141.31
Pseudo-R2	0.2105			0.1634

Notes: Std. Err.: standard error; ***, **, and * represent significance at 1%, 5%, and 10%.

shows the results for the DS.

It is important to note that for the age variable the reference group is with ages from 18 to 30; hence, the variables presented in the table represent a comparison with this reference group. Regarding income, the reference group is the highest income (higher than EUR 4000). For driving time, the reference group is the one driving less than 30 min.

Considering the first model, where the dependent variable is owning or not owning an EV, the results obtained demonstrate that there is no statistically significant difference between the different categories of the variables of age group, educational qualifications, professional status, and driving time. On the other hand, women have a lower probability of owning an EV, which is in accordance with the literature [24]. Regarding net monthly income, it was found that, compared to individuals with incomes greater than EUR 4000, individuals with incomes of EUR 500–999, EUR 1000–1.900, and EUR 2.000–3.900 have a statistically significant lower probability of owning an EV. Furthermore, the lower the income, the lower that probability. These results confirm Hypothesis 1, that is, there are sociodemographic factors that affect the probability of owning an EV. In particular, men are more likely to own an EV, probably due to a higher desire to own the latest technologies. Individuals with higher income levels are also more likely to own an EV, which relates to the higher purchase cost of these vehicles.

Regarding the incentives for the acquisition of an EV, our results demonstrate that the variables considered do not have statistical significance. Hence, Hypothesis 3 is rejected. Despite being slightly surprising, this result is probably explained by the fact that consumers who already own an EV are those who desired one and had the economic means for its purchase, regardless of any incentives. Hence, in the current moment, actual purchases are more connected to the personal preferences of individuals. Regarding the barriers to the acquisition of EVs, the results show that only cost and security concerns are statistically significant. Hence, individuals that consider higher costs and security issues as relevant barriers have a lower probability of owning an EV. This result implied that Hypothesis 5 is accepted even though only two factors are relevant. Concerns about security are likely related to a lack of proper information dissemination. EVs are not less safe than conventional ones, but there may have been a certain image created in the media indicating that they are.

The results obtained in M2 demonstrate that there is no statistically significant difference between the different categories of the variables of gender, educational qualifications, and professional status for the acceptability of EM. On the other hand, older individuals (with ages between 31 and 40 years) are less likely to consider EM a good option when compared to younger ones (with ages between 18 and 30 years). This shows that acceptance of EM is higher for younger generations, which probably relates to behavioral aspects such as resistance to change in older generations and technological acceptance in younger ones. Regarding net monthly income, it was found that, compared to individuals with incomes greater than EUR 4000, individuals in lower income groups have a statistically significant higher probability of finding EM to be a good option. Therefore, Hypothesis 2 is accepted. This result raises an interesting observation. Contrarily to the effective purchase of an EV, the desire to own one is more connected to lower income groups. This means that richer individuals probably already own an EV if they want to. However, for lower income groups, the desire may exist but the economic availability for the purchase does not exist. Hence, promoting broader EV fiscal and economic incentives from the government is key. Some lower income individuals would buy an EV if they could afford it. The difference in results when compared to M1 may be explained by the high initial investment required to buy an EV. Hence, lower income individuals find EM to be a good option but do not effectively own an EV. It is worth keeping in mind that in M2 we are dealing with stated preferences regarding the acceptance of EM. Therefore, individuals may like the idea of having an EV but may not be able to afford one. In the case of daily driving time, individuals who drive between 30 min and 1 h show a lower acceptance of EM than those who drive less than 30 min. The perceived duration of the battery can explain such a result. Despite the fact that most EVs are currently offering sufficient driving ranges for longer distances, it is important to consider that the general population does not have proper and reliable information. This is particularly relevant because in the case of the acceptance of EM we are dealing with revealed preferences, that is, individuals do not own an EV and therefore do not know for sure their technical characteristics.

The results regarding the factors that encourage the purchase of EVs demonstrated that only the variable of free parking and priority circulation positively influence the acceptance of EM. The other variables are not statistically significant. Still, this means that Hypothesis 4 is accepted. The lack of explanatory power of the existing incentives for the acceptance of EM probably means that these incentives are not perceived as sufficient. This is in line with the idea previously presented that individuals who already own an EV are those that could afford one, but some individuals would like to own one but cannot afford it.

Regarding the barriers to the acquisition of EVs, the results showed once again that most variables are not statistically significant. Only security concerns led individuals to have a lower acceptance of EM. Still, this result validates Hypothesis 6. As before, it also shows the importance of providing reliable information to consumers, since EVs are not less safe than conventional ones, but there seems to be a general perception that this is the case. De Jesus et al. [25] had already called attention to the relevance of reliable information in informing potential buyers, influencing the decision-making process. It is interesting to note that concerns regarding the charging stations were not found to be relevant for our sample. This is contrary to the most common literature (e.g., [25,31,33]).

Given that most independent variables included in the models were not statistically significant, we re-estimated the models considering only statistically significant independent variables. The results are presented in

Table 7. Binary logit model results using only statistically significant independent variables for the Domestic Sector.

Included Observations: 254 Coefficient Covariance Computed Using Observed Hessian
M1—Dependent Variable: Have EV				M2—Dependent Variable: Good Option
Variable	Coefficient	Std. Err.	t-Value	Coefficient	Std. Err.	t-Value
Gen	−0.71 *	0.39	−1.81	-	-	-
Age_31_40	-	-	-	−0.98 ***	0.34	−2.90
Income_500_999	−1.89 **	0.78	−2.43	1.00	0.69	1.46
Income_1000_1999	−1.73 **	0.76	−2.27	1.05	0.68	1.46
Income_2000_3999	−1.25 **	0.72	−1.73	1.46 **	0.70	2.08
Driving_30 min_1 h	-	-	-	−0.71 **	0.29	−2.42
Incentive factors for the acquisition of EVs
Parking	-	-	-	1.01 ***	0.39	2.60
Barriers to the acquisition of EVs
Cost	−0.91 **	0.40	−2.27	-	-	-
Unsafety	−1.46 *	1.00	−1.47	−2.09 ***	0.44	−4.79
Constant	−0.44	0.71	0.62	−0.11	0.65	−0.17
Log-likelihood	−83.94			−145.74
Pseudo-R2	0.1092			0.1372

Notes: Std. Err.: standard error; ***, **, and * represent significance at 1%, 5%, and 10%.

.

Analyzing the results in Table 7, it is possible to see that most variables continued to be statistically significant, except in Model 2 where two of the variables related to family income became statistically insignificant at 10%. This shows that a more complete model, with more variables, even if not statistically significant has a higher explanatory power. This is also visible by the decrease in the value of Pseudo-R².

Similarly, two binary logistic regressions were estimated for the BS, considering the same dependent variables as in the DS (in M1 and M2), and the Hosmer and Lemeshow tests also confirmed that the model selected was appropriate.

As before, in the first regression (M3) the dependent variable used was the answer to the question “Do you own an electric car?” and in the second regression (M4) the dependent variable used was the answer to the question “Do you consider electric cars a good alternative to conventional vehicles?”

Table 8. Binary Logit Model Results for the Business Sector.

Included Observations: 56 Coefficient Covariance Computed Using Observed Hessian
M3—Dependent Variable: Have EV				M4—Dependent Variable: Good Option
Variable	Coefficient	Std. Err.	t-Value	Coefficient	Std. Err.	t-Value
Workers_49	−0.28	1.27	−0.22	−2.21	1.82	−1.22
Cars_5	−0.66	1.22	−0.54	1.84	1.59	1.16
Incentive factors for the acquisition of EVs
Fiscal_Incentives	1.17	0.77	1.51	0.20	0.95	0.21
Economic_Incentives	1.45	0.92	1.58	−1.53	0.93	−1.64
Environmental_Incentives	0.89	1.02	0.88	2.10 **	0.93	2.25
Parking_Circulation	−0.99	0.99	−1.00	2.62 **	1.13	2.33
Technology	−1.65 *	0.92	−1.79	−1.09	1.05	−1.04
Barriers to the acquisition of EVs
Price	−0.46	0.85	−0.54	0.51	0.89	0.58
Cost	0.16	0.78	0.21	0.90	0.76	1.19
Uncertainty	−0.66	0.74	−0.88	−0.53	0.90	−0.58
Technical_Restrition	−0.24	0.78	−0.31	−1.63 *	0.83	−1.96
Unsafety				−1.01	1.57	−0.64
Constant	−1.80	1.35	−1.33	0.65	1.66	0.39
Log-likelihood	−27.54			−27.84
Pseudo-R2	0.1514			0.2580

Notes: Std. Err.: standard error; ** and * represent significance at 5%, and 10%.

shows the results of the models for the BS.

The results demonstrated that, in both models, the number of employees and the number of vehicles in the fleet are not statistically significant.

In the first model, the results regarding incentives for the purchase of na EV demonstrate that only the variable of new technology and driving pleasure is statistically significant. However, the negative sign associated with this variable is a surprising result and is not easily explained. Simultaneously, none of the barriers to the purchase of EVs is statistically significant.

In the second model, only environmental benefits and the variable of free parking and priority circulation positively influence the acceptance of EM. Regarding the barriers to the purchase of EVs, we find two of them to be statistically significant. Companies that considered technical restrictions and unsafety important factors showed a lower acceptance of EM in comparison to those who did not consider these variables important. Overall, our results lead to the rejection of Hypothesis 7. Hence, the DS and BS present different behaviors regarding the purchase of EVs and the acceptance of EM. This result indicates the need to design and implement differentiated policies for each of these sectors.

To maintain the consistency of the analysis, and since it was also verified that most of the independent variables included in the model were not statistically significant, we also re-estimated the models for the BS, including only statistically significant independent variables. The results are presented in

Table 9. Binary logit model results using only statistically significant independent variables for the Business Sector.

Included Observations: 56 Coefficient Covariance Computed Using Observed Hessian
M3-Dependent Variable: Have EV				M4-Dependent Variable: Good Option
Variable	Coefficient	Std. Err.	t-Value	Coefficient	Std. Err.	t-Value
Incentive factors for the acquisition of EVs
Environmental_Incentives	-	-	-	1.40 *	0.77	1.81
Parking_Circulation	-	-	-	1.64	1.16	1.42
Technology	−0.71	0.85	−0.83	-	-	-
Barriers to the acquisition of EVs
Technical_Restrition	-	-	-	−1.43 *	0.74	−1.92
Constant	−0.79	0.32	−2.45	0.17	0.61	0.29
Log-likelihood	−33.22			−33.34
Pseudo-R2	0.0226			0.1115

Notes: Std. Err.: standard error; * represent significance at 10%.

.

As in the DS, here we can also see the necessity of considering other variables in the regression to obtain more comprehensive conclusions.

6. Discussion and Conclusions

Climate change has been leading the planet towards a potentially irreversible situation [39]. We are currently seeing an increase in the occurrence of phenomena such as heat waves and floods and increased mortality, which even the most pessimistic climatologists expected would only occur within two decades. Undoubtedly, EVs represent a better alternative in environmental terms than conventional vehicles. EV sales continue to break records, and some manufacturers plan to electrify their fleets even before the targets set in the applicable legislation [40]. However, the route to the electrification of the transport sector appears long and we are only at the beginning. Among other factors, consumers’ preferences will be essential for this process [41].

Hence, understanding consumers’ perceptions of EVs and acceptance of EM is essential. In this article, we studied the Portuguese case, focusing on the comparison between the domestic and business sectors. We explored research hypotheses regarding the influence of sociodemographic factors, incentives, and barriers on the purchase of an EV and on the acceptance of EM. We estimated two models where the first explored the determinants of the effective purchase of EVs while the second focused on the determinants of the acceptance of EM.

Our results showed that that sociodemographic characteristics were the main drivers for EV purchases. Males were more likely to own an EV, which is in accordance with the literature [24]. Additionally, consumers with higher income were also more likely to own an EV. In the domestic sector, use and maintenance costs were found to be significant in reducing the likelihood of owning an EV, as well as concerns with security. This concern could be explained by the lack of consumer information on the safety of the electrical circuit, as lithium-ion batteries have already been shown to have a much lower risk of explosion than conventional petrol vehicles. Our results highlight the need to reduce EV costs in order to accelerate their market penetration. Simultaneously, it is key to provide consumers with comprehensive and reliable information to decrease uncertainty. The current lack of information may be responsible for the fact that most factors in the model were not statistically significant. For instance, individuals may not be aware of the fiscal incentives available for the purchase of EVs [13,33]. This observation is a relatively new result in the literature. Additionally, most people do not yet own an EV, which may relate to the lack of a need to replace their current vehicle, lack of monetary power to purchase an EV, lack of information, or the natural tendency of humans to resist change.

In the business sector, only technology-related topics seemed to be relevant. Costs and security are important for domestic buyers but not for business ones. Hence, the BS and DS display distinct behaviors. This result implies that policies directed at each sector need to be differentiated.

Regarding the determinants of the acceptability of EM, for domestic buyers we observed that younger generations (with ages between 18 and 30) are more likely to consider EV a good option than older consumers. This factor highlights the importance of time in the energy transition process. It is going to take some years for consumers to get used to the new technologies and surpass the habitual resistance to change and new technologies. One interesting observation was that consumers with lower income levels seem to find EVs appealing, even though they do not effectively own one. This indicates the importance of the price barrier, although price was not found to be statistically significant in the explanation of the acceptance of EM. This result can only indicate the difference between revealed and stated preferences. Consumers do not think or want to admit that price is a barrier to the purchase of an EV. However, they indicate that EVs are a good option but do not effectively buy one or at least have not yet bought one. This result also highlights the importance of providing broader fiscal and economic incentives if a proper transition in the transport sector is desired. This would allow consumers with lower income levels, who have good acceptance of EM, to effectively buy an EV. Furthermore, consumers with driving times between 30 min and 1 h are less likely to consider EVs a good option than consumers who drive less than 30 min. This can be explained by the concerns related to the battery autonomy [42]. Facilities regarding free parking spaces also seemed to increase the acceptability of EM, while concerns regarding security decreased that acceptability in the domestic sector. This result shows the importance of providing reliable information to consumers, since EVs are not less safe than conventional ones, but there seems to be a general perception that this is the case.

When it comes to the business sector, environmental incentives appeared as a statistically significant variable that increased EM acceptability. Here, we note that respondents were answering the questionnaire in person and on behalf of the firm they were representing. Hence, environmental concerns may be important for the image of the company. So, this variable may not only represent a real concern with the environment, which still may exist, but also a form of “warm glowing”. Additionally, as in the domestic sector, parking facilities increase the acceptability of EM. On the other hand, technical restrictions decrease this acceptability. These restrictions may include, for instance, low battery duration, which is an understandable concern when we are talking about commercial, likely long-distance, transport. Once again, the comparison between the domestic and business sectors is very limited since only parking facilities appear relevant for both sectors. Most explanatory variables are not statistically significant in the two models.

One surprising result of our models is the lack of statistical significance of fiscal incentives. These have been identified as important in the literature [30,31,32]. This result can be explained by a general lack of information regarding the existing incentives or by the fact that EVs are already purchased by richer consumers. Another possible explanation is that individuals who already own an EV are the ones that had the desire and economic means to do so, regardless of any incentives. Simultaneously, for lower income consumers, the incentives may not be enough for them to buy an EV. In fact, obtaining these incentives is not immediate for consumers and depends on meeting eligibility criteria [43], reinforcing the fact that an individual wishing to acquire an EV must have the financial power to do so. Still, it is surprising that even for the acceptance of EM these incentives do not have explanatory power. This result can be explained by the lack of available information or by the differences found in stated preferences when compared to revealed ones. Another surprising result was the lack of explanatory power of the variables related to charging the vehicles. This has been an important factor in the literature [13]. This result is probably explained by the low number of respondents with EVs and therefore a low level of information regarding this topic. Another aspect worth mentioning is the lack of relevance of environmental concerns for the domestic sector, both for the effective purchase of the EV and for the acceptability of EM. Once again, this aspect reveals a strong need to increase information and education among the general public.

Overall, our results show a relatively small penetration of EV but higher acceptability of EM. There seems to be a need for greater economic support for consumers with lower income levels that do not own an EV but consider it a good option. Hence, improved tax and economic incentives would be required. Additionally, a larger amount of information seems to be advisable since some consumers are resistant to change due to concerns, such as security, that are not justified. The environmental benefits of EVs also need to be emphasized. Finally, younger generations will probably adopt EM more easily; hence, in a few years, EV penetration will be much more visible.

It is important to note the limitations of this study, namely, the short sampling of respondents related to the difficulties in disseminating the questionnaires, especially in the business sector. This low representativity requires caution in extrapolating the results. Hence, our results are dependent on the sample and the methodology used, which is also a limitation of our work. Another important limitation of our study is that it is based on stated preferences, which can be biased. In future research, it would be interesting to test other methodologies and broaden our sample to test the results obtained.