The Differences that Electric Vehicle Diffusions make:
An Investigation of Observers’ Adoption Attitude and
Status Inferences.
by
Patrick Hill
The Differences that Electric Vehicle Diffusions make:
An Investigation of Observers’ Adoption Attitude and Status
Inferences
Master Thesis
by
Patrick Hill
26
thJune 2014
Submitted to the:
Faculty of Economics and Business, Marketing
University of Groningen
Supervisor: dr. Jan Willem Bolderdijk
Data of the Author: Student number: S2555123
TABLE OF CONTENTS
Management Summary ... II List of Figures ... III List of Abbreviations ... IV
1 Introduction ... 1
2 Theoretical Foundations ... 3
2.1 Factors Influencing the Adoption of Electric Vehicles ... 3
2.2 Diffusion Theory and Related Observability of EVs ... 4
2.3 Inferences of Status ... 6
2.4 Individuals’ Need for Uniqueness & Corresponding Behaviour ... 8
2.4.1 Influence on Adoption of Electric Vehicles ... 8
2.4.2 Influence on Status Inferences ... 9
2.5 Overview of the Present Research ... 10
3 Research Design & Method ... 11
3.1 Participants & Data Collection ... 11
3.2 Design and Procedure ... 11
3.2.1 Manipulation of Perceived EV Diffusion ... 12
3.2.2 Measurement of EV Adoption Attitude ... 13
3.2.3 Measurement of Inferred Status ... 15
3.2.4 Measurement of Need for Uniqueness ... 16
4 Results ... 16
4.1 Results for Adoption Attitude ... 17
4.2 Results for Inferred Status ... 19
4.3 The potential Relationship between Adoption Attitude and Inferred Status ... 20
5 Discussion ... 21
6 Limitations and Directions for Future Research ... 24
7 Managerial Implications ... 26
References ... 28
Appendices ... 35
Management Summary
List of Figures
Figure 1: Conceptual framework and hypotheses ... 10
Figure 2: Depiction of the “scare EVs” condition in the survey ... 12
Figure 3: Depiction of the “common EVs” condition in the survey ... 13
Figure 4: Depiction of the questions regarding adoption and interest ... 14
Figure 5: Inferred Status between EV Diffusions for Levels of Need for Uniqueness ... 18
List of Abbreviations
BEV ... Battery Electric Vehicle (for a detailed explanation compare appendix 1) DUCP ... Desire for Unique Consumer Products (scale developed by Lynn & Harris,
1997)
EV ... Electric Vehicle (for a detailed explanation compare appendix 1)
HEV ... Hybrid Electric Vehicle (for a detailed explanation compare appendix 1) ICE ... Internal Combustion Engine
M ... Mean (statistical)
NUCP ... Need for Unique Consumer Products (scale developed by Tian et al., 2001) PEV ... Pure Electric Vehicle (for a detailed explanation compare appendix 1) PHEV ... Plug-‐in Hybrid Electric Vehicle (for a detailed explanation compare
appendix 1)
REEV ... Range-‐extended Electric Vehicle (for a detailed explanation compare appendix 1)
1
Introduction
While electrified cars still seem to be futuristic they have already been on the streets more than a century ago. After all, Mismanagement, bad practicability, and strong competition let the electric vehicle (EV) finally fail and internal combustion engine (ICE) cars took over by the 1930s and are still predominating in the 21st century (International Energy Agency, 2013; Rae, 1955). Propelled by climate change, an increase in global pollution, and a desire for independence from limited fossil fuels, consideration for the formerly inglorious EV has risen again in the last years. While the EV could be the ultimate long-term solution, this change will not happen quickly considering current estimates of 10 % market share of EVs for the year 2020 (McKinsey&Company, 2013). The actually slow development of EVs is underlined by a recent press release by visionary Tesla Motors CEO Elon Musk who stated that EVs are still too underrepresented in markets and that he could have not been more wrong in fearing that the large car companies would squeeze Tesla when entering markets1 (Musk, 2014). Nevertheless, when being faced with uplifts in EV sales and aligned different local market developments, future management and marketing is challenged by an inevitable puzzle: How will the diffusion of EVs affect people’s adoption attitudes of these cars? And what inferences do people make when coming across EVs in their surrounding?
As highly innovative products, EVs are subjects of the spreading of their idea and potential advantages over ICE cars (e.g., Bass, 1969 & 1980; Eggers & Eggers, 2011; Rogers, 1962, 1976, & 2010). Thus, diffusion, embedded in a host of other influencing factors, is inevitably driving the success of EVs. Yet, despite a host of sources emphasising the importance of market share of EVs on their (1) adoption potential there is almost no research taking into account effects of different diffusion phases. Not only can diffusion be expected to change within one market over time but also are there currently immense differences in EV diffusion between different markets. Compare, for example, California, which can be seen as the benchmark market for electric vehicle mobility, with Germany. Although being one of the big automobile countries of the world, Germany still has a relatively low share of EVs compared with the US west coast state (International Energy Agency, 2013). This research tries to close this gap and proposes that EVs are more likely to be adopted when potential car
1 Musk feared that huge market power of large car companies would rule out Tesla’s competitive
customers observe (more) others using an innovative EV. A second limitation that is addressed by this research is derived from findings that detect identity-signalling motives as important attributes of the adoption of EVs (Gärling & Thøgersen, 2001; Heffner, Kenneth & Turrentine, 2005 & 2007; Noppers, Keizer, Bolderdijk & Steg, 2014). As status is often suggested to be an underlying motive in existing literature, this research expands findings by questioning what (2) status observers actually confer to an EV owner depending on the previously mentioned different diffusion stages. Such knowledge can be insightful because it shows whether desired status is actually perceived like it was intended to. Further, this can give an idea of the impression people have of EV owners. In addition, this research measures the role of individuals’ uniqueness need on both (1) adoption and (2) status inferences because the importance of uniqueness is often recalled with regards to new and innovative products (Gärling & Thøgersen, 2001; Rodgers, 1962 & 2010, Vandecasteele & Geuens, 2010). Because EVs are still quite new phenomena this research aims at providing deeper understandings of underlying effects that drive individual perception and adoption of EVs. By doing so, findings of this paper should be helpful in understanding effects of different diffusion of EVs in practice and theory. Future decision makers and theorists are hereby hoped to profit from the results that might also function as an initiation for further research. This paper is organised as following. First, the theoretical foundations for this research are identified. This section specifically identifies (1) factors that have been found to influence the adoption and diffusion potential of EVs, (2) the inferences of status, and (3) the role of individuals’ desire for uniqueness on both concepts (1) and (2). Moreover, each of these parts is accompanied by the proposition of the evolved hypotheses, which are summarised in a graphical conceptual model. Hereafter, the conceived research design is described and consequent results presented and discussed. Next, an identification of limitations is accompanied by an array of future research opportunities. Finally this paper is closed with some important implications for future decision makers. Note that, for simplification reasons, this paper uses “EV” as a placeholder for all electric vehicle forms. Important for the classification applied here is the fact that a car can be clearly identified as having an electric engine (either pure or in addition), which makes it recognizable different from a conventionally powered car (PricewaterhouseCoopers, 2011) 2.
2 More specifically EVs identified yet incorporate forms like “All Electric Vehicles” (EV/BEV/PEV),
2
Theoretical Foundations
2.1 Factors Influencing the Adoption of Electric Vehicles
EVs have increasingly become part of literature in areas such as sustainable consumption3,
transportation, energy, and marketing. Starting early in the 1990s, partly in response to California’s zero-emission vehicle mandate4, a host of research started to deal with EVs and adoption forecasts for California (e.g., Becker, Sidhu & Tenderich, 2009; B. Brownstone, Bunch, Golob & Ren, 1996; Brownstone, Bunch & Train, 2000; Hidrue, Parsons, Kempton & Gardner, 2010; Segal, 1995) and regions outside the US market (e.g., Dagsvik, Wennemo, Wetterwald & Aaberge, 2002; Eggers & Eggers, 2011; Eppstein, Grover, Marshall & Rizzo, 2011). Most of the existing studies were conducted either to forecast EV demand or to investigate what factors fuel and hinder EV adoption potential. Here, one frequently finds three attributes, namely instrumental, environmental and symbolic5, as important factors for the adoption and usage of sustainable products. These attributes are applicable for cars (and thus EVs) as well (Steg, Vlek & Slotegraaf, 2001). Due to their key function for understanding EV adoption these attributes are shortly summarised hereafter.
Instrumental attributes of EVs include consumers’ evaluation of factors, such as mileage,
road performance and comfort. Especially the (up to now) limited range of batteries and the corresponding restriction in perceived travelling distance is often considered as a restraint (Eppstein et. al, 2011; Franke, Bühler, Cocron, Neumann & Krems, 2012). Another important instrumental driver is the price tag of EVs, which often inheres a significant premium compared to conventional car options (e.g., Franke et. al, 2012; Heffner et. al, 2007; Kotler, 2011). However, this price premium can be counteracted by governmental tax subsidies (Eppstein et. al, 2011) and adjustments in EV price structure (Thøgersen, 2005) – a strategy that can affectively stimulate initial EV pushes into markets6. These instrumental attributes
3 Consumption that meets the patterns of present generations without harming future generations’
ability to meet their own needs (World commission on Environment and Development, 1987)
4 Enacted in 1990 the regulation requires car companies to sell a constantly increasing percentage of
their cars being on sale in California as zero-emission vehicles. In 2001 this regulation was adapted by a category system that also incorporates other forms such as HEVs (California Environmental Protection Agency, 2014)
5 Every author uses his own terms here. The terms used here are adapted from Noppers et al. (2014). 6 Skerlos & Winebrake (2010) found that tax credits are more effect full when being adapted to
mostly have rather negative connotations while only some research identifies instrumental benefits of EVs, such as increased comfort due to nearly silent work mode (e.g., Franke et. al, 2012) or independence from gas stations and oil producers (Heffner et. al, 2007).
Environmental attributes can also fuel EV adoption. Consumers’ intentions to preserve the
environment when opting for a (more) sustainable car have been highlighted, for example, in Heffner et al. (2007), Thøgersen (2005) and Gärling & Thøgersen (2001). Noppers et al. (2014) found that environmental attributes are important for an increase in sustainable products consumption (EVs being one part herein) or even more important than instrumental attributes, when they asked respondents to rate instrumental attributes in an indirect manner.
Symbolic attributes pose the third influencing factor that can be predominantly found in
psychological, consumer, and marketing research. Cars are inevitably conspicuously consumed/used products and thus perfectly suited to function as a symbol (Berger & Heath, 2007; Hsee, 1999; Tian, Bearden, & Hunter, 2001; Veblen, 1899/1994). Further research suggests variety seeking and individuality motives as part of this symbolism (Ariely & Levav, 2000; Heffner et. al, 2007). In a California based interview with early HEV owners by Heffner et. al (2005) some interviewees also stressed their pioneering role and technology affinity. After all, a host of psychological research within the EV realm is dedicated to the role of status motives. An EV can create identity for the owner, making a statement (Heffner et. al, 2007) or showing altruism (e.g., Kollmuss 2002). Griskevicius et. al (2010) found that green vehicle choice could function as a costly signal, which is again associated with status.
2.2 Diffusion Theory and Related Observability of EVs
products7. Similar to Rogers’ findings, Bass (1969 & 1980) identifies a certain critical mass that is reached when about 10 – 25% of the market start using an innovation. To date, EVs can be placed at the beginning of the curves’ ascending slope characterised by relatively low awareness of products, systems, and technologies (Becker & Sidhu, 2009; Eggers & Eggers, 2011; Eppstein et al., 2011; Thøgersen, 2005). Yet, EV adoption rates are currently rather different depending on geographic markets (compare vehicle fleet shares of EVs between discrete worldwide regions, (International Energy Agency, 2013)).
As a result, the extent to which EVs are prominent in consumers’ habitat is crucial. Observability of EVs will fuel consumers’ initial interest and curiosity, which then influences their adoption of EVs (Fisher & Price, 1992; Gärling & Thøgersen, 2001). Consumers will use their observations as informational input (Hoyer et. al, 2013) and evaluate EVs on at least some of the attributes described beforehand. The reasoning of the classic diffusion theory together with discussed constraints will pose a warning signal when EVs are rather scarce on the roads. Golder & Tellis’ (2004; for an overview: 2010) stated that “because most individuals do not purchase a new product during the introduction stage, most consumers use this information to decide likewise” and vice versa. Doubts about the advantages of the product and that it might become an industry standard in the future pose further uncertainties to observers (Ziamou & Ratneshwar, 2002; Hoyer et. al, 2013). Eggers & Eggers (2011) concluded that the achievement of a critical mass is even more important, when innovations depend on the availability of corresponding services or products, which is certainly the case for EVs and attendant charging stations. Once EVs are not that scarce anymore this information can imply falling prices, which again matches a greater number of people’s “budgetary limitations” (Eggers & Eggers, 2011). Further, an increasing number of users and an increased service level can fuel positive network externalities and can be interpreted as a signal of higher relative quality (Hellofs & Jacobson, 1999; Katz & Shapiro, 1985). After all, the simple fact that others are doing so can function as a positive sign towards adoption as found by Nolan, Schultz, Cialdini, Goldstein & Griskevicius (2008) for a case of engagement in environmental protection.
7 The very first users of an innovation, the “innovators”, are usually the first 2.5% of the market,
Hence, observers will use information, regardless of which kind, to arrive at a certain adoption attitude. The higher the diffusion of EVs and availableness of information obtained from that, the smaller people’s EV related constraints should become, which consequently makes a larger number of people to opt for an EV. Based on the high visibility of EVs and the discussed dynamics the following hypothesis is proposed:
H1: The more people can be observed using EVs, the more likely it is that observers have a more favourable attitude towards the adoption of an EV.
2.3 Inferences of Status
In order to understand what status people confer to an EV owner the concept of status is briefly summarised hereafter. Magee and Galinsky (2008) define status as a high(er) position compared to others on dimensions such as wealth, respect, hierarchy, etc. Consumption is one form to gain status because products not only serve as means to satisfy basic needs but they can also have symbolic connotations (Levy, 1959). Individuals use, own, and display some goods to present a certain image of for example what they are, feel, and what social relationship they desire (Belk, 1985; Braun & Wicklund, 1989; Eastman, Goldsmith & Flynn, 1999). Status display of products has received a lot of attention ever since Veblen (1899 & 1994) who used the term “conspicuous consumption” to label products that are used to signal status aspirations to others. Such products can be described as “status symbols”.
studies realised status as a motive when people were asked about their purchase intention, preference, or actual ownership of an EV. In their study Griskevicius et al. (2010) already raised the question how the ownership of a pro-environmental car would be perceived and interpreted by others. Using this suggestion as a starting point one might wonder how much status externals confer to an EV owner. Such insights could reveal whether status seeking EV owners are actually able to attain their desired status. If not, it would inevitably produce some undesired reactions by others. Also, observers’ discrete inferences of status display a certain image that is attached to EV ownership.
High status individuals8, in contrast to rather low status individuals, can afford to depart from common and conventional expectations (c.f. Bellezza et al., 2014; Galinsky et al., 2008; Haslam, 2004) and thus tend to engage in nonconforming behaviour (Hollander, 1958). Nonconformity refers to a belief or behaviour that is inconsistent with existing norms and standards (Nail, Macdonald, and Levy, 2000) and can lead to positively afflicted inferences by other (Simonson and Nowlis, 2000). Moreover, observers’ ability to make inferences of a person’s status based on visible signals such as behaviour and consumption choices is identified in various literature (Belk, Bahn, & Mayer, 1982; Bellezza et al., 2014; Calder & Burnkrant, 1977; Dubois, Rucker, & Galinsky, 2012, Magee & Galinsky, 2008, Ridgeway & Correll, 2006). The present study translates a rare diffusion of EVs as something that is deviating from the commonly accepted fact that petrol cars are nowadays standard. In such a situation, a deviating EV owner signals some freedom from the pressure to conform to the standard (car choice), thereby accepting hurdles or inconveniences. Such a individual might be more powerful, wealthy, or important in its social and/or job position. In contrast to that, EV ownership will not be perceived as that deviating anymore once a critical mass already uses them. Based on this reasoning the following hypothesis can be derived:
H2: The more people own an EV, the smaller will be the status external observers derive from the ownership of an EV.
8 Increased status can be gained through contrary credits or a collection of positive impressions in the
2.4 Individuals’ Need for Uniqueness & Corresponding Behaviour
Uniqueness theory implies that a need for uniqueness can be seen as an extension of the self (Snyder & Fromkin, 1977). In their study the authors found that uniqueness seeking arises from “early childhood socialization” in which norms were either followed or creativity and individuality stressed. Other studies emphasise “counter conformity” motivation or a threat to be too similar to others as drivers of strong need for uniqueness (Snyder, 1992; Snyder & Fromkin, 1980; Nail, 1986). According to Tian et al. (2001) an individual’s “pursuit of differentness is achieved through the acquisition, utilization, and disposition of consumer goods for the purpose of developing and enhancing one’s personal and social identity”. This social identity goal is also highlighted in Ariely & Levav’s (2000) “wine study”, which found that, in a group context, individual’s choice behaviour could function as a self-presentation cues in the form of uniqueness.
2.4.1 Influence on Adoption of Electric Vehicles
Some authors found uniqueness seeking to be important for highly conspicuous and personal important product categories (Berger & Heath, 2007; Campbell, 1986; Kernis, 1984), thus highlighting the importance for EV adoption related research. Considering the relative early stage of EV diffusion, one can find some characteristics typical for consumers in these very first adoption stages. Innovativeness, uniqueness, and non-conformity are constantly found herein (Kirton, 1976; Skinner, 1996). Skinner’s (1996) results even suggested that a high need for uniqueness also represents a preference for innovativeness - and thus for innovative products like EVs. Because degrees of uniqueness-seeking motivation vary in individuals9, it is expected to see differences among observers regarding EV adoption.
When being scarce, the possession or use of an EV should make a unique impression, which is socially accepted and recognisable by observers (Simonson and Nowlis, 2000; & Berger & Heath, 2007). Lynn & Harris’s (1997) findings suggest that individuals with a high uniqueness need prefer a product that deviates from the perceived norm or mass and hence are more in favour of an EV. This preference is well confirmed by extracts from Heffner et al.’s (2007) Californian-based interviews, in which one HEV owner said, “he wanted to make
9 Literature suggests various scales that make individual uniqueness need measurable (Lynn & Harris,
his statement; he wanted to be the one in the block that had the [Toyota] Prius”. Other interviewees said they “got a little bit of pleasure of being ahead of the crowd”. Literature also suggests that high uniqueness scorers have a strong sense for detecting unique behaviours (Snyder, 1992). An EV owner’s unique car choice is likely to attract followers that also have a high need for uniqueness and that share a common link with early adopter groups (Fisher & Price, 1992; Rogers, 1962 & 2010). In contrast to that mechanism Berger & Heath (2007) point out to the fact that products, formerly perceived as cool, may loose their identity signalling ability once outsiders or even the mainstream begin to also consume these products. This statement emphasises motivation of high uniqueness-seekers can change when EVs are no scarce event anymore. As a result of these mechanisms and the symbolic motivation of high uniqueness seeking the following is postulated:
H3a: Observers’ levels of need for uniqueness will moderate adoption attitude depending on the perceived EV diffusion. High uniqueness seekers will prefer adopting an EV when EVs are scarce but not when they are commonplace.
2.4.2 Influence on Status Inferences
As described previously, high need-for-uniqueness individuals often deviate from existing norms and standards in order to arrive at a higher social status. However, one might wonder how a high uniqueness need of observers is related to status observers confer to an EV owner. A valuable answer to that question can be derived from uniqueness literature (Snyder & Fromkin, 1977; Tian et al., 2001) and particularly from recent findings by Bellezza et al. (2014). The authors expanded existing literature on uniqueness and differentness by finding that people who have a high uniqueness need infer more status (and competence) to a nonconforming person as compared to people with rather low uniqueness needs10. Consequently, high uniqueness seekers are also more likely to derive higher status inferences from nonconforming behaviour or product possession. In a surrounding of scarce EVs the ownership of an EV should depict such a nonconforming possession leading to higher status inferences by observers with a high uniqueness need. When EVs are quite commonplace on the contrary a high uniqueness need of observers will not lead to higher status inferences
10 The authors tested how much status (and competence) respondents conferred to a person of which
compared to observers with a low need for uniqueness since there is no nonconforming event detectable anymore. Therefore, the following hypothesis is predicted:
H3b: Need for uniqueness (and nonconformity) of observers will moderate status inferences of an EV owner depending on the perceived EV diffusion. When EVs are scarce high uniqueness seekers will confer more status to an EV owner than low uniqueness seekers.
2.5 Overview of the Present Research
As depicted in figure 1, both dependent variables discussed in the theoretical foundations, namely (1) adoption attitude and (2) status inferences, will be examined with the help of an online survey. In addition, the moderating effect of observers’s need for uniqueness on the two dependent variables will be measured.
3
Research Design & Method
3.1 Participants & Data Collection
In total 124 individuals completed an online survey (44% female, Mage = 32). Average age of
respondents was under average thereby impairing representativeness for overall population. Response rate of all individuals who followed the web link was 74%11. Most of all respondents were either working (48%) or were students (44%). Average net income of respondents was in the range between 1,501 € and 2,000 €. Of all respondents 43% indicated that they reside in Germany, 35% in the Netherlands, 11% in the USA, and 11% in other countries (compare appendix 2 for a overview). All participants filled in the survey in private. The link to the survey was accessible for two weeks. Finally, the survey data was downloaded and analysed with the help of the IBM SPSS program.
3.2 Design and Procedure
To prevent suspicion a cover story was developed (Aronson & Brewer, 1998). Goal of the cover story was mainly to raise validity of results by preventing respondents from presenting themselves in a better light or in a way of which they think it would be most “appropriate” to answer the questions. As a part of this complex cover story respondents were told to imagine a situation in which they have recently moved to a new (geographic) region X in which they will both live and work for a certain time12 (compare appendix 3 for the whole survey). They were further told that moving there left them with some purchase decisions they have to make throughout the survey. These decisions involved (1) renting an accommodation, (2) acquiring a car, and lastly (3) choosing their preferred dress style for work. To assure that the essential information presented in the survey was comprehended carefully respondents were also told, that “throughout the survey memory of presented details will be tested”. Intensive pre-testing of early test versions as well as some conversations with respondents who had participated revealed the cover story worked well since respondents were not sure what the exact focus of the survey was and/or did not identify EVs as the clear focus of the survey.
11 The relatively low dropout rate is assumed to be a result of respondents’ familiarity with the
displayed situations and the survey design that used a lot of appealing pictures
12 Such a situation can be considered as very realistic since over 90% of all respondents indicated that
3.2.1 Manipulation of Perceived EV Diffusion
Two different diffusion stages of EVs were randomly manipulated between-subjects and embedded into the cover story. Respondents were presented with photographs and additional information artificially conveying an environment in which EVs are either scarce (compare figure 2) or rather common (compare figure 3). In the following of this study, these two conditions are labelled as “scarce EVs” and “common EVs”. Respondents were asked to imagine that each situation represents typical car usage in the environment they have been set in with the help of the cover story. Newness of the environment was created to prevent that respondents draw spontaneous inferences based on their knowledge of EV diffusion in their usual environment, thus making an artificial manipulation of EV diffusion possible in the first place.
Figure 2: Depiction of the “scare EVs” condition in the survey13
13 Pictures of the “company car park“ are photographs of the (partly electrified) Google Headquarter
Figure 3: Depiction of the “common EVs” condition in the survey
3.2.2 Measurement of EV Adoption Attitude
distance was chosen because commuting plays an important role in nowadays (electric) mobility and it does not include long-distance travels (Cooper, 2000; Eppstein et al., 201114; Segal, 1995). To create a realistic car acquisition scenario, respondents were told that it would be an attractive option to lease a car from the employer’s company and that the most important difference of cars available to the company car pool is the way they are powered (either electric or petrol-fuelled). After manipulation and introduction to the car acquisition possibility, respondents were presented with the two car types and corresponding questions regarding adoption and interest. Figure 4 shows the question design that was applied here.
Figure 4: Depiction of the questions regarding adoption and interest
14 With the help of an agent-based model the authors found that 73% of all agents commute more than
Car examples used here were represented in form of photographs of similar BMW and Mercedes-Benz mid sized electric and petrol cars.15. The cars were chosen because their manufacturers originate from the same country and they are similar in size, utility and styling16. Moreover, both car types can be found in a similar price range of 35,000 to 45,000€ (Jordan, 2014) and this fact was communicated to respondents by putting them in the same company leasing range. These cars are also expected to represent an option suitable for the mass since they are compact/medium sized, reasonable priced, and suitable for commuting. Since functional drawbacks, economic and environmental advantages were not the explicit focus of this research; respondents did not receive any detailed information on any of this. In consequence they were left with only what they already knew (or not knew) about EVs. The use of (fixed) slider scales for answering the questions was expected to make the tension between both opposing car types more apparent17. The first two questions were integrated to
gain deeper insight into respondents’ attitude towards EVs, whereas the focus regarding actual adoption was clearly on the car type preference question.
3.2.3 Measurement of Inferred Status
After respondents had answered the questions on adoption attitude they moved to the next page of the survey. Next to a question on memory of the details tested beforehand (which was used as a manipulation check) respondents were asked to rate the status of a person who owns an electric car in region X. Inferred status was depicted by a discrete five-star scale (1 star = very low status, 5 stars = very high status). The placement of this measurement right after the diffusion manipulation was supposed to ensure full imagination of the depicted EV situation while at the same time respondents were not able to jump back to check/change previous answers (compare appendix 3 for the whole survey and question flow).
15 To prevent simple preferences based on extras all four cars were shown in a similar exterior styling
and respondents were also told that the depicted models and colours were only exemplary.
16 Indeed, both electric cars have approximately the same one-charge driving range of circa 140 km (87,5 miles). Both Mercedes-Benz cars’ base and body are exactly the same from the outside (Jordan, 2014). Note that, interestingly, the B Class is only available as the Electric Drive version in the US in 2014.
17 After extensive pre-testing this answering design was modified a few times to finally arrive at a
3.2.4 Measurement of Need for Uniqueness
At the end of the survey respondents were asked to fill out an 8-item five point bi-polar scale measuring need/desire for uniqueness. This scale was placed at the end of the survey because a positioning at the beginning of the survey was feared to raise respondents’ detection that uniqueness would be part of the following questions, thereby distorting the findings. For the purpose of this study, Lynn & Harris’ (1997) Desire for Unique Consumer Products (“DUCP”) 8-item scale was used. A clear advantage of this scale is its relative shortness (and thus user friendliness) in comparison to other uniqueness-measuring scales (Snyder & Fromkin, 1977 & 1992; Tian et al., 2001)18 often used in existing research. The used DUCP scale comprised statements like “I am very attracted to rare objects”, “I enjoy having thinks that other do not”, and “I like to try new products and services before others do”. Respondents had to indicate their agreement with the statements (1 = Strongly disagree, 5 = Strongly agree). Internal consistency of the DUCP scale was estimated with the use of Cronbach's coefficient alpha (α = 0.84). The scale displayed adequate internal reliability and was thus used to draw conclusions up on observers’ need/desire for uniqueness.
4
Results
First, a manipulation check was conducted to see whether both EV diffusion conditions evoked the desired perception in observers’ minds. Masked as a question on memory respondents indicated on a five-item scale (1 = very common, 5 = very scarce) how they rate the diffusion of EVs in the presented conditions. The manipulation check confirmed that participants perceived EVs as more scarce when confronted with the “scarce EVs” situation compared to the “common EVs” situation (Mscarce EVs = 4.18, SDscarce EVs = 1.06 vs. Mcommon EVs = 2.11, SDcommon EVs = 1.16, t(124) = 10.32, p < 0.001). Because the manipulation was
successful it was proceeded with the results section. First, the results for adoption attitude and the role of uniqueness need as a moderating variable are presented. Second, the same tests are applied in a similar fashion for observers’ inferences of status of an EV owner and the moderating effect of observers’ uniqueness need.
18 Scales like Tian et al.’s “Need for Unique Consumer Products scale” (“NUCP”) are quite non
4.1 Results for Adoption Attitude
The two questions measuring for which car type respondents show more interest and for which car type they desire more additional information revealed no significant differences between conditions19. Note here, that all respondents, regardless of perceived EV diffusion, indicated a high desire for additional information on EVs (M = 7.08, SD = 2.72). Finally, the results for the question on car type preference were investigated. Respondents showed a (marginally) higher preference for EVs when EVs were perceived as more scarce vehicles (Mscarce EVs = 5.77, SDscarce EVs = 3.14 vs. Mcommon EVs = 4.74, SDcommon EVs = 3.03, t(124) =
1.86, p < 0.10). Note, that, with the p-value of 0.065, significance tends towards a significance level of 0.05 and that difference in scoring on car type preference is more than one unit (compare appendix 6 for test outputs). With these results the hypothesis 1 was interestingly not confirmed but turned the other way around. Respondents did show relatively more preference for an EV when they had been confronted with EVs depicted as more scarce compared to respondents that had been confronted with EVs depicted as more commonplace.
Need for Uniqueness as Moderator. The moderating role of observers’ need for uniqueness
(as measured by the DUCP scale) was analysed with the help of the PROCESS procedure for SPSS (Hayes, 2012 & 2013). Here, car type preference functioned as dependent variable with the following independent variables: EV diffusion condition, observers’ uniqueness need represented by the mean of the 8-item DUCP scale, and their interaction. The regression depicted by the applied PROCESS procedure revealed a (marginal) main effect of the EV condition (β = -0.94, t(124) = -1.71, p < 0.10), a significant main effect of observers’ need for uniqueness (β = 0,78, t(124) = 1.98, p < 0.05), and a non-significant interaction between these two variables (β = 0.23, t(124) = 0.29, NS). Conditional effects of EV diffusion on car type preference at one standard deviation above and below the mean of need for uniqueness showed no significance (compare appendix 7). Likewise, the Johnson-Neyman technique (Spiller, Fitzsimons, Lynch, & McClelland, 2013) run by PROCESS in SPSS detected no statistical significant transition points within the observed range of observer’s need for uniqueness as a moderator. Although observers’ need for uniqueness showed a significant main effect on car type preference (correlation = 0.20, t(124) = 2.24, p < 0.05) there was no significant effect for the interaction of EV diffusion and need for uniqueness detectable. To
19 Tested with an independent sample T-test; Compare appendix 4 & 5 for statistic outputs and exact
further explore the effects of observers’ need for uniqueness on general car type preference (independent from both presented EV diffusions) an analysis of different scorings on the DUCP scale was conducted. In order to do that, scoring on the applied DUCP scales was divided into percentiles . The statistical tests revealed no significant differences in mean scoring in car preference when lower and upper quartiles were compared20. However, when comparing sound lowest quartile (Q1) and highest quartile (Q4) of the DUCP scale, a significant difference in car type preference was detected (Mlowest quartile = 4.32, SDlowest quartile =
2.64, N = 31 vs. Mhighest quartile = 6.18, SDhighest quartile = 3.18, N = 33, t(64) = -2.54, p < 0.05)
with scoring on car type preference showing a difference of 1.86 units on the scale ranging from 0 to 10. Figure 5 depicts car type preference of low (Q1) and high uniqueness scorers (Q4) for EV diffusion conditions and regardless of conditions (compare also appendix 8).
Figure 5: Inferred Status between EV Diffusions for Levels of Need for Uniqueness
Although specific hypothesis 3a for uniqueness scoring in combination with the two EV diffusion conditions was not supported, low and high scorers on the need for uniqueness scale indicated significant differences in car type preference regardless of which EV diffusion was presented.
20 M
lower quartiles = 5.19, SDlower quartiles = 3.17 vs. Mupper quartiles = 5.33, SDupper quartiles = 3.08, t(124) = 0.25,
4.2 Results for Inferred Status
Similar to the structure of the proceeding results section both EV diffusion conditions were examined with respect to their effect on status observers confer to an EV owner. Again, an Independent-Samples T Test was therefore used with observers’ inferred status as dependent variable and both EV diffusion conditions as predictors. The results supported hypothesis 2 since respondents conferred significantly more status to an EV owner when respondents perceived EVs as scare rather than common vehicles (Mscarce EVs = 3.81, SDscarce EVs = 0.92 vs.
Mcommon EVs = 3.34, SDcommon EVs = 0.96, t(124) = 2.77, p < 0.05).
Need for Uniqueness as Moderator. Similar to the moderation analysis under the adoption
attitude part the PROCESS procedure was used again. More specifically, inferred status functioned as dependent variable with the following independent variables: EV diffusion condition, observers’ uniqueness need represented by the mean of the 8-item DUCP scale, and their interaction. The regression depicted by the applied PROCESS procedure revealed a significant main effect of the EV condition (β = 0.44, t(124) = -2.64, p < 0.05), a marginally significant main effect of observers’ need for uniqueness (β = 0,22, t(124) = 1.80, p < 0.10), and a non-significant interaction between these two variables (β = 0.31, t(124) = -1.29, NS). However, when conducting a spotlight analysis (Fitzsimons 2008) and looking at the conditional effect of EV diffusion on inferred status at one standard deviation above the mean of need for uniqueness a significant difference between conditions was revealed (β = -0.67, t(124) = -2.79, p < 0.05); observes with high need for uniqueness inferred significantly more status to an EV owner when EVs were perceived to be scarcely diffused. In contrast, when looking at the conditional effect of EV diffusion on inferred status at one standard deviation below the mean of need for uniqueness difference between conditions was not significant (β = -0.22, t(124) = -0.91, NS). The specific scorings of both high and low need for uniqueness are depicted in figure 6 showing the ascending slope of status inferences for the scarce EVs condition. To further explore the conditional effects of EV diffusion conditions on inferred status effects at different values of the moderator need for uniqueness were explored with the Johnson-Neyman Technique run by PROCESS in SPSS (Spiller et al., 2013). Here, results showed a Johnson-Neyman significance region ranging from 0.27 under the mean of observers’ need for uniqueness (MDUCP – 0.27 = 3.08, β = -0.18, t(124) = -1,98, p = 0.05) to
Figure 6: Inferred Status between EV Diffusions for Levels of Need for Uniqueness
4.3 The potential Relationship between Adoption Attitude and Inferred Status
Recapitulating the previous results, hypothesis 1 was rejected since respondents indicated more preference for an EV when they actually were confronted with the “scarce EV condition”. At the same time hypothesis 2 regarding status inferences by observers was accepted because observers conferred more status to an EV owner in the “scarce EV condition”. The rejection of hypothesis 1 and the aligned scoring on the car type preference scale thus imply a correlation of preference for an EV on the one hand and a higher status conferred to an observed EV owner on the other hand. Actually, both variables showed a correlation of 0.46 indicating a higher inferred status also reveals a higher EV preference. In consequence, this unexpected conclusion will be investigated further making use of a mediation test incorporating the discussed variables.
significant effect of the EV diffusion condition on the mediator inferred status (B = -0.47, t(124) = -2.77, p < 0.05) and a significant effect of the mediator on car type preference (B = 1.44, t(124) = 5.39, p < 0.001). However, when integrating inferred status in the model the effect of EV diffusion condition on car type preference became nonsignificant (B = -0.36, t(124) = -0.69, NS) while it was marginally significant before including the mediator (B = 1.03, t(124) = -1.86, p < 0.10). With bootstrap analysis, it was found that the 95% bias-corrected confidence interval for the size of the indirect effect excluded zero (95% CI = -1.24 to -0.21), suggesting a significant indirect effect (compare appendix 10). The findings consequently implied that inferred status (respondents attributed to an observed EV owner) completely mediated car type preference. However, Hayes (2013) argued that the realisation of a complete mediation should be handled carefully when there is low statistical power for the initial effect of predictor on outcome. Since exactly this effect was not significant at a 0.05 level (2-tailed) but only marginally significant it lacks the statistical power Hayes (2013) pointed to. In consequence, the findings of the applied mediation analysis can be seen as a hint towards a strong relationship between car preference and inferred status but not as a perfect proof for it.
5
Discussion
The present paper manipulated two different EV diffusion stages in a (artificially created) real world setting and introduced a conceptual model in which third party observers’ reactions with respect to their (1) adoption attitude and (2) status inferences of observed EV owner were measured. In addition, the moderating role of observers’ need for uniqueness was measured.
As expected, observes attested more status to an EV owner when his/her electric car was a rather scarce vehicle in the perceived surrounding. It can thus be concluded that an EV is well suited for its owner to display status to externals21 – but only as long as it is unusual or nonconforming to own an EV. This psychological mechanism was recently proved in the work by Bellezza et al. (2014) and the present study expands their findings to the category of cars and EVs. Furthermore, this research broadens EV related theory with the notion that
21 Note, that the findings of this study neglected the fact whether it was the main intention of the EV
status, externals confer to an EV owner, will go down once EVs reach a recognisable critical mass. This, in return, also suggests that (perceived) factors such as uniqueness, newness, or scarcity of these vehicles play an important role for such status inferences. Even though EVs will have the same or even more positive performance features, such as low noise level and advanced driving performance, as well as environmental benefits in a future with higher EV market diffusion, such factors will not lead to the same status inferences compared to situations in which all these factors seem to be quite new and special for consumers.
Looking at the results regarding adoption attitude, this study’s findings interestingly provided an unexpected outcome that counteracted the hypothesis that people should show relatively more preference for an EV when they perceive it as not that scarce and “inconvenient” anymore. The theoretical framework identified a host of arguments, such as increased observability, reduced innocence and scepticism, and the classic diffusion theory, to underline the proposed hypothesis. In the present study however, respondents showed more preference for an EV when they perceived EVs as more scarce. Hence, there must have been some other factor(s) in action that made people prefer EVs when they were rather scarce. Even in theory, there is some ambiguity because some symbolic motivations for acquiring an EV would indeed turn the proposed hypothesis around. After all, findings implied a relationship between inferred status and EV preference. The conducted mediation analysis consequently suggested, at least in the present model, that the extent of status attributed to an EV (owner) was the driving factor that determined respondents’ preference for an EV. Yet, it is important to note that this analysis showed quite small statistical power, thus requiring a crtitcal reflection of the mediation role of inferred status.
type preference.
In contrast to the previous paragraph, there is also evidence suggesting why status intentions could not explain car type preference perfectly. Some researchers emphasised the notion that status motives, especially for highly visible products, referrers to costly signalling. For example, Grikevicius et al. (2010) found that status activation let more people chose an EV when it actually costed more but not less than its conventional counterpart and when it posed some functional disadvantages at the same time. Interestingly, in the present survey, respondents could choose between both car types together with the information that the lease rate for both types would be equivalent22 and there were no noticeable advantages or disadvantages of both car types displayed23. As a consequence it seems unlikely that respondents tended to the EV in the scarce EV condition because the evaluated the EV as more cost intensive than a conventional petrol-fuelled car. Recapping these interrelated dependencies between potential drivers it can be concluded that status might indeed be an important driver when looking at EV adoption. Nevertheless, the argued ambiguity still leaves room for other drivers than status.
Uniqueness seeking, also a symbolic motivation, was integrated as a moderating variable into effects of EV diffusion condition on both dependent variables. Findings of the study showed that respondents with a high need for uniqueness indicated more preference for adopting an EV than respondents with a low need for uniqueness did. In contrast to the proposed hypothesis this effect was surprisingly not significant across EV diffusion conditions. This implies that EVs inhere some attributes that appeal to high uniqueness seekers not only when EVs are scarce objects. Yet, these findings accidentally cannot contribute to explaining the puzzle why respondents in the present study preferred an EV when they perceived it as more scarce. Furthermore, the expectations regarding the moderating role of high need for uniqueness on status inferences were fulfilled. Respondents with a high need for uniqueness conferred more status to an observed EV owner – but only in a surrounding in which EVs were scarce. In sum it can be concluded that individuals with a high uniqueness need play an important role in understanding the psychological dimensions that underlie EV adoption and opinion formation.
22 Both car types (EV and petrol) by BMW and Mercedes-Benz actually fulfil this condition with
price tags ranging from 30,000 to 40,000 €.
23 Previous studies, like the one by Grikevicius et al. (2010), actually presented advantageous and
6
Limitations and Directions for Future Research
Due to this research’s new approach of manipulating two rather different diffusion stages and due to its complexity a few limitations can be identified and avenues for future research derived. In order to raise external validity, both dependent variables could be investigated in a more separate manner. Such an approach would prevent possible order effects or biased answering by respondents. In particular, a separate investigation of status inferences could be applied to the presented mediation analysis, thus verifying reliability of the present findings. A second limitation is derived from the method different EV diffusion was manipulated. The displayed EV usage rate in observers’ relevant surrounding was evoked by a story, which was realistic but still artificially created for the purpose of this study. To even increase external validity of the findings the method of this study could be expanded to the field of real geographic regions and surroundings that could function as the predicting variable. Although a host of regions fulfil the requirements for conditions in which EVs are scarce, the opposite condition is rather hard to detect in up to date reality. Nevertheless, specific areas in California or Norway, which both are benchmark markets for EVs, could are suitable for displaying such a common appearance of EVs24. Extending this study to the real world could also investigate whether respondents still display more preference for EVs in regions with quite low or rather high EV market penetration. At the same time, such a real world setting could incorporate an array of various factors that could influence EV adoption under different EV diffusion. Additional factors could be derived from the theoretical framework of this study. Rather than measuring such factors, future research could activate them in respondents.
Regarding this paper’s findings on status inferences, future research could also analyse on what base(s) observers exactly confer status to an EV owner. Literature from consumer psychology and behavioural science suggests various indicators of status, such as the pure uniqueness of the conspicuous product (Berger & Heath, 2007; Tian, Bearden, & Hunter, 2001), otherness as a result of avoiding blatant and conspicuous consumption (Berger & Ward, 2010; Han, Nunes, & Drèze, 2010; Heffner, Kurani, & Turrentine, 2007), the
24 California is the largest EV market in the US. The US make up 38 % of all EV stocks in the world
acceptance of a higher EV price (Eastman, Goldsmith, & Flynn, 1999; Griskevicius, Tybur, & Van den Bergh, 2010), that an EV owner can afford to deviate from conventions (Galinsky et al., 2008), or even intelligence and moral of the observed person (Heffner, Kurani, & Turrentine, 200725). In addition, one could examine which of these factors influences observers to what extent when forming their attitude towards EVs and their adoption.
A similar approach for further research could be applied to the role uniqueness seeking has on EV adoption. The present study indicated that EVs appealed to high uniqueness seekers not only because they were a rather rare phenomenon on the roads. Future research could raise the question, what actually creates the uniqueness of EVs versus conventional cars. One could suspect that unfamiliar performance features (such as seamless immediate acceleration), a different noise level (Tschampa, 2014), a completely different “fuelling” (charging) experience, or an unusual product design pose perceived uniqueness of EVs. In addition, validity of results as well as insights into uniqueness could be further increased by replacing the applied 8-item scale (which was chosen to prevent a high drop-out rate in the survey) by, for example, the more complex 31-item scale developed by Tian et al. (2001)26.
Another suggestion for further research is related to the field of information. This was only indirectly covered in this study but is expected to have a critical meaning for EV adoption. Consumers will need to have a set of information (of whichever kind) to draw conclusions upon EVs. This study only provided respondents with some pictures and short texts as basic information on EVs. Since researchers (Gatignon & Robertson, 1985, Rogers, 2010, Thøgersen, 2005) rewarded communication with an essential role for products in the early innovation life cycle stages, prospective research could investigate effects of different information and communication sources on people’s EV adoption. Especially in the automotive industry, advertising, for example, has a high importance (Horsky, 1983; Steg, Vlek, & Slootegraaf, 2001) and could be investigated for its moderation effect in EV adoption.
25 The authors found this self-reported attribute when interviewees in their study emphasised the
moral difference between acquiring a HEV or a quite unneeded Sports Utility Vehicle.
7
Managerial Implications
One important implication that should be highlighted first refers to the findings attached to a scarce diffusion of EVs, which is exemplary for the status of most markets in 2014. The fact that EVs are afflicted with higher status in relation with the unexpected effect that respondents showed more preference for an EV in this early market stage (even though reasons for this unexpected behaviour are not fully clear), produces an important bottom line for strategic planning: car companies should not wait any minute longer to bring some serious supply within the EV spectrum to markets, and this is especially true for if companies do not want to not fall behind pioneering brands/companies that already serve EV demand. This implication is even more substantial for companies that offer product ranges, which are intended to appeal to status seekers, or are marketed with status related characteristics. Since, in early diffusion stages, an EV seems to pose status per se, a “top down” market penetration strategy in terms of car luxury level and prices might be a fruitful strategy. Early adopters usually can afford to invest more money into their purchases (Rogers, 2010). Especially premium car suppliers could use this early base of progressive customers to work their way down to more affordable cars and a larger share of customers. This strategic approach is underlined by a recent Forbes article that emphasized an already 3 % market share of EVs in the luxury car segment (Gartner, 2014). There is no recognizable price mark-up for EVs in this car segment and a potential customer might opt for the EV more easily. The strategy applied by California based EV company Tesla Motors might be exemplary here.
Findings from this study also suggested what happens to EV related status attributes once a critical market penetration is reached. Form this point on, inferred status will decrease and the sheer appearance of EVs as rather special and unique vehicles might not be in operation anymore. In consequence, managers of premium car companies that sell EVs with status attributes should be prepared for such a future situation. Once EVs become commonly diffused, status of EV models must be leveraged by other means than sheer otherness. Lessons here could be learned from the petrol-fuelled car when increasing numbers of car models were differentiated in order to attract discrete target groups in the first decades of the 20th century.
difference towards the conventional car by a unique design vocabulary. In reality, there is some ambiguity regarding this design approach. While some brands try to implement EV technology into familiar car design (as Mercedes-Benz does with the B-Class displayed in this study), others try to communicate EV innovativeness with futuristic and quirky designs (BMW, for example, created its all-new “i” range solely for EVs). Managers should thus decide if a prominent display and glorification of electric mobility can fit their strategy. Support for this decision might be derived from special consumer groups that have a relatively high uniqueness need. These individuals can embody a highly valuable clientele for EVs in early market penetration phases. Such target groups should thus be identified and targeted with appropriate marketing tools.
Promoting EVs by inserting them into car fleets might be another fruitful implication. Respondents in this particular study were confronted with quite equivalently presented car alternatives a company car pool offered. Since respondents showed a tendency towards the EV in this situation, a certain “entrance barrier” seemed to be diminished that way. This idea is already used in carsharing services (for example, BMW’s Drive now & car2go) and could be extended to car rental companies, car fleets within public services, company car fleets, and also taxi fleets. The often-cited problem of inconvenient charging processes seems to pose no major difficulty for such fleets. As an additional advantage, such attempts would increase availability and observability of EVs, thus promoting EV adoption in general.
References
Ariely, D., & Levav, J. (2000). Sequential choice in group settings: Taking the road less traveled and less enjoyed. Journal of Consumer Research, 27(3), 279-290.
Aronson, E., & Wilson, T. (86). D-, & Brewer, MB (1998). Experimentation in social psychology. The handbook of social psychology, 99-142.
Bass, F. M. (1969). A new product growth model for consumer durables. Management
Science, 15 (January): 215-27.
Bass, F. M. (1980). The relationship between diffusion rates, experience curves, and demand elasticities for consumer durable technological innovations. The Journal of Business, 53(3), S51-67.
Becker, T. A., Sidhu, I., & Tenderich, B. (2009). Electric vehicles in the United States: a new
model with forecasts to 2030. Center for Entrepreneurship and Technology, University
of California, Berkeley, (2009.1).
Belk, R. W. (1985). Materialism: Trait aspects of living in the material world. Journal of
Consumer Research, 12(3), 265.
Belk, R. W., Bahn, K. D., & Mayer, R. N. (1982). Developmental recognition of consumption symbolism. Journal of Consumer Research, 4-17.
Bellezza, S., Gino, F., & Keinan, A. (2014). The red sneakers effect: Inferring status and competence from signals of nonconformity. Journal of Consumer Research, 41. DOI: 129.125.175.115
Berger, J., & Heath, C. (2007). Where consumers diverge from others: Identity signaling and product domains. Journal of Consumer Research, 34(2), 121-134.
Berger, J., & Ward, M. (2010). Subtle signals of inconspicuous consumption. Journal of
Consumer Research, 37(4), 555-569.
Braun, O. L., & Wicklund, R. A. (1989). Psychological antecedents of conspicuous consumption. Journal of Economic psychology, 10(2), 161-187.
Brownstone, D., Bunch, D. S., & Train, K. (2000). Joint mixed logit models of stated and revealed preferences for alternative-fuel vehicles. Transportation Research Part B:
Methodological, 34(5), 315-338.
Brownstone, D., Bunch, D. S., Golob, T. F., & Ren, W. (1996). A transactions choice model for forecasting demand for alternative-fuel vehicles. Research in Transportation
Economics, 4, 87-129.
California Environmental Protection Agency (2014). Zero Emission Vehicle (ZEV) Program. Retrieved from http://www.arb.ca.gov/msprog/zevprog/zevprog.htm
Campbell, Jennifer D. (1986), Similarity and Uniqueness: The Effects of Attribute Type, Relevance, and Individual Differ- ences in Self-Esteem and Depression, Journal of
Personality and Social Psychology, 50 (2), 281–94.
Cooper, L. G. (2000). Strategic marketing planning for radically new products. Journal of
Marketing, 64(1), 1-16.
Dagsvik, J. K., Wennemo, T., Wetterwald, D. G., & Aaberge, R. (2002). Potential demand for alternative fuel vehicles. Transportation Research Part B: Methodological, 36(4), 361-384.
Diamond, D. (2009). The impact of government incentives for hybrid-electric vehicles: Evidence from US states. Energy Policy, 37(3), 972-983.
Dubois, D., Rucker, D. D., & Galinsky, A. D. (2012). Super size me: product size as a signal of status. Journal of Consumer Research, 38(6), 1047-1062.
Eastman, J. K., Goldsmith, R. E., & Flynn, L. R. (1999). Status consumption in consumer behavior: Scale development and validation. Journal of Marketing Theory and
Practice, 41-52.
Eggers, F., & Eggers, F. (2011). Where have all the flowers gone? Forecasting green trends in the automobile industry with a choice-based conjoint adoption model. Technological
Forecasting and Social Change, 78(1), 51-62.
Eppstein, M. J., Grover, D. K., Marshall, J. S., & Rizzo, D. M. (2011). An agent-based model to study market penetration of plug-in hybrid electric vehicles. Energy Policy, 39(6), 3789-3802.
Fisher, R. J. and Price, L. (1992), An Investigation into the Social Context of Early Adoption Behavior, Journal of Consumer Research, 19 (December), 477-86.
Fitzsimons, G. J. (2008). Editorial: Death to dichotomizing. Journal of Consumer Research, 35(1), 5-8.
Franke, T., Bühler, F., Cocron, P., Neumann, I., & Krems, J. F. (2012). Enhancing sustainability of electric vehicles: A field study approach to understanding user acceptance and behavior. Advances in Traffic Psychology.
Galinsky, A. D., Magee, J. C., Gruenfeld, D. H., Whitson, J. A., & Liljenquist, K. A. (2008). Power reduces the press of the situation: implications for creativity, conformity, and dissonance. Journal of personality and social psychology, 95(6), 1450.
Garcia, R., & Calantone, R. (2002). A critical look at technological innovation typology and innovativeness terminology: a literature review. Journal of product innovation
