Is it time to drive green? Determinants of green consumption behavior for a high involvement innovation, a choice-based conjoint analysis

(1)

Is it time to drive green?

Determinants of green consumption behavior for a high involvement innovation,

a choice-based conjoint analysis

(2)

Is it time to drive green?

Determinants of green consumption behavior for a high involvement innovation, a choice-based conjoint analysis

Author: B.S. (Berend) Venema University of Groningen

Faculty of Economics and Business

Marketing Management & Marketing Research Master Thesis December 14, 2011 Address: Kneppelhoutstraat 2 bis A 3532 EX Utrecht +31(0)6 249 504 25 berendv@gmail.com Student number: 1467549

Supervisor: dr. J.E. (Jaap) Wieringa

(3)

Management summary

Environmental issues have become more and more a daily concern for consumers, which reflects on their behavior and consumption decisions. A major development is found in personal transportation, namely the introduction of the electric car. Academic research into consumption behavior with regard to green or sustainable products has increased, but there is still a great deficit of research focusing on high involvement green products. This is an important research topic since the antecedents of high involvement purchase decisions are different from low involvement purchase decisions and curtailment behavior. In addition, understanding green consumption behavior is essential in order to optimize product performance, infrastructural issues and policies that support consumers to adopt a more sustainable lifestyle.

A literature study is carried out to identify the determinants of green consumption behavior. Three general factors influencing green consumption behavior were found, these are individual factors, environmental factors and product factors. It is conceptualized that environmental factors are incorporated in the individual factors since both factors interact. Subsequently, variables representing the factors were identified. Attitudinal factors, habits and personal capabilities represent the individual factor. The product factors are price, performance, environmental impact and availability. In order to measure the relationship between the independent variables and green consumption behavior of a high involvement product, the electric car was selected. More specifically, data was collected with regard to the preference and purchase intention for battery electric cars (BEVs) and plug-in hybrid electric cars (PHEVs). Since there is a deficit of actual consumer behavior data regarding electric cars, consumer preferences and purchase intention are used as measures for green consumption behavior.

Factor analysis was first performed, which created three consumer attitude components. These three attitude components were used as independent variables in the regression analysis and set as active covariates in the conjoint analysis. Second, a multiple regression analysis was performed to investigate the individual factors’ influence on green consumption behavior. It is found that consumer age, gender, car ownership, occupation and the three consumer attitude components significantly influence purchase intention and therefore green consumption behavior. Interestingly, current driving habits were not found to have any influence. Third, a choice-based conjoint analysis was performed in which all the identified product factors were found to significantly influence green consumption behavior. In addition, five heterogeneous consumer segments were identified, with each different product preferences and a variant consumer attitude. Moreover, price and performance were found to be the most important product factors. The influence of environmental impact is also a major consumer concern, but this differs greatly per consumer segment.

(4)

for a large group of consumers. This especially holds for the first adopters who are not yet open to fully rely on electric driving. Moreover, a smaller group of wealthy, mainly male consumers is found to be attracted to driving plug-in hybrid electric cars, whether for business or pleasure driving purposes. For this segment, sporty and high performance electric cars (e.g. the Tesla Model S) will be ideally suited.

(5)

Preface

A conversation with my friend Ernst about his thrilling experience driving a Tesla Roadster, made me realize a new form of personal transportation would probably soon take over the world. The intriguing subject of the electric car was born, and studying consumer behavior with regard to this green innovation instantly became my goal for the next few months.

I would like to thank Jaap Wieringa for supervising this research, and for his structured and honest feedback. Furthermore, I want to thank Jenny van Doorn for her thorough and helpful feedback in the final phase of this thesis. Moreover, this thesis would not have been completed without the help of Ellen, Yasmin and Henk.

I hope you enjoy reading! Kind regards,

(6)

1. Introduction

1.1 Background information of the problem

The last decade, environmental issues have become more and more an everyday concern. Consumers are increasingly aware of green issues, the state of the environment and the fact that their consumption behavior has the potential to contribute to a better environment. There has been an increase in media attention towards pollution and environmental disasters around the world. In addition, more laws are being made that force companies (and to some extent consumers) to be more aware of the environment (Faiers et al., 2007). Consumers are better educated and able to find information about green issues much easier. There are also much more green products available that substitute environmentally harmful versions, which make it easier for consumer to ‘go green’ (Young et al., 2010).

One of the major developments is found in personal transportation, more specifically; the introduction of the electric car. Most cars currently driven are still propelled by limited available fossil fuels. These cars pollute the environment with their exhaust fumes, of which carbon dioxide (CO2) is a

well known malefactor. Pollution by cars is still a big issue, even though for over two decades car manufacturers have been working on making car engines cleaner and more fuel efficient, partly motivated by stricter governmental regulations. Manufacturers also introduced several types of alternative fueled cars of which the hybrid engine car has won great popularity since the introduction of the Toyota Prius in 1998. High awareness for global warming and the rising gas prices seem to show a suitable situation for the introduction of a new generation of electric cars (Lieven et al., 2011; Oliver and Rosen, 2010). Electric Vehicles (EVs) seem to be a promising technology for reducing greenhouse gas emissions and other impact on the environment by road transportation. Indeed almost all major car manufacturers have developed (concept) cars that use battery electric or hybrid electric technologies (Walter, 2010), although high initial purchase price, long refueling time, a limited driving range and the lack of supporting infrastructure might pose a threat to adoption (Oliver and Lee, 2010; Eggers and Eggers, 2011).

It is still unclear if consumers are willing to adopt this innovation and although academic research into consumption behavior with regard to green or sustainable products has increased in the last few years (e.g. Stern, 2000b; Kaiser et al., 2005; Hughner et al., 2007), studies tackling high involvement technology-based products (e.g. washing machines, cars) in relation to green consumer purchase decisions have been rare (Young et al., 2010).

Increasing the understanding of green consumption behavior is important for environmental and business reasons. From an environmental perspective, it is important to gain understanding about consumption with a negative impact on the environment. This is essential in order to fulfill environmental goals set by the international community (EU, 2007; EU, 2009; OECD, 2002; UNEP, 2007). From a business and marketing point of view, the development of more environmentally friendly products is ineffective when consumers are not willing to adopt new green technology and behavior into their lives. Therefore, understanding of consumer preferences and the underlying psychology is essential (Stern, 2000a; Young et al., 2010).

(9)

(2010) point out, more research needs to be done across product categories (e.g. cars and white goods) that reflect more on consumer’s image, have a higher social risk, are more expensive and require more research prior to the purchase. Results from low involvement green products (i.e. purchasing products that involve little risk, such as coffee, chocolate or meat) cannot be copied directly to purchase decisions concerning high involvement green products, since the latter type has a number of aspects that clearly distinguishes them (e.g. higher perceived risk, higher need for information, more influence of social context; Young et al., 2010). In other words, current literature calls out for more insight in green consumption behavior with regard to high involvement environmentally friendly products.

1.2 Problem statement

The problem statement of this study is to examine the main determinants of green consumption behavior. The results will be obtained to answer the following problem statement:

What are the main determinants of green consumption behavior?

The results in this study are based on consumer preference for a battery electric or plug-in hybrid electric car. In order to find an answer to the problem statement, several other questions will be addressed. These are:

- What is green consumption behavior?

- How is consumption behavior of green products different from consumption behavior of other products?

- What role does consumer (social) environment play in determining green consumption behavior? - What role do consumer capabilities, habits and attitude play in determining green consumption

behavior?

- What role do the product and product characteristics play in determining green consumption behavior?

- Can consumers be segmented based on their preference for green consumption behavior?

1.3 Research uniqueness and relevance

The uniqueness of this research can be addressed to its attempt to not only identify the determinants of green consumption behavior (in the case of preferences for battery electric and plug-in hybrid electric cars), but also to determine their relative importance, by using choice-based conjoint analysis. There is a lack of scientific research determining the (relative) importance of the different attributes of high involvement green products (e.g. electric car). Current research is either linked to isolated determinants of (environmentally friendly) behavior (e.g. Cleveland et al., 2005; Eriksson et al., 2006), incentives effectively changing behavior (e.g. government incentives, Diamond, 2009; Potoglou and Kanaroglou, 2007), developments and evolution of the industry (e.g. Dijk and Yarime, 2010) or in the specific case of automobiles; focused on traditional hybrid cars and other alternative fuelled cars (e.g. Oliver and Lee, 2010; Caulfield et al., 2010). There is in other words a deficit of scientific research into green product attributes, especially in their relationship to personal characteristics, current use and consumer attitude (Lieven et al., 2011; Eggers and Eggers, 2011).

(10)

is a gap in the understanding of green consumption behavior in relationship to high involvement products that are marketed as green (Oliver and Lee, 2010, Young et al., 2010). Follows and Jobber (2000) identify that this is because of a lack of high involvement green products available on the market and a dominant focus on post purchase and non-consumption behavior in research. Therefore, there is limited understanding of the factors that obstruct the diffusion of green products that have the potential to contribute to a more environmentally friendly future.

Limited understanding is also due to the fact that high involvement green products (e.g. electric cars) are only marginally distributed. Research conducted in this field is therefore often not based on actual consumer experience and purchase data, but on current consumer attitude and knowledge about the new innovations. Nonetheless, Rogers (2003) states that it is important to examine the adoption process while it is still in progress, when the innovation is still spreading and not already adopted on a large scale. Furthermore, since the adoption process is still in infancy, only the future will tell if the identified relationships are indeed valid. Looking back on this field of research in the future can therefore contribute to the optimization and validation of research methods of research into innovative high involvement environmentally friendly products. The relevance of this research can therefore be seen in adding new insight in the adoption of high involvement green products and identifying important hurdles that have to be taken in the diffusion of the new innovation.

The insights presented in this research have managerial contributions, since they will clearly identify the relative importance of determinants of green consumption behavior. This information is ideally suited for managers when predicting the adoption rate of their green products, and has the potential to serve as input for marketing and green product development strategies. Concluding, several studies are available about preferences when buying an internal combustion engine car (e.g. De Hoop, 2009; Manski and Sherman, 1980). Expanding that knowledge in the area of electric cars is essential for companies in order to determine how they should adapt their current marketing strategies to successfully launch their innovation.

1.4 Structure report

(11)

2. Theoretical framework

In this chapter the theoretical framework will be presented. First, green consumption behavior will be defined and the context of this study will be stipulated. Furthermore, an analysis of the determinants of green consumption behavior is provided, also explaining the difference with ‘normal’ consumption behavior. In addition, the motives for adopting green consumer behavior are discussed, leading to hypotheses, which are the basis for the conceptual model.

2.1 Defining green consumption behavior

“Green, a symbol of life, good health, and vigor, a color that reminds us of hope.”

1 The word ‘green’ is associated with environmental protection and energy conservation. In addition, Hoyer and MacInnis (2008, p. 3) give the following definition of consumer behavior: “Consumer behavior reflects the totality of consumers’ decisions with respect to the acquisition, consumption, and disposition of goods, services, activities, experiences, people, and ideas by (human) decision-making unit (over time).” The term ‘green consumption behavior’ therefore entails the full range of decisions and activities, regarding acquiring, consuming and disposing of green products, recycling of materials, efficient use of energy, environmental protection and the preservation of species (Alfredsson, 2004). Moreover, every time a consumer makes a decision about whether or not to purchase a product (or service), this decision has the potential to contribute to a sustainable pattern of consumption (Young et al., 2010). With regard to this, Stern (2000b, p. 408) identifies green consumption behavior as ‘environmentally significant behavior’ and defines this construct according to its impact: “The extent to which it (i.e. behavior) changes the availability of materials or energy from the environment or alters the structure and dynamics of ecosystems or the biosphere itself”.

Moreover, Stern (2000b) distinguishes several distinct types of pro-environmental behavior (i.e. green consumption behavior) and states that different combinations of causal factors are determinants of the different context types of environmentally friendly behavior. Determining the context of green consumption behavior is therefore essential for the definition, and in order to identify relevant determinants. The four context types will be discussed briefly in order to identify the context in which this study should be seen. The context types are: (1) environmental activism, (2) non-activist behavior in the public sphere, (3) private sphere environmentalism and (4) behavior in organizations.

Environmental activism: Entails the active involvement in environmental organizations and demonstrations. The major focus of research in this context is therefore on social movement participation. Much research focuses on the process of ‘recruitment’ of individuals who become activists (McAdam et al., 1988)

Non-activist behavior in public sphere: This context type is related to behavior like signing a petition on environmental issues and joining environmental organizations, which is viewed as more active kind of environmental citizenship. It includes the acceptance of public environmentally friendly policies.

1

(12)

This kind of behavior affects the environment indirectly, but can however have a large effect, since it affects public policies (Dietz et al., 1998; Stern et al., 1999).

Private sphere environmentalism: Is defined as: “The purchase, use, and disposal of personal and household products that have environmental impact.” (Stern, 2000b). A distinction should be made between the purchase of major household products that impact the environment (e.g. cars, household energy), the use and maintenance of environmentally harming or friendly goods (e.g. air conditioning, sun beds), waste disposal and other green product consumption (e.g. the purchase of organic food and recycled products). Private sphere behavior differs from public-sphere environmentalism in a way that the former has a direct consequence for the environment. In addition, the environmental impact of individual behavior is limited, but does have a major effect on an aggregate level.

Behavior in organizations: By working and acting within an organization, individuals may significantly influence the environment. This effect is caused by the actions and choices that individuals make when working in an organization. For example, an engineer chooses to design a product in a more environmentally friendly way or a banker takes environmental criteria into account when investing (Stern, 2000b). According to several authors (Stern and Gardner, 1981; Stern, 2000a), this type of behavior greatly impacts the environment because actions of organizations are a large direct source of many environmental problems. In addition, individual behavior determinants within organization are likely to be different from political and household behavior.

In conclusion, this study focuses on green consumption behavior determinants when purchasing consumer products and should therefore be seen in a ‘private sphere environmentalism’ context. In other words, determinants are discussed that influence purchase, use and disposal of environmentally friendly products by consumers.

2.2 Introduction into determinants of green consumption behavior

The factors determining green consumption behavior (i.e. environmentally significant behavior) in a private sphere environment, may differ greatly per person and per situation. This section serves as an introduction and outlines the complexity of the relationships that are examined in this study.

Green consumption behavior can be part of personal habits or household routines (e.g. switching the lights off, turning down the thermostat). It can be limited by income or infrastructure (e.g. use of public transportation, insulation of the house) and environmental factors can be a marginal consideration in a bigger decision (e.g. deciding on central air conditioning in a new house or the size of the engine in a car). In addition, the environmental impact of behavior does not always have to be known by the individual, who may be unaware of the production process of different companies. It could also be that people believe they take the environment into consideration, but in fact their behavior does not have any effect (e.g. deodorant sprays do not contain harmful substances for the ozone layer). Moreover, people might not act based on environmentally conscious considerations, but do positively affect the environment with their behavior. For example, behavior could be caused by the possibility to save money (e.g. subsidy on green products) or to confirm a sense of personal competence or status.

(13)

consumption behavior has multiple determinants, depending on the type of behavior and involvement with the product (Stern, 1997; Black et al., 1985; Cleveland et al., 2005; Roberts and Bacon, 1997). As mentioned previously, there is limited understanding of green consumption behavior in relation to high involvement products (Young et al., 2010; Follows and Jobber, 2000; Minton and Rose, 1997; Thøgersen, 1999). On the other hand, rich knowledge has been developed on the behavior of consumers from other perspectives, like environmental psychology. Those insights are potentially useful in order to understand the determinants of green consumption behavior.

Faiers et al. (2007) produced a review and categorization of consumer behavior theories. They identified critical internal (i.e. individual or attitudinal) and external (i.e. environmental or contextual) factors that influence consumer choices with respect to energy use. Faiers et al. (2007) based their theory selection on a thorough review from Jackson (2005), who showed that beyond rational choice and cognitive assessment there are emotional, social and cultural influences and issues that also effect individual consumption and behavioral choice. According to Faiers et al. (2007), this leads to three general factors that influence the consumption decision; (1) the environment, (2) the individual and (3) the product. Identifying a causal relationship between these three central factors is critical in understanding green consumption behavior (Faiers et al., 2007).

In the following section the three factors (i.e. environment, individual and product) will be discussed, structured and elaborated on in more detail. Furthermore, hypotheses will be formed leading to a conceptual model of green consumption behavior. Since individuals are part of and interact with their environment both factors will be taken into account together (Stern, 2000b; Black et al., 1985).

2.3 Environmental and individual factors

(14)

Figure 1: ABC model. Adopted from Guagnano et al. (1995)

Stern (2000b) points out that the ABC model, although providing valuable insight, is a simplified model that needs further elaboration to fully grasp consumer behavior. He elaborates on the ABC model, and goes deeper into the personal-contextual or organism-environmental distinction. In his research Stern (2000b) groups the causal variables of consumption behavior in four major types that are significant determinants of environmentally friendly behavior (i.e. green consumption behavior). He identifies (1) contextual forces and (2) attitudinal factors which resemble the variables in the ABC model. Furthermore, Stern (2000b) found (3) habits (or routines) and (4) personal capabilities to be additional determinants of environmentally friendly behavior. The four general categories will be elaborated on below.

2.3.1 Contextual forces

The first causal variable of environmentally friendly behavior, according to Stern (2000b), is the external or contextual force. Like the term suggests, there is a wide variety of external factors influencing an individual’s behavior. These include expectations of the community on how to behave, interpersonal influences (e.g. persuasion from friends and colleagues), regulations of governmental organizations, advertising and other institutional and legal factors. Furthermore, there are monetary costs and incentives (e.g. reduced car taxation, BPM, for fuel efficient cars), the physical difficulty of specific actions (e.g. living 30 kilometers away from work makes it impossible to cycle to work, for most people) and the limitation of capacity and constraints by technology and built environment (e.g. availability of bicycle lanes, solar and wind energy technology). In addition, government policies (e.g. recycling programs for batteries and plastic) and several broader societal, political and economic factors influence behavior (e.g. oil price, interest rate on financial markets, government sensitivity to interest group pressure). Moreover, contextual forces may have different meanings to people with different attitudes or beliefs. For example, the higher price of biological products may be an indication of a superior product to some people, whereas for others it is an economic barrier.

Black et al. (1985) have conceptualized contextual forces as affecting behavior indirectly through attitudinal factors. This places attitudinal factors in the center for understanding green consumption behavior from both marketing and psychological perspective (Alwitt and Berger, 1993; Stern, 2000b). An individual’s attitude is influenced by his environment and therefore environmental factors are incorporated in consumer attitude. Consumption attitude can therefore be seen as a context-specific tendency that connects personal values to consumption level attitudes and behavior (Cleveland et al.,

(15)

2005; Pickett-Baker and Ozaki, 2008). Moreover, environmental psychology research has mainly been focused on attitudinal factors due to their convenience in explaining green consumption behavior across contexts and segments of the population (Diamantopoulos et al., 2003). Therefore, no hypotheses are formulated concerning the direct influence of contextual forces (i.e. the environmental factor) on green consumption behavior. In addition, the specific relationship between contextual forces and attitudinal factors is outside the scope of this research.

2.3.2 Attitudinal factors

The second determinant of environmentally friendly behavior is consumer attitude (Stern, 2000b). Attitude can be defined as the way an individual views or behaves towards an object, often in an evaluative manner (Moore, 2001). Attitudes, also named ‘social–psychological antecedents’ (Faiers et al., 2007), have been found to be key determinants of environmentally conscious behavior (Stern, 2000b). In addition, attitudes, such as political orientation, environmental concern, and in particular ‘perceived consumer effectiveness’ (i.e. PCE) have been found to have a causal link with behavior (Lee and Holden, 1999; Roberts, 1996). Furthermore, research by Laroche et al. (2001) on factors influencing ‘willingness to pay’ has proven that attitudes are effective predictors of environmentally friendly behavior.

There are several theories that take into account attitudinal factors influencing (green) consumption behavior. According to Stern (2000b) attitudinal factors include values, beliefs and norms, which guide consumers to act with pro-environmental intention and can influence all behavior an individual considers to be important from an environmental perspective. The value-belief-norm theory (VBN theory) has been found valid in a variety of research into green consumption behavior, such as conservation behavior (Kaiser et al., 2005), household energy use (Poortinga et al., 2004) and reduction of car usage (Nordlund and Garvill, 2003; Eriksson et al., 2006). The separate items of the value-belief-norm theory (also presented graphically in figure 2) will be briefly discussed below.

Figure 2: Schematic presentation of variables in the VBN theory. Adopted from Stern (2000b) Values: Several types of values have been proven to influence green consumption behavior. So called biospheric, social-altruistic and egoistic values have been found to be strongly related to activating pro-environmental personal norms (De Groot and Steg, 2008; Hansla et al., 2008; Stern et al., 1999). Social-altruistic and biospheric values have a positive relation with green consumption behavior and egoistic values have a negative relation (Cleveland et al., 2005; Nordlund and Garvill, 2002). People that

(16)

are aimed at biospheric values, will mainly base whether they will act green or not on the perceived costs and benefits for the biosphere and ecosystem as a whole. Individuals with a socio-altruistic orientation will mainly base their green consumption choices on the perception of costs and benefits for others. Finally, people with an egoistic value orientation will mostly consider their personal costs and benefits when considering green behavior; if the benefits exceed the costs for them personally they will act in an environmentally friendly way (De Groot and Steg, 2008).

Beliefs: Values influence an individual’s beliefs, and beliefs have been found to effect green consumption behavior, resulting in an ecological worldview. A belief is an assumed truth and individuals create beliefs to anchor their understanding of the world. Once a belief has been formed, individuals will tend to persevere with that belief (Stern et al., 1999). Stern et al. (1999) found that norm based actions (i.e. green consumption behavior) result from three factors: (1) the acceptance of particular personal values (i.e. certain ecological worldview), (2) the belief that things important to those values are under threat (i.e. adverse consequences for valued objects), and (3) the belief that behavior initiated by the individual can help alleviate the threat and restore the values (i.e. perceived ability to reduce threat).

It is proven that if an individual believes that the consequences of his or her behavior affects the environment (i.e. adverse consequences, AC) and ascribes responsibility (AR) for taking pre-emptive action, a pro-environmental norm will develop with the potential to actually influence behavior (Bamberg and Schmidt, 2003; Stern, 2000b). This so called ascription of responsibility (AR), by Schwartz (1977) also termed as responsibility denial, is closely related to Thøgersens’ (1999) concept of ‘perceived consumer effectiveness’ (PCE). The belief of ascription of responsibility and perceived consumer effectiveness are found to be positively related to several types of consumer behavior, like recycling (Guagnano et al., 1995), reducing car use (Tanner, 1999) and the acceptance of environmental policy (Steg et al., 2005).

Norms: The final concept of the value-belief-norm theory is personal norms. Personal norms are experienced by individuals as a feeling of moral obligation to act, and are needed to create a willingness to act pro-environmentally (Stern, 2000b). Personal norms are shaped by the adoption of social norms in a consistent personal value system. In several researches personal norms have been found to be predictors of green consumption behavior. For example, research found that personal environmental norms are positively related to the purchase of “a simple market basket of mundane, non-food, non-durable consumer goods” (Minton and Rose, 1997, p. 40). In addition, personal norms have a positive effect on environmentally friendly travel mode use (Hunecke et al., 2001; Nordlund and Garvill, 2003), positively influencing the purchase of organic wines (Thøgersen, 2002) and influencing the willingness to pay a higher price for pro-environmental food (Widegren, 1998).

(17)

The way consumers view themselves, the consumer self-image, is likely to influence the preference for attributes and the willingness to purchase a product (Oliver and Lee, 2010). Sirgy (1982) states that self-image congruence captures the way a consumer feels a product relates to his view of who he would like to be and of who he is. This is related to social value because others can influence the consumer’s self-image. Nevertheless, according to Bearden et al. (1989) consumer’s self-image is an independent predictor, even though others can shape or enhance the consumer’s view of himself. Consumers are most likely willing to pay more for products and services that are in line with their belief system, which leads to:

Hypothesis 1a: Consumers who believe green consumption behavior reflects positively on their self-image are more likely to adopt green consumption behavior.

In addition, Laroche et al. (2001) identified that consideration for environmental issues when making a purchase decision, is a type of behavior that affects consumer willingness to pay more for an environmentally friendly product. This consideration was included in their research by asking consumers whether they avoid companies that pollute and use non-environmental friendly packaging. Especially in the case of high involvement products, consumers are likely to search for information regarding the product attributes and the company offering the product, to learn if the company has a reputation as a polluter. Therefore, consumer motivation to seek information about the environmental impact of the product they consider purchasing, can affect their behavior. What is more, Oliver and Lee (2010) found a positive association between consumers in both the United States and Korea seeking green product information and the intention to purchase a hybrid car, a high involvement green product. This leads to:

Hypothesis 1b: Consumers seeking green product information are more likely to adopt green consumption behavior.

As stipulated before, the purchase of high involvement products cannot be examined correctly without an environmental or social context. The social surrounding (e.g. friends, neighbors, colleagues, opinion leaders etc.) influences consumer consumption patterns. In addition, social value is assigned by individuals to a product, which evolves during the interaction with information and with others (Deffuant et al., 2005). If a consumer feels that a product has high social value he will look for information that will help to evaluate the benefits of purchasing a product. This especially holds for high involvement products, which are often symbols of personal identity, status and luxury (Kadirov and Varey, 2010). That is why the way others perceive the consumers choice of a product is an important factor in the purchase consideration (Oliver and Lee, 2010; Jansson et al., 2010), leading to:

(18)

What is more, it is important to consider the degree to which consumers feel their own behavior and decisions can make a difference (Oliver and Lee, 2010). If consumers are convinced that their own actions or purchase decisions can contribute to a better environment (i.e. higher perceived consumer effectiveness), they are more likely to consider behavior that positively influences the environment (Oliver and Lee, 2010; Thøgersens, 1999). On the other hand, consumers who believe their purchase decisions are too insignificant to affect the environment, will be less motivated to adopt green consumption behavior. This leads to:

Hypothesis 1d: Consumers who perceive that their own behavior can affect the environment positively, are more likely to adopt green consumption behavior.

2.3.3 Habits or routines

The third distinct variable of Stern’s (2000b) determinants of environmentally friendly behavior is habit (or routine). A habit can be described as an automatic link between a goal and specific behavior that demands very little attention and subsequent behavior (Verplanken et al., 1997). Changing behavior often requires a person to change old habits and establishing new habits (Dahlstrand and Biel, 1997). Thøgersen and Ölander (2006) found there are three requirements in habit research that must take place for a habit to evolve: (1) frequent repetition of behavior, (2) behavior must take place in a stable surrounding and (3) rewarding consequences should be available. Habits therefore influence intentions and willingness to change behavior.

A good example of habitual behavior is travel mode choice. Under normal conditions, travel mode choice was found to have an extremely high level of repeated behavior (i.e. temporal stability, Thøgersen, 2006). Research found mounting evidence of this being one of the reasons that travel mode choice becomes habitual (Bamberg et al., 2003; Gärling et al., 2001; Verplanken et al., 1998; Aarts et al., 1998). Since habits result in automatic response, they are independent from an individual’s attitude, and therefore may deviate from what an individual is intended to do in a specific situation. Therefore, it can be hypothesized that if the adoption of green consumption behavior requires a change in habitual behavior it is less likely that green consumption behavior will be adopted, leading to:

Hypothesis 2: If habits have to be adjusted, it is less likely an individual adopts green consumption behavior.

2.3.4 Personal capabilities

(19)

educational level. In addition, significant financial investments (i.e. higher income) are required to renovate a house so that it becomes energy conserving.

Due to the practical segmentation possibilities, several studies try to find the ‘environmentally friendly consumer’ based on profiling using socio-demographic data. Research was conducted into the impact of age, gender (Roberts, 1993; Olli et al., 2001) and the impact of educational level (Olli et al., 2001; Vlosky et al., 1999; Pedersen, 2000). Although solely looking at the impact of socio-demographic variables could lead to an over simplification of the causal relationship, these insights lead to the conclusion that the stereotypical view of the ‘green consumer’ is a young female, well educated, liberal and wealthy (Gilg et al., 2005).

Moreover, there are several insights found in research concerning the relationship between personal capabilities and preference for alternative fueled vehicles (AFVs), a type of high involvement green product. In addition to previous findings concerning the ´environmentally friendly consumer´, income and level of education were again found to have a positive association with green consumption behavior with regard to AFVs (Ong and Hasselhoff, 2005; Potoglou and Kanroglou, 2007; Segal, 1995). A report of J.D. Power and Associates (2008) also identified hybrid vehicle owners to be higher educated, having a higher household income, but found that they were about 4 years older than the average new car buyer.

Female consumers supposedly care more about the environment and feel more responsible for environmental issues (Roberts, 1993). This could also hold for slightly older consumers who are less self-centered and more aware of their individual responsibility for the environment (Olli et al., 2001). In addition, as stipulated before, a higher educated consumer could have more knowledge about environmental issues, and for consumers with a higher income it is easier to afford new (more expensive) environmentally friendly products. These insights lead to the following hypotheses:

Hypothesis 3a: Female consumers have a higher chance of adopting green consumption behavior.

Hypothesis 3b: Higher educated consumers have a higher chance of adopting green consumption behavior.

Hypothesis 3c: Consumers with a higher income have a higher chance of adopting green consumption behavior.

Hypothesis 3d: Older consumers have a higher chance of adopting green consumption behavior. In addition, although no literature background is provided in this field, marital status and household size possibly play a role in adopting green consumption behavior with regard to high involvement products and will therefore be taken into consideration in this study.

2.3.5 Modeling environmental and individual factors

(20)

environment). Environmental intent and impact are often not the same. This is identified in literature as the ‘attitude/ behavior gap’ or ‘value/ action gap’ (Young et al., 2010). For example, consumers in the United Kingdom showed favorable attitude towards organic food (between 46 to 67% of the population), but only 4 to 10% actually purchased (i.e. behavior) the different products (Hughner et al., 2007). In addition, Truffer et al. (2001) found that although 20% of the consumers state that they are willing to pay 10 to 20% more for green energy products, less than 1% actually adopts green energy sources.

Nevertheless, intentions can be seen as the main antecedent of consumer behavior (Ajzen, 1991). According to Ajzen (1991), intention captures the motivational factors and helps to predict behavior through these factors. The stronger the intention to behave in a certain manner, the more likely a person engages in that behavior. These insights lead to the model presented in figure 3.

Figure 3: Model of environmental and individual factors in relation to green consumption behavior. Interpreted from Stern (2000b), Black et al. (1985) and Ajzen (1991)

2.4 Product factors

In the previous section the individual and the environmental factors influencing consumption behavior were discussed. As found in section 2.2., the product itself also plays an important role in determining behavior. Therefore, the product factor (i.e. product attributes) will be elaborated on in the following section.

From a marketing perspective it is useful to look at the relationship between the product attributes and green consumption behavior. Determining which product attributes are perceived to be most important by consumers, leading to behavior (i.e. purchase), can help to shape marketing and communication strategies (Jobber, 2004). In addition, this provides useful insight for product design and for understanding the willingness to adopt the new product or service (Faiers et al., 2007).

A functional theory to analyze product attributes of innovative products is the Diffusion Theory (Rogers, 2003). The theory models consumer expectations (of services and products) against (innovative) product attributes. It suggests that these expectations affect the willingness to purchase (i.e. consumption behavior). Moreover, consumer expectations are formed by the product attributes. Five predominant attributes of the product are presented: relative advantage, compatibility, complexity (or simplicity), trialability and observability (Rogers, 2003). Consumers assess the attributes in a stepwise process, starting with relative advantage, followed by compatibility and complexity. In each step the consumer can reject the new product. In addition, the attributes relative advantage, compatibility and complexity have been found have the most influence on consumption behavior (Dunphy and Herbig, 1995; Martinez et al., 1998; Rogers 2001). Therefore, these three attributes and the subsequent hypotheses will be discussed below.

Attitudinal factors Habits (or routines) Personal capabilities

Individual factor

Intention Green consumption behavior Contextual forces

(21)

The first attribute ‘relative advantage’ is defined as the marginal advantage of an innovative product over existing products (Rogers, 2003). It is essential to notice that relative advantage is the way potential consumers perceive this advantage, rather than the actual performance of the product (Aggarwal et al., 1998). This can for example be influenced by the enhancement of status related to the ownership of a product (Martinez et al., 1998; Rogers, 2003). Several authors found price and performance to be most significant when consumers perceive relative advantage (e.g. Bhate and Lawler, 1997; Janssen and Jager, 2002; Faiers et al., 2007).

To elaborate on this, it is common knowledge that price has an inverse relationship with demand. Although a higher price may also indicate a higher quality and imply that a consumer is buying a technological and environmentally superior product. In addition, a higher price may influence consumer expectations about additional performance or other attributes of the product, which affects consumption preference (Jobber, 2004). Nevertheless, it is assumed that consumers prefer the lowest price, if this price is still reasonable for the presented product. This leads to:

Hypothesis 4a: A higher price has a negative influence on green consumption behavior.

Performance is also identified as a significant variable when viewing relative advantage. In order for consumers to consider and adopt a new green product, the performance of the product should at least be competitive with the current offering or the loss in performance should be compensated by other product attributes (Janssen and Jager, 2002). This leads to:

Hypothesis 4b: A higher performance has a positive influence on green consumption behavior. The second Diffusion Theory attribute ‘compatibility’ is described as how the product or service fits an individual’s attitude, values and behavior (Rogers, 2003). According to several studies, the product impact on the environment and the product performance are part of compatibility (Janssen and Jager, 2002; Stern, 2000b). The fit between the individual and the product (i.e. compatibility) is captured by the moderating effects presented in section 2.5. In addition, the relationship between performance and green consumption behavior is captured in hypothesis 4b. With regard to product impact on the environment, it is only logical that compared to non-green products, green products have a positive impact on the environment. Furthermore, there is a difference in impact on the environment between different green products, due to production methods and impact during the use of the product (e.g. emission level of greenhouse gases, electricity consumption of washing machine). Due to rising awareness of the importance of more sustainable consumption, products that have less impact on the environment (when all other attributes are the same) are supposedly preferred (Stern, 2000b; Roberts and Bacon, 1997). This leads to:

Hypothesis 4c: The more a product contributes to a better environment, the higher the chance of green consumption behavior.

(22)

The first element is the availability of a green product at purchase locations (i.e. stores and dealers). The more purchase locations that offer a product, the more a consumer is exposed to the product, leading to a higher awareness, which may lead to a higher chance of consumption (Williams et al., 2001).

The second availability element relates to choice options. For many product categories (e.g. coffee, chocolate) there are many brands and different flavors and sizes. Often a green product is only one of these many options in the category. If consumers have or perceive more choice (i.e. variety), it is more likely that one of the choice options fits their needs and preferences, which will lead to more consumption (Kahn and Wansink, 2004).

The third element of availability relates to infrastructure and facility that supports the consumer in the use of the product after it is purchased, especially in the case of high involvement products. For high involvement products that have a sophisticated technology (e.g. computers, cars) it is important that consumers know that after the purchase, they can rely on essential infrastructure and supporting facilities (Young et al., 2010). This is important, in order to use the product (e.g. availability of internet connection, refueling stations) and in the case of maintenance or reparation.

In subsection 3.2.6 the three elements of availability will be discussed in more detail. It is argued that only the third element of availability is directly included in this research, this leads to:

Hypothesis 4d: A better availability of infrastructure and supporting facilities has a positive influence on green consumption behavior.

To summarize the previous findings, there are four product factors hypothesized to influence green consumption behavior: (1) price, (2) performance, (3) impact on the environment and (4) availability. These four product factors represent the significant attributes from the Diffusion Theory as shown in table 1 below. In addition, the corresponding hypotheses are shown.

Diffusion Theory Attributes Product characteristic Hypothesis Relative advantage Price

Performance

4a 4b Compatibility Impact on environment

Performance

4c 4b

Complexity Availability 4d

Table 1: Overview product characteristics and hypotheses

2.5 Moderating effects

In the previous chapters, the direct effect of individual and product factors on green consumption behavior is discussed. In addition to the direct effects, moderating effects are identified. These moderating effects will be discussed in this section.

(23)

to a better environment, he or she will probably prefer a product that is more environmentally friendly. Taking these findings into consideration, leads to:

Hypothesis 5a: Consumer attitude moderates the effect of product factors on green consumption behavior.

In addition, consumers will (subconsciously) look for products that fit their habits, in order to minimize the necessity of changing their habitual behavior (Dahlstrand and Biel, 1997). This affects the preference for product attributes (e.g. performance, availability), and therefore moderates the effect these attributes have on green consumption behavior, which leads to:

Hypothesis 5b: Habits or routines moderate the effect of product factors on green consumption behavior.

Moreover, it is hypothesized that personal capabilities moderate the effect of product factors on consumption behavior (Stern, 2000b; Faiers, 2007). To illustrate this with regard to socio-demographics: for consumers with a higher income it is easier to pay a higher price, which influences the price relationship with green consumption behavior. Furthermore, consumers with higher income were found to drive more kilometers than consumers with a lower income (Steg et al., 1997), which has an effect on the performance requirements. In addition, individuals who are higher educated may be more aware of the effect of pollution on the environment, and place more weight on the environmental impact of a product. Moreover, men use a car more often as a status symbol and women value the practical aspects more (De Hoop, 2009), which influences performance and availability preferences. These insights lead to:

(24)

2.6 Modeling green consumption behavior

In the previous chapter the three factors (environmental, individual and product) resulting in intention and influencing the green consumption behavior were discussed. In addition, a model for environmental and individual factors influencing green consumption behavior was presented (figure 3, in subsection 2.3.5). Adding the product factors and the stated hypotheses to this model leads to the conceptual model shown in figure 4. The conceptual model shows the relationship between product factors of innovative high involvement products and individual factors influencing green consumption behavior. In addition, the moderating effect of the individual factors on the product factors is graphically presented (H5). As shown before, the environmental factor has been found to be incorporated in the attitudinal factors and therefore no separate hypothesis is formulated.

Figure 4: Conceptual model with hypotheses

As discussed previously, little is known about the relationship between environmental, individual and product factors and consumer behavior with regard to innovative high involvement green products. This research attempts to clarify the connection between the product factors, individual factors and green consumption behavior, by investigating consumer preference for battery electric cars and plug-in hybrid electric cars on the Dutch market. In the next chapter (chapter 3) the research methodology and the measurements for the presented variables will be discussed.

Price difference Performance difference

Environmental impact difference

(25)

3. Research design

In this chapter the research design choices and considerations will be discussed. This is structured by first discussing the research method, followed by an elaborate discussion about the measurements of the variables. In addition, the data collection method will be described, sampling issues will be stipulated and finally the plan of analysis will be briefly discussed.

3.1 Research method

Two multivariate data analysis techniques are used in this research: multiple regression analysis and conjoint analysis. In order to investigate the stated hypotheses a questionnaire was sent asking respondents several survey questions (multiple choice and rating) that measured the individual factor constructs, suitable for the regression analysis. In addition, the questionnaire consisted of a choice-based conjoint section in which 16 choice tasks were presented to the respondents. Each choice task consisted of three different electric car profiles and respondents were asked to choose the profile that had their highest preference. A dual-response none option was also included after each choice task.

3.1.1 Multiple regression analysis.

According to Malhotra (2007) multiple regression analysis is ideally suited to determine whether a relationship exists between variables (i.e. if independent variables significantly explain variation in the dependent variable), what the strength of the relationship is (i.e. amount of variation explained) and what the structure or form of the relationship is. In addition, multiple regression is able predict values of dependent variables and control for other independent variables when evaluating the contribution of specific variable or set of variables. Further regression analysis issues will be discussed in section 4.2.

3.1.2 Conjoint analysis

Conjoint analysis assumes that any set of objects (e.g. product, brand, company) or concepts (e.g. images, benefits, positioning) is evaluated as a bundle of separate attributes (Hair et al., 2010). Conjoint analysis is different from other multivariate techniques in that (1) it is decompositional of nature, (2) it has a different specification of the variate, (3) estimates can be made at an individual level and (4) the relationship between dependent and independent variables can be flexible (Hair et al., 2010).

There are several types of conjoint analysis methods with each their advantages and disadvantages (see appendix A for overview). Due to the realistic nature, the possibility of fractional factorial design and the fact that fewer profiles are needed, a full profile method was used. In a full profile conjoint analysis questionnaire, respondents are asked to evaluate realistic product profiles (described by several levels of attributes) and to choose the product they would buy (i.e. consumption behavior).

(26)

analysis can be conducted on an individual and an aggregate level, and the interaction effects can be measured (see table 2 for conjoint methodology overview).

Characteristic Traditional conjoint Adaptive/Hybrid conjoint Choice-based conjoint

Upper attributes limit 9 30 6

Level of analysis Individual Individual Aggregate or Individual

Model form Additive Additive Additive + interaction

Choice task Evaluating full-profile stimuli one at a time

Rating stimuli containing subsets of attributes

Choice between sets of stimuli

Data collection format Any format Generally computer-based Any format Table 2: Comparison of conjoint methodology (adopted from Hair et al., 2010, p. 278)

It is important to note that conjoint analysis investigates consumer preference and not actual consumption behavior, which is the key concept in the conceptual model (presented in section 2.6). The technique is nonetheless very well suited since there is currently a deficit of purchase data of high involvement green products. Consumer preference is a good proxy for actual consumption behavior, since it is only logic that a higher preference for a certain product leads to a higher chance of consumption (Folkers, 1988). In addition, the dual-response none option, asking respondents if they would actually buy the selected product, indicates purchase intention and therefore consumption behavior (Ajzen, 1991).

The choice sets in this study are generated using Sawtooth Software (Sawtooth SSI web). Using fractional factorial design, one Conjoint Paper en Pencil Interview was created. Because of money and location restrictions and with regard to the expected amount of respondents, this is the preferred research methodology. It was essential to identify the key decision criteria that are communicable and actionable measures, and could be translated into attributes with distinct levels (Hair et al., 2010). The key decision criteria are presented in section 3.2.

In addition, the Balanced Overlap Method was used to create the profiles. The Balanced Overlap Method is a middling position between the random and the complete enumeration strategies. With regard to the main effects, the method is comparable to the Complete Enumeration and Shortcut methods (in terms of efficiency), but the method is much better in measuring the estimates of interaction (Sawtooth Software, 2008). No prohibitions and conditional relationships were included and no concept sorting was applied, which infers full randomization. The presented attribute order per profile was not randomized, but placed in a logical order. This was also done to make the choice task more convenient for respondents in order to ensure full response.

Sixteen dual-response choice tasks (i.e. 48 profiles) were created, of which three served as hold-out sets. Generally, twelve to eighteen choice tasks are recommended (Hair et al., 2010; Sawtooth Software, 2008). A dual-response none option was included in order to ensure retention of information (Sawtooth Software, 2005). In a test version with an integrated ‘none option’, the ‘none option’ was selected in about half of the cases, which severely limited the information that would be obtained.

(27)

2005). The D-efficiency value (Kuhfeld et al., 1994) summarizes how precisely this design can estimate all the parameters of interest with respect to another design.

The outcomes of the design test showed that efficiency values meet requirements. In addition, standard error values are roughly equivalent, no larger than 0,05 for all effects (Sawtooth Software, 2005), and standard errors are lower for the dual-response design. Moreover, D-efficiency scores indicate that a design with an included ‘none option’ would be 14% less efficient, supporting previous design decisions.

3.2 Measurements of the variables

As discussed in chapter 2, the influence of individual factors and product factors on green consumption behavior will be measured. In order to measure the relationship between the independent variables and green consumption behavior of a high involvement product, the electric car was selected. More specifically, the research will focus on consumer preference for battery electric (BEV) and plug-in hybrid electric (PHEV) cars on the Dutch market (further background is provided in subsection 3.2.2). Electric cars can be seen as typical green products (reducing the carbon footprint) and are clearly marketed as being environmentally friendly. Furthermore, electric cars are high involvement products since they require a significant financial investment. In addition, car choice is related to consumer values and relates to status and social identity. Cars are thus ‘consumed’ in a social context which stimulates higher involvement (Corfman, 1991; Janssen and Jager, 2002). The measurements for the variables therefore relate to the purchase (i.e. consumption) of an electric car.

The variables that were identified as individual factors (in section 2.3) are consumer attitude, habits and personal capabilities. The measurements for these individual factors will be discussed in subsection 3.2.1. The product factors that were identified (in section 2.4) are price, performance, impact on the environment and availability. In subsection 3.2.2 to 3.2.6 these variables will be translated towards electric car attributes, which will be the basis for the choice-based conjoint analysis.

3.2.1 Individual factors

(28)

Individual factors and measurement levels Supporting literature Consumer attitude

- Self-image effects

- Seeking green product information - Social value

- Own emission contribution

Oliver and Lee, 2011; Stern, 2000b; Staughan and Roberts, 1999; Roberts, 1996; Jackson 2005

Car or transportation usage (i.e. habits) Car ownership

- Private or lease - Amount of cars

- Type of car (most frequently used) - Energy Label current car

Car usage

- in km during week and weekend - % of use for (driving to) work

- Use of public transportation in relation to car use In the case of no car ownership

- Preferred means of transportation - km of travel during week and weekend - % transportation use for work

Bamberg et al., 2003; Gärling et al., 2001; Thøgersen, 2006; Verplanken et al., 1998; Aarts et al., 1998; Jackson 2005 Socio-demographic characteristics - Gender - Age - Education - Occupation - Income

- Household size (and composition) - Province (in the Netherlands) - Size of town or city

- Driver’s license ownership

Potoglou and Kanroglou, 2007; Segal, 1995; J.D. Power and Associates, 2008; Ong and Hasselhoff, 2005; Gilg et al., 2005; Staughan and Roberts, 1999

Table 3: Overview of measurement levels of the individual factors

3.2.2 Introduction into the product factors

(29)

European market General Motors introduced the Opel Ampera in the second quarter of 2011, which is technically identical to the Chevrolet Volt (Auto-press, 2009).

Viewing these developments, a clear distinction is made between BEVs and PHEVs in this research. This distinction is conceptualized by using the main differences between both types of cars, namely impact on the environment and driving range.

Picture 1: BEV 2 Picture 2: PHEV 2

A study by Milieu Centraal (De Hoop, 2009), conducted via a large scale digital panel research in the Netherlands, is ideally suited as a preliminary research for the investigation of other key decision criteria when buying a car in general. The research identified purchase price, safety, comfort, running costs and fuel consumption (i.e. impact on the environment) as being the top five most important criteria when looking for a new car. In addition, engine performance appeared to be an important factor when buying a car.

Car safety regulations in the Netherlands are very strict and car manufacturers have to make sure they comply with these regulations, especially in a category that has to earn consumer trust. It can therefore be assumed that safety is similar (i.e. 4 or 5 star) for all new models (Euro NCAP, 2011). In addition, the concept of comfort is found to be too abstract for the scope of this research and is already incorporated in other attributes (e.g. driving range, recharge possibilities). Running costs and fuel consumption (i.e. environmental impact) are linked to the type of car (i.e. battery electric and plug-in hybrid electric) and therefore are taken into account indirectly.

Based on the research of De Hoop (2009) and other findings in literature, the following attributes are included in the analysis (see table 4): environmental impact, price, driving range, battery switch (see subsection 3.2.5 for background), engine power and fast charging at home (see subsection 2.3.6). Further literature background for the choice of attributes and the attribute levels will be provided in the subsections below.

2

(30)

Attributes Attribute levels3

1. Environmental impact BEV zero tailpipe CO2 emission

BEV 75% less CO2 emission than ICE

PHEV 50% less CO2 emission than ICE

PHEV 25% less CO2 emission than ICE

2. Price € 15.000

€ 30.000 € 45.000 3. Performance – Driving range 150 km

350 km 550 km

4. Performance – Recharge time With battery switch Without battery switch 5. Performance – Engine power 100 HP

150 HP 200 HP

6. Availability – Fast charge Fast charge at home No fast charge at home

Table 4: Attributes and attribute levels for choice-based conjoint analysis

3.2.3 Product factor - Environmental impact

Driving an electric car reduces the carbon footprint (Eggers and Eggers, 2011). Although in the past two decades internal combustion engines have become cleaner (see appendix C, figure A and B), BEVs and PHEVs have even less direct impact on the environment. In addition, they do not depend (solely) on fossil fuels, of which the global reserves are limited.

Environmental impact is measured in CO2 emission values, since this is also the base for the

well-known Energy Label system (used for cars, washing machines and housing). In order to determine realistic emission values, a thorough analysis was performed using the extensive car database of Autotrack.nl and several other online sources (see appendix D). Appendix D shows that, when driving on electricity, the CO2 tailpipe emission of a BEV is zero and the CO2 emission of a PHEV is about half of

the emission of comparable gasoline cars (in the compact car and midsize car segment). In the case of PHEVs, emission level heavily depends on usage. Since a PHEV can drive about 50 kilometers on the all-electric engine, tailpipe emission is only an issue after these 50 kilometers (Walter, 2010; Pooley, 2010; Vliet et al., 2011).

In addition to environmental impact of the emission of the car itself, the electricity needed for the electric engine can either be generated in a conventional coal burning installation or the electricity can be green electricity (e.g. wind or solar power). These forms of electricity production have different impact on the environment (Brey et al., 2007). To account for this, a separation of two extra levels of emission values is included, both having 25% more CO2 emission.

3

Is it time to drive green? Determinants of green consumption behavior for a high involvement innovation, a choice-based conjoint analysis

Is it time to drive green?

Determinants of green consumption behavior for a high involvement innovation,

a choice-based conjoint analysis

Is it time to drive green?

Management summary

Preface

Table of contents

1. Introduction

2. Theoretical framework

“Green, a symbol of life, good health, and vigor, a color that reminds us of hope.”

3. Research design