Abstract
This present study examines whether highly environmentally involved consumers respond better to truthful green marketing advertisements than those that use greenwashing tactics for
clothing advertisements. The greenwashing techniques used in this experiment are fibbing, vagueness, irrelevance, and false labels. In line with existing literature, a mediation analysis
showed that both the green marketing and greenwashing advertisement positively affect brand attitudes by shaping advertisement attitudes. Contrary to expectations, consumers with
higher levels of green involvement are found to respond more negatively to brands that use green marketing advertisements. Moreover, greenwashing advertisements, on the other hand,
are found to appeal to both low and highly green involved consumers.
Keywords: green advertising, greenwashing, brand attitude, environmental involvement, effectiveness, fashion industry
Joann Naomi Noslin 13286323
Supervisor: Dr. Ayşe Gul Mermer PhD Second supervisor: Prof. Dr. Arthur Schramm
University of Amsterdam Amsterdam School of Economics
MSc Economics, Behavioral Economics, and Game Theory August 15, 2021
Green involvement as a mechanism against
greenwashing in the fashion industry
ii Statement of originality
This document is written by Student Joann Naomi Noslin, who declares to take full
responsibility for the contents of this document. I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it. UvA Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
iii Table of Contents
Abstract ... i
Statement of originality ... ii
I Introduction ... 1
II Literature review ... 5
2.1 Green marketing ... 5
2.2 Greenwashing ... 6
2.2.1 The impact of greenwashing ... 8
2.2.2 Greenwashing in the fashion industry ... 9
2.2.3 Consequences of greenwashing ... 10
2.3 Environmental involvement ... 11
III Methodology and Models ... 14
3.1 Research design ... 14
3.2 Experimental manipulation ... 15
3.3 Measures ... 17
3.4 Theoretical model ... 18
3.4 Sample and procedure ... 20
IV Results ... 21
4.1 Attitude toward the advertisement ... 21
4.2 Attitude toward brand ... 22
4.3 Structural equation model ... 22
4.3.1 Environmental involvement ... 23
4.4 Interaction effects ... 26
4.4 Robustness checks ... 30
V Discussion ... 31
5.1 Summary of main findings ... 31
5.2 Practical implications ... 33
5.2.1 Managerial implications ... 34
5.2.2 Governmental policy implications ... 35
5.3 Limitations and recommendations for future research ... 36
VI Conclusion ... 38
Reference list ... 39
Appendices ... 46
Appendix 1. Stimuli ... 46
Appendix 2. Questionnaire ... 47
Appendix 3. Descriptive Statistics ... 48
iv
Appendix 4. ANOVA tests ... 50
Appendix 5. Regressions and SEM ... 52
Appendix 6. Interaction effects regressions ... 57
Appendix 7. Robustness ... 65
1 I Introduction
A growing number of consumers are becoming more environmentally conscious in their consumption decisions. More importantly, a report by Nielsen (2015) found that 66% of consumers are willing to pay more for environmentally friendly products. As a response, the consumer market for green products and services is expected to continue to grow. As a result, an increasing number of companies use environmental appeals in their advertisements to entice these consumers (Easterling, Kenworthy, and Nemzoff 1996).
However, national consumer protection authorities in the European Union found that in 42% of the cases where green claims were used on the web, the claims were exaggerated, false, or deceptive and could therefore potentially qualify as unfair commercial practices under the Unfair Commercial Practices Directive (European Commission 2021).
Attempts to mislead consumers by making them believe products are more environmentally friendly than they are is called greenwashing and is most prevalent in consumer goods markets (Delmas and Burbano 2011). Companies do this through various marketing tools by strategically highlighting the product's positive environmental effects while occluding the negative aspects.
Greenwashing commonly occurs in consumer market industries; even though consumers have become more skeptical nowadays, companies can still significantly profit from its usage (Pimonenko et al. 2020). Although much research has been done on
greenwashing in the last couple of years, as greenwashing practices increase in prevalence, the topic is still relevant and demands more detailed research (Matthes, Wonneberger, and Schmuck 2014).
The Intergovernmental Panel on Climate Change (IPCC) estimated that the fashion industry produces approximately 10% of global carbon dioxide emissions annually and uses around 1.5 trillion liters of water each year (Davis 2020; McKinsey 2020). On this basis, this study will take a closer look at greenwashing in the fashion industry. One of the main
developments in the industry that is deemed to be aggravating the problems is high street fashion, also known as fast fashion, relatively cheap clothing that is cast aside as trends change rapidly (Davis 2020).
At present, the number of major fashion brands engaging in green marketing is growing (H.-S. Kim and Hall 2015). Not surprisingly, the fashion industry is experiencing a
2
concurrent rise in greenwashing practices. Fashion brands commonly utilize green imagery, labels, and names to appear green and ethical to target the conscious consumer (Henninger, Alevizou, and Oates 2016). As there is a constant trade-off between sustainability, trendiness, and price in the fashion industry, the current move towards a more sustainable economy has significant implications (Jung and Jin 2016). Various high street fashion brands have recently begun to release lines targeted at the more conscious consumers, for example, H&M's
conscious line.
However, in its essence, high street fashion creates clothing that is meant to be replaced within a short period; therefore, some may argue that these companies can never be eco-friendly as their business model is inherently unsustainable (Davis 2020). Hence, some argue that sustainable fashion, in general, is an oxymoron as fashion signifies something trendy (Clark 2015). Therefore, in this paper, sustainable fashion represents clothing designed, manufactured, and distributed in an environmentally friendly manner.
The importance of greenwashing is apparent as prior research found that the
increasing level of greenwashing can have profound adverse effects on consumer confidence in green marketing (Chang 2011; Naderer, Schmuck, and Matthes 2017). As a result, NGOs and government agencies try to limit the benefits of greenwashing by educating consumers, suing companies, creating ecological labels, and participating in the creation of stronger laws (Sun and Zhang 2019).
Nonetheless, the question arises as to whether those more involved with
environmental issues respond more negatively to greenwashing. For example, it was often believed that those high in green involvement are more skeptical towards green
advertisements (Urbański and Haque 2020). However, some studies both found that green consumers are less skeptical about advertisements than others and respond more strongly to emotional attributes in green advertisements, a common tactic in greenwashing (Matthes and Wonneberger 2014; Matthes, Wonneberger, and Schmuck 2014). Therefore, it may not be the case that higher levels of green involvement and more knowledge about environmental issues protect consumers from the greenwashing used by companies.
For companies, it is interesting to determine whether consumers with higher levels of green involvement respond differently to green advertisements than those low in green involvement. This can be used to target consumers with different levels of green involvement and optimize the profitability of their campaigns. Therefore, this study will examine how consumers respond to a misleading advertisement that uses common greenwashing tactics. In
3
addition, their responses will be compared to those for a green advertisement that does not employ any greenwashing tactics.
The purpose of this study will thus be to determine whether green involvement can help protect consumers against green advertisements that use standard greenwashing
practices. The first objective of this study is to provide empirical evidence for the suggestion that both green marketing and greenwashing appeals improve people's attitudes toward a brand. Secondly, the effectiveness of these appeals might depend on the consumer's level of green involvement. Understanding the effectiveness of these appeals and whether the level of green involvement impacts their effectiveness may provide policymakers with valuable insights on how to disincentivize greenwashing. Moreover, it may aid marketers when trying to target customers with higher levels of green involvement.
Consequently, the following research question arises:
Do consumers with higher levels of green involvement respond better to green marketing appeals than greenwashing?
As the market for more environmentally friendly products continues to grow and greenwashing becomes more common, many research papers investigate the different aspects. Prior research has, for instance, addressed that greenwashing can have an indirect negative effect on green purchase intention and consumers' perception of green marketing as they become unable to trust green advertisements (Hamann & Kapelus 2004; Naderer, Schmuck, and Matthes 2017; Polonsky et al. 2010; Tarabieh 2021). Many of these studies, therefore, recommend that companies should not employ greenwashing methods. However, based on the high number of instances in which campaigns can be seen to engage in some sort of greenwashing practices, its usage appears to remain widespread (TerraChoice (Underwriters Laboratories) 2010). Hence, it is interesting to determine how consumers respond to greenwashing compared to green marketing advertisements because little is still known about whether greenwashing, or green marketing effectiveness depends on consumers' level of environmental involvement (Matthes, Wonneberger, and Schmuck 2014).
This study intends to add to the literature in the following ways. Prior studies often explore how consumers feel about greenwashing or are usually specific to an industry like the food and beverage industry (Chang 2011; Nyilasy, Gangadharbatla, and Paladino 2014;
4
Schmuck, Matthes, and Naderer 2018). However, little research has been conducted regarding greenwashing in the fashion industry (Henninger, Alevizou, and Oates 2016).
Moreover, the effectiveness of underlying processes for greenwashing campaigns remains understudied. While previous studies have focused on the efficacy of either
emotional or functional appeal concerning levels of green involvement, this study will instead investigate greenwashing advertisements by trying to determine whether greenwashing advertisements' effectiveness is subject to someone's environmental involvement (Matthes, Wonneberger, and Schmuck 2014). A study on greenwashing and green involvement found that consumers high in green involvement, despite being skeptical, still incorrectly identify greenwashed products as sustainable. However, to test this, this study asked participants to what extent they trust labels that say "recycled," "No CFCs," or "Green-certified." It then determined how people respond to vague, irrelevant, and false labels (Urbański and Haque 2020).
As this does not mimic realistic advertisements, this thesis will add to the present literature by establishing if greater awareness of purchases' environmental impacts leads to lower vulnerability to common greenwashing practices. Suppose it is the case that consumers respond better to factual green marketing statements. In that case, this might help companies become more truthful about their ecological impact if the share of those high in green
involvement grows. However, if the opposite is found to be true, this can indicate that
governments and NGOs should focus more on creating universal standards for environmental certifications.
Data is collected through an online experiment to determine whether a higher level of green involvement makes consumers respond better to a green advertisement without any greenwashing tactics (N = 247). The experiment has a between-subjects design. Participants are shown either a control advertisement, a greenwashing advertisement, or a green
marketing advertisement. Both advertisements are based on advertising strategies and slogans used by major fashion brands. The greenwashing advertisement uses four common
greenwashing tactics: irrelevance, false labels, fibbing, and vagueness. On the other hand, the green marketing advertisement is based on claims made by fashion brands and companies that focus on shedding light on greenwashing.
The data is then analyzed using Stata, and a structural equation model was built to run robust regressions and conduct a mediation analysis. The results of this study indicate that both green marketing and greenwashing appeals improve brand evaluation by shaping the
5
attitudes toward the advertisement. Interestingly, a negative interaction between the green marketing advertisement and higher levels of green involvement is found; no interaction effects are found for the greenwashing advertisement. People with higher levels of
environmental involvement are thus observed to view the brand that uses green marketing appeals in their advertisements more negatively.
The organization of this thesis is as follows—Section 2 reviews relevant literature on green marketing, greenwashing, and green involvement. The four hypotheses are also introduced in this section. Section 3 presents the quasi-experimental design and
manipulations. This is followed by Section 4 shows the results. In Section 5, the practical implications of the findings are discussed. Lastly, concluding remarks are made.
II Literature review
In this section, an overview of related papers is given. First, a clear definition of green marketing is given. Afterward, the most critical literature about greenwashing is covered.
Additionally, a definition for environmental involvement is given. Lastly, the hypotheses will be presented accordingly.
2.1 Green marketing
Consumption habits are widely seen as contributing to environmental degradation, global warming, and the depletion of non-renewable resources (Grimmer and Woolley 2014;
Hartmann and Apaolaza Ibáñez 2006). Therefore, more consumers are becoming willing to pay a premium for green products and services, which have a non-negative impact on the environment. Consequently, the market for green products and services is growing rapidly;
companies have realized the importance of distinguishing their products as having a positive impact on the environment (Grimmer and Woolley 2014; Naderer, Schmuck, and Matthes 2017).
To develop a positive image as a green brand in the eyes of environmentally
conscious customers, companies started utilizing green marketing (Chang, 2011; Chen & Wu, 2015; Chen & Chang, 2013; Chen, Tien, Lee, & Tsai, 2016; Wu & Lin, 2016). Literature provides a wide range of definitions of the concept of green marketing. It is commonly defined as creating and advertising products or services based on their (non-negative) environmental impact and has become essential to target environmentally conscious
6
consumers (Polonsky 1994). Green advertising is expressed by three criteria, of which at least one has to be fulfilled: (1) Explicitly or implicitly communicate the (positive) relationship between the product and the environment, (2) Emphasize a green lifestyle, and (3)
communicates a company's responsibility to the environment (Banerjee, Gulas, and Iyer 2013).
Important to note is that consumer pressure is one of the critical drivers for firms to become more environmentally friendly. Therefore green marketing is more common in consumer product markets, such as the fashion industry (Delmas and Burbano 2011). There are various techniques for companies to use green marketing communications to persuade consumers. For example, Hartmann and Ibáñez (2006) claim that companies can increase consumers' perceptions of personal benefits by adding emotional value to goods as
consumers primarily want to feel good after a purchase. These benefits can either be tangible or intangible. Intangible value focuses on the idea that consumers feel good about themselves after purchasing green products or services; they obtain a so-called warm glow (Grimmer and Woolley 2014).
With this information, the first hypothesis is formulated
H1: Green marketing advertisements have a positive impact on brand attitudes by shaping advertisement attitudes.
Nonetheless, as consumers' interest in sustainable products grew, the words commonly used in green marketing, such as "eco-friendly," "biodegradable," and "more sustainable," became popular even among firms that do not necessarily behave sustainably.
However, sometimes, these green statements or advertisements are not sincere or truthful, becoming misleading. This is called greenwashing.
2.2 Greenwashing
The act of misleading consumers regarding the environmental practices of a company or the environmental benefits of a product or service is defined as greenwashing (Delmas and Burbano 2011; TerraChoice (Underwriters Laboratories) 2010). To bring attention to this phenomenon, the term greenwashing was coined by Jay Westerveld in 1986, specifically to inform people about the exaggerated claims the hotel industry used about the effect of reusing towels on the environment (Koh et al. 2010; Sun and Zhang 2019). As nudges inherently focus on altering the choice environment without restricting any decisions, they will often
7
mislead consumers regarding the environmental benefits of a company (Thaler and Sunstein 2001). Green nudges may thus not be a harmless way to entice consumers to behave more sustainably as previously perceived.
The environmental performances of firms range from bad to good; these firms can be broadly divided into two groups poor environmental performers (known as "brown" firms) or good environmental performers. Further categorization is between those firms that
communicate about having a positive environmental performance, vocal firms, and those who do not share about their environmental impact at all, silent green firms. A greenwashing firm thus engages in two behaviors alongside each other; namely, their communication can be described as vocal, but their environmental performance is bad (Delmas and Burbano 2011).
It is thus a way of masking a firm's weak environmental performance.
Greenwashing can occur at either the firm level, by overstating a company's environmental practices, or at the product level, by overstating a product's environmental benefits (Delmas and Burbano 2011). There are essentially two types of greenwashing—
companies claiming that new methods to cut costs are done from an eco-friendly perspective.
The second form of greenwashing exists of companies who mislead consumers about the ecological impact of a product by, for instance, using irrelevant claims or environmental imageries. This study will focus on the second form of greenwashing, which is most relevant to the fashion industry.
To determine whether and why specific green marketing claims fall into the
greenwashing territory TerraChoice Environmental Marketing (2009) created the seven sins of greenwashing shown in figure 1. Based on the sins shown in figure 1, a study by
TerraChoice (2009) concluded that 98% of the green product marketing they researched in North America used at least one of these sins of greenwashing. Therefore, they are classified as greenwashing.
8
Figure 1. The seven sins of greenwashing by Terrachoice (2009)
Firms can greenwash with ease, mainly due to a lack of consumer awareness and inadequate product information. Another key driver is the current state of lax regulations in most of the world (Delmas and Burbano 2011). Terms frequently used in green (fashion) advertisements are 'eco-friendly,' 'sustainable,' 'biodegradable,' 'recyclable,' 'non-toxic,' and 'plant-based, as they are all broadly defined and therefore its authentic meaning, will likely be misunderstood by the target customer (van Niekerk and Conradie 2020). Thus, to enjoy the benefits of the sustainability trend, companies may take advantage of these vague definitions.
2.2.1 The impact of greenwashing
The impact of greenwashing on the societal level is apparent as the urge to buy ecologically conscious products often stems from greater public awareness of the current environmental crisis. Therefore, persuading consumers by using non-factual, inaccurate, or false information in green advertising diminishes the utility of green advertising in guiding consumers who want to make eco-conscious purchases (Naderer, Schmuck, and Matthes 2017). Thus, greenwashing possibly affects the way society deals with the ecological crisis.
The most important way greenwashing harms society is by possibly causing consumers to distrust all green advertising. One study found that greenwashing potentially leads to more negative than positive outcomes, hurting both purchase intention and perceived
9
lower integrity (Jong, Harkink, and Barth 2017). Additionally, Tarabieh (2021) found that greenwashing, specifically false environmental claims, has an indirect negative effect on green purchase intention. This study further enhanced the idea that all green marketing activity would be harmed by greenwashing as consumers ultimately become unable to trust firms' green advertising (Chen and Chang 2013; Hamann 2004).
However, in these studies, the green advertisements were always perceived as greenwashing, while this may not be as noticeable in more realistic settings. Hence, the advertisement in this study will mimic an actual advertisement used by major brands.
Consequently, greenwashing may not be spotted as easily. According to Akturan (2018), if a brand is simply associated with being green, this could potentially lead to a more positive evaluation of the brand. Moreover, Hartmann & Apaolaza-Ibáñez (2015) argue that green advertising that uses informational claims can lead to a higher brand attitude.
Therefore, the second hypothesis is as follows:
H2: Greenwashing advertisements have a positive impact on brand attitudes by shaping advertisement attitudes
2.2.2 Greenwashing in the fashion industry
Greenwashing practices are rising rapidly in the fashion industry, especially among high-street fashion brands (Henninger, Alevizou, and Oates 2016; H.-S. Kim and Hall 2015).
Common forms are releasing a conscious product line while continuing to focus on mainly selling unsustainably made products. Other companies use marketing statements with a message like "buy this new product to help the environment," an argument that is contradictory by nature. Lastly, many companies simply overstate their ethical or environmental efforts (Henninger, Alevizou, and Oates 2016; Madeleine Hill 2021). It remains unclear, however, whether consumers notice that these efforts are simply greenwashing.
One of the incentives for companies to use greenwashing is to make prospective customers feel good about their potential purchases. Oliver Balch (2013) pointed out an example of a big brand using insubstantial feel-good claims, such as H&M's recycling
scheme. He stated that "the promise of closed-loop textile recycling currently remains exactly that: a promise," as there is a lack of publicly disclosed results of the campaign (Blach 2013;
Gould 2017).
10
To determine which fashion brands are sustainable the website, Good on You, a company that focuses on educating customers about the impact of many fashion brands on the planet, people, and animals, is popularly used by consumers. This website educates consumers on the practice of greenwashing and therefore tries to use detailed descriptions as to why a brand is sustainable (Madeleine Hill 2021). According to Good on You, H&M and Uniqlo, well-known high-street fashion brands, are on their way to becoming sustainable brands, as they currently employ a medium proportion of eco-friendly materials and set science-based targets (Hymann 2020; Lara Robertson 2020).
On the other hand, Inditex, the world's largest fashion retailer, is currently not doing enough to cause less harm to the environment, according to Good on You, as there is no evidence that its brands, like Zara, minimize textile waste when manufacturing its products (Dockrill 2020). However, all three of these companies have launched "sustainable" clothing lines and campaigns to attract more conscious consumers. Nonetheless, as previously
explained, the business model of high street fashion brands is focused on creating a throw- away culture. Consequently, the business model of these companies may be inherently unsustainable.
2.2.3 Consequences of greenwashing
As mentioned previously, the impact of greenwashing on society can be disastrous, as it affects how the current environmental crisis is being handled (Dahl 2010; Naderer,
Schmuck, and Matthes 2017). Another lesser-known consequence of the current rate of greenwashing is that some firms will stay silent about their positive environmental impact even if this is backed by a third-party certification (Carlos and Lewis 2018). IKEA, for example, is a leader in sourcing Forest Stewardship Council (FSC) certified lumber but rarely communicates these efforts to consumers.
Fortunately, corporations are increasingly being held accountable for corporate social irresponsibility, and some have even faced lawsuits for engaging in greenwashing (Delmas and Burbano 2011; Stäbler and Fischer 2020). For example, Windex settled a class action suit for deceptive statements regarding the environmental safety and soundness of the cleaning product (Lane 2010). To which extent an event of corporate social irresponsibility damages a company is primarily boosted by whether it is covered by the media, as by construction, unethical behavior has no consequences unless it is publicized (Stäbler and Fischer 2020).
Moreover, Stäbler and Fischer (2020) found that positive brand news can crowd out media
11
coverage of unethical behavior. For these reasons, companies may still practice greenwashing as the reward may still be higher than the risk. Moreover, as most fashion industry firms simply use vague slogans, environmental icons as labels, and irrelevant imagery, this is often not strictly illegal; hence, lawsuits are less common.
Because of the negative impact of greenwashing on the environmental crisis, various NGOs and companies are focused on educating consumers through their social media
platforms and websites on the greenwashing practices used by firms to minimize the benefits of greenwashing (Two sides 2018; Urbański and Haque 2020). However, the effects of this form of education depend strongly on the idea that those who are higher in green
involvement, in other words, those who know more about the environment, can notice the greenwashing practices employed by companies. Little research has, however, investigated this topic. Furthermore, the popularity of greenwashing techniques among companies could show the effectiveness of attracting customers with these practices.
Consumers who have no concerns for the environment at all should not be affected by greenwashing, as the environmental impact of a product should not influence their opinion of a brand (Delmas and Burbano 2011; Urbański and Haque 2020). Hence, it might indicate that people who are somewhat interested in the impact of their purchase on the environment are the people these greenwashing advertisements are aimed at. All in all, it would be interesting to see whether people with higher levels of green involvement would respond better to green marketing, companies showing information about what they are truly doing for the
environment, compared to using greenwashing claims.
For the remainder of this paper, green marketing is defined as the practice of advertising products or services based on their actual environmental impact without
committing any of the seven sins as designed by Terrachoice (2009). Greenwashing will be used in the context of advertising products and services to mislead consumers regarding the environmental practices by using at least one of the seven sins of greenwashing. Moreover, green, environmental, and environmentally friendly are used as synonyms.
2.3 Environmental involvement
A wealth of research suggests that consumers differ in how well they respond to advertisements. One key determinant of how they react to persuasive appeals is consumers'
12
level of involvement. Involvement denotes the degree to which one perceives an attribute object as personally relevant (Matthes, Wonneberger, and Schmuck 2014; Petty and
Cacioppo 1990). Petty and Cacioppo (1990) found that the quality of the presented arguments becomes a more relevant determinant of persuasion as the level of involvement increases. In terms of green advertising, consumers high in environmental involvement are thus expected to mainly be persuaded by arguments that explain why a product or service has a positive environmental impact. In contrast, those low in environmental involvement may rely more on heuristic cues like emotions to form an opinion, used by various forms of greenwashing.
Based on this, hypothesis three is as follows.
H3: The effects of the greenwashing advertisement will be more substantial for consumers low in green involvement.
Most studies define consumers high in green involvement as those that are highly concerned with the environment. However, there is a minimum of three key
conceptualizations related to environmental involvement, (1) environmental concern, (2) attitudes towards green products and services, and (3) green purchase behavior (Matthes, Wonneberger, and Schmuck 2014). These three sides of environmental involvement are deeply intertwined but can still have different antecedents and outcomes. Therefore all three should be included when studying green involvement.
Environmental concern is associated with a consumer's awareness of the environment and its perceived necessity to protect it (Matthes, Wonneberger, and Schmuck 2014). It has been found to play an important role in green purchasing decisions (Grimmer and Bingham 2013; Grimmer and Woolley 2014). Moreover, scholars treat consumers with a positive attitude towards green products as highly involved (Chan 2001). Chang (2011) described these positive attitudes as relating to advantages, favorability, and quality of green products or services. Lastly, green purchase behavior, both actual behavior and behavioral intentions, is an indicator of green involvement (Schwepker and Cornwell 1991; Stone, Barnes, and Montgomery 1995).
Furthermore, previous studies have shown that environmental claims have a more substantial effect on consumers with higher levels of green involvement. For instance, research by Grimmer & Bingham (2013) found that environmental claims based on a company's environmental performance cause more substantial positive effects on purchase intentions for consumers high in green involvement.
13
More importantly, a study that looked at the effects of environmental messages in fashion advertisements found that subjects concerned about the environment responded more positively to advertisements with an environmental message. Conversely, subjects without such a concern responded more positively to fashion advertisements without an
environmental message. Therefore, the findings demonstrated the importance of consumers' environmental concerns as a mediator to responses to fashion advertisements (Y. K. Kim, Forney, and Arnold 1997). However, this study did not look at the link between
environmental claims in fashion advertisements and consumers' level of environmental involvement, nor did it look at the difference between misleading environmental messages and truthful messages.
It is essential to do so as, in contrast, findings by Matthes, Wonneberger, and Schmuck (2014) find that consumers high in green involvement do not devalue emotional appeals used in advertisements for a laundry detergent, nor do they respond better to advertisements that give arguments explaining the product's positive environmental
attributes. Conversely, they also find that consumers who are high in green purchase behavior or have positive attitudes towards green products and services respond better to functional green advertising for this laundry detergent. Therefore, looking at the effects of all
determinants of environmental involvement in other industries is critical.
Moreover, consumers are frequently mentioned to be skeptical of environmental claims, especially consumers high in green involvement. These claims are often
misunderstood, and consumers often evaluate environmental advertising as vague or even misleading. However, Matthes and Wonneberg (2014) found that consumers high in green involvement are less skeptical of green advertisements. These studies also demonstrate the importance of distinguishing between these three axes of environmental involvement.
All in all, based on the model by Petty and Cacioppo (1990) and the findings by Grimmer & Bingham (2013), the green marketing advertisements will lead to a more positive brand evaluation for consumers high in green involvement.
This leads to the following hypothesis:
H4: The effects of the green marketing advertisement will be stronger for consumers high in green involvement.
14 III Methodology and Models
First, an overview of all the previously derived hypotheses will be given. Next, the research design and the basis of the experimental manipulations will be explained. Then, the
procedures and the questionnaire design will be discussed. Finally, a plan for analysis of the data will be presented.
The current study strives to establish a cause-and-effect relationship between different appeals used in green advertisements, participants’ environmental involvement, and attitude towards the advertisement or brand. Therefore, the conclusive research design should be causal. An experimental approach is frequently considered the most reputable and primary method to test causality, as no other method offers as much control over the research procedure (Malhotra, Nunan, and Birks 2017). Therefore, this method fits this study best.
Moreover, as this study focuses on what participants think about an advertisement or brand, this approach draws valid conclusions regarding the causal relationship with a satisfactory degree of confidence. An overview of the hypotheses is shown in Table 1.
Table 1. Summary of hypotheses
H1 Green marketing advertisements have a positive impact on brand attitudes by shaping advertisement attitudes.
H2 Greenwashing advertisements have a positive impact on brand attitudes by shaping advertisement attitudes.
H3 The effects of the greenwashing advertisement will be stronger for consumers low in green involvement.
H4 The effects of the green marketing advertisement will be stronger for consumers high in green involvement.
3.1 Research design
A quasi-experimental research design is employed for this study, with two independent variables: the various advertisements and environmental involvement, and two dependent variables, namely, brand and advertisement perception. A convenience sample of 237 participants is used where respondents were randomly assigned to either one of the three conditions. Therefore, the experimental design is a between-subjects design and can be considered a true experimental design (Malhotra, Nunan, and Birks 2017). The between
15
subjects’ design ensures that different participants are shown different advertisements. The study employed a 3 (advertisement without additional stimuli, typical greenwashing advertisement, and green marketing advertisement) X 3 (low, moderate, and high environmental involvement) between-subjects factorial design.
3.2 Experimental manipulation
Three separate advertisements are used. One represents an advertisement for the product that promoted no environmental benefits (the control group), a greenwashing advertisement that uses four greenwashing sins: fibbing, vagueness irrelevance, and false labels (experimental group 1). The explanation of these sins can be found in figure 1. Lastly, a green marketing advertisement, which uses factual non-exaggerated statements based on the information given on the Good on You website, does not employ any of the greenwashing sins (experimental group 2).
These four sins are chosen as they are commonly employed in consumer goods markets (Naderer, Schmuck, and Matthes 2017). Additionally, brands in the fashion industry seem to use this practice, for instance, H&M and their ‘Now made with more sustainable cotton’ campaign.
The brand used, Towkow, is a non-existing brand to ensure that participants have no bias towards the brand. The decision to use a non-existent brand is relevant in the fashion industry as brand names are a crucial determinant of purchasing decisions in this industry (Ahuja 2015). Moreover, doing so increases the experimental control to eliminate possible alternative explanations and guarantee that only the relevant manipulations influence the results (Boush and Loken 1991).
The background picture is from a gender-neutral campaign by Bethnals for basic clothing in the hopes that participants are not familiar with the photo, nor are they heavily biased by the style. The clothing in the picture could consequently be found in most high- street fashion stores, such as H&M and Uniqlo.
As stated before, various high-street fashion brands have been adding sustainable collections to attract more environmentally conscious customers. Many brands set targets to ensure that their clothing will be produced consciously within the next decade (Bowman and Mccammon 2019; Lara Robertson 2020). H&M, the second-largest clothing retailer globally,
16
currently uses green clothing tags that promise that at least 50% of the product is made of sustainable materials (Noto 2020). Subsequently, the choice is made to use the statement that more than half of the clothing was made in a sustainable fashion. However, how to market this statement differs for the two experimental groups. The experimental conditions vary in their visual and textual layout, but the slogans and arguments are based on real brands in both conditions.
For experimental group 1, three specific sins of greenwashing were used (1) vagueness, (2) fibbing, and (3) false labels, as many of the high-street fashion brands use these sins to market their clothing. A study by Baum (2012) revealed that the most frequently applied form of greenwashing was the presentation of vague claims (51.7%), and has been found to occur in the United States, India, and China (Naderer, Schmuck, and Matthes 2017).
Examples of vague words often used in the clothing industry are organic, recycled, and plant- based. In this study, organic plant-based cotton represents both a vague claim and an
irrelevant claim. As organic cotton is not necessarily better for the environment but is often believed to be, cotton is nearly always plant-based and is thus irrelevant to mention (Kooistra, Termorshuizen, and Pyburn 2006).
The sin of fibbing occurs when firms make environmental claims that are simply false. This is often represented by marketing statements in the fashion industry that encourage customers to buy new products to help the environment or no longer feel bad while buying your style. ASOS, for instance, advertises its sustainable line as follows: “Giving you the confidence to shop with both sustainability and style in mind” (ASOS n.d.). Based on this, the statement “No longer choose between fashion and protecting the environment” was chosen. This represents the ideology that buying, in this case, organic clothing, offsets any negative impact that buying new clothing has on the environment.
Lastly, the sin of false labels is frequently used by clothing companies by adding small icons, often leaves, hands, the earth, or hearts, sometimes combined with buzzwords that hold little legislative power. This is usually done to mimic the idea of third-party endorsements (Dahl 2010; UL 2020). For example, Segev et al. (2016) concluded that in all green advertisements across different industries that presented a logo, 67.7% of these logos were illegitimate. For these reasons, in this advertisement, icons of leaves and hands are used to portray an image that the products are eco-friendly, organic, and help save water.
Moreover, the company logo, a heart, is turned into a heart with arrows to signify the idea of being a circular brand.
17
For experimental group 2, the green marketing treatment, the statements are not exaggerated and are based on slogans of sustainable clothing brands. To determine which brands are sustainable the website, Good on You, which focuses on educating customers about the impact of many fashion brands on the planet, people, and animals, was used. The slogan “Buy better, wear longer” has been used by Levi’s in its most recent campaign to focus on sustainability (Levi Strauss & Co 2021). The idea of choosing quality products so that one needs to buy less clothing over time is fundamental to slow fashion (Jung and Jin 2016). Hence, slow fashion is key to creating a more sustainable clothing industry. As this slogan perfectly captures, it is used in the green marketing treatment.
To further explain why this clothing line is environmentally friendly, information from the website, Good on You is utilized. Whenever a brand has a sustainable supply chain and uses organic cotton, the website mentions that the company tries to limit the chemicals, water, and wastewater used in the production process.
Besides, Dahl (2010) mentions that third-party certifications are a great way for consumers to determine whether a company is truly sustainable. The Global Organic Textile Standard (GOTS) is the world’s most recognizable certificate for organic cotton (Global Organic Textile Standard n.d.). Additionally, brands such as H&M hold this certificate. For these reasons, this certification is used in the advertisement. As a small environmental icon is one of the sins of greenwashing, the brand logo was changed back into a heart instead of a heart with arrows. Overall, the explanation as to why this collection is green reads as follows,
“More than 50% of our collection is made from GOTS* cotton to limit the chemicals, water, and wastewater used in production”. This statement is based on the information given by Good on You (Good on you n.d.). All three experimental conditions can be found in appendix 1.
3.3 Measures
The full survey can be found in appendix 2. Five five-point semantic differential scales measure the two dependent variables, namely attitude towards the advertisement, (Cronbach’s α = 0.83; M = 3.65; SD=0.74), and attitude towards the brand, (Cronbach’s α = 0.90; M = 3.63; SD=0.77). As for the independent variables (1-“strongly disagree” and 5-
“strongly agree”), the study employs the three previously mentioned forms of environmental
18
involvement after the experimental stimulus. This is done to ensure consumers are not affected by their prior answers.
To determine whether someone is high in green involvement, questions related to three categories (1) environmental concern, (2) attitudes towards green products and services, and (3) green purchase behavior are used. Environmental concern contains three items based on Schuhwerk and Lefkoff-Hagius (1995), (Cronbach’s α = 0.76; M = 4.25; SD=0.72).
Attitude towards green products consists of four items based on Chang (2011), (Cronbach’s α
= 0.84; M = 4.15; SD=0.69). Third, green purchase behavior entails four items based on Kim
& Choi (2005), (Cronbach’s α = 0.86; M = 3.67; SD=0.95). Total green involvement (Cronbach’s α = 0.89; M = 4.00; SD=0.67) is then determined by all these questions and averaged to range from 1 to 5.
Next, like previous studies, the study uses two additional control variables (Matthes and Wonneberger 2014; Matthes, Wonneberger, and Schmuck 2014). First, negative sentiment towards marketing is added as the perceived effectiveness of advertisements
depends on a consumer’s sentiment towards marketing in general. Negative sentiment towards marketing consists of two items, (Cronbach’s α = 0.50; M = 3.03; SD=0.94) (Mohr et al 1998). Second, perceived consumer effectiveness (PCE) is added since it is crucial in green advertising as those who believe they cannot change anything about the environment are less likely to respond to information about the environmental impact (Ellen, Wiener, and Cobb-Walgren 1991; Y. Kim and Choi 2005). Two items based on Ellen et al. (1991)
measure perceived consumer effectiveness, and these are reverse coded to measure consumer effectiveness (Cronbach’s α = 0.66; M = 3.43; SD=1.23).
No items could be deleted to increase the Cronbach’s Alpha for any variable. Also, as only one control variable has a Cronbach Alpha of lower than 0.50, slightly increasing the number of items for this item may lead to an acceptable value in future research (Taber 2018). Lastly, control questions for age, gender, and education are added as well.
3.4 Theoretical model
Empirical evidence from previous studies indicates that besides the treatments, and environmental involvement, the previously mentioned control variables also influence the attitude toward advertisement and brand. Therefore, an ordinary least squares (OLS) regression will be performed to determine whether these effects are present for this data.
19
Moreover, as the treatments significantly influence both the attitude toward the advertisement and the brand, a logical next step is to analyze the role attitude toward the advertisement plays in this relationship. To determine the extent to which a person’s attitude toward advertisements mediates the treatments, structural equation modeling (SEM) is used (Hayes 2017). This model uncovers the links between observed variables and latent variables, which fits the analysis used in this study (Cheung and Lau 2007). The corresponding model contains two dependent variables (attitude toward advertisement and attitude toward brand) and the independent variables. This model is depicted in figure 2. Mediation analyses are commonly used in psychological research and are appropriate in this study, as they provide more insights into the underlying mechanisms of consumer attitudes.
To test this method, a simple mediation model that follows Hayes (2017) is used.
Following this model, the total effect can be found by regressing attitude toward the brand (AtB) on either greenwashing (GW), green marketing (GM), environmental involvement (EI), and the control variables (C).
𝐴𝑡𝐵! =∝"+ 𝑏"𝐺𝑊 + 𝑏#𝐺𝑀 + 𝑏$𝐸𝐼 +d"𝐶 + 𝜀 (1)
Subsequently, it is crucial to find the effect of these variables on the mediator variable, attitude toward advertisement (AtA). Doing so helps determine whether the key variables are correlated with the mediator, which is essential for mediation analysis.
𝐴𝑡𝐴! =∝#+ 𝑎"𝐺𝑊 + 𝑎#𝐺𝑀 + 𝑎$𝐸𝐼 +d#𝐶 + 𝜀 (2)
Lastly, to find the indirect effect, a final regression needs to be created. The attitude toward the brand is regressed on both the mediator, which shows the indirect impact and the variables that determine the direct effect.
𝐴𝑡𝐵! =∝$+ 𝑐𝐴𝑡𝐴!+ 𝑏′"𝐺𝑊 + 𝑏′#𝐺𝑀 + 𝑏′$𝐸𝐼 +d$𝐶 + 𝜀 (3)
Finally, by combining the estimates of these models, the amount of mediation can be determined. Following the preceding discussion, Figure 2 depicted below represents the conceptual framework of this study. It portrays the mediation model, and the letters in the figure correspond to the coefficients in the model mentioned above.
20
Figure 2. Theoretical structural equation model without interaction effects
Note. Control variables are omitted for clarity.
3.4 Sample and procedure
237 people participated in this between-subjects experiment, of which 45.1% identify as a man (N=107), 51.9% as a woman (N=123), and 3.0% either as non-binary or preferred not to declare their gender (N=7). According to a study conducted by the World Economic Forum (2021), approximately 1% of adults globally identify as neither man nor woman.
Among those younger than 24, approximately 3% identify as non-binary, and 2% prefer not to declare their gender (World Economic Forum, 2021). This is evidence of a new shift towards more than two genders. Hence, the choice was made to include these participants under other genders instead of disregarding their answers. Doing so has no substantial effect on the results of the study. The respondents’ gender is equally distributed among the three conditions, χ2 (6, N=237) = 8.10, p = 0.23).
The participants’ ages range from 17 to 73 (M=29.37, SD=10.76), with a strong tendency towards a younger age range of 17-29 with 68.4%. The proportions of age groups are equally distributed across the three conditions, χ2 (10, N=237) = 6.34, p = 0.79). The majority of the participants have Dutch nationality (62.4%). The proportion of Dutch and non-Dutch nationalities is equally assigned across the three conditions, χ2 (2, N = 237) = 1.27, p = 0.53). Therefore, randomization for all demographics was successful. Extensive descriptions of the full data set can be seen in appendix 3.
21
The experiment was conducted online with the usage of the Qualtrics system. Each participant could open the survey and would be randomly assigned to one of the experimental conditions. When showed the advertisement; participants were asked to examine the
advertisement carefully. They could freely choose to shorten exposure time and continue with the survey via a mouse click. Plausibly, the exposure time may vary among different
experimental conditions as the control group shows no textual information. The green marketing condition (n=83) had an average exposure time of 249 seconds, while the
greenwashing condition (n=77) averaged around 280 seconds, and the control group (n=77) averaged around 236 seconds.
The Qualtrics software warned participants when they did not answer a question.
Participants were thereby unable to proceed with the survey if any answers were blank, which prevented the occurrence of missing values. Moreover, no extreme values were found after reviewing box plots and scatter plots of the individual variables. Consequently, no additional data was deleted.
IV Results
In this section, the results of the previously covered models are discussed. The first part deals with differences in attitudes towards both the advertisement and brand for the various
experimental groups. Secondly, a structural equation model is built, and regressions are run to determine the main effects of greenwashing and green marketing advertisements; both direct and indirect effects are looked at. Next, mediation analysis is performed to examine both the indirect and direct effects of the key variables on the attitude toward the brand.
These regressions are done both for total environmental involvement and its three determinants. Subsequently, interaction terms are added to the model to determine whether the effects of either greenwashing or green marketing depend on consumers’ environmental involvement. Again, regressions and mediation analyses are run. Finally, the chapter ends with further assessments of the robustness of the results.
4.1 Attitude toward the advertisement
The first two hypotheses aim at the main effects of greenwashing and green marketing treatment. A one-way ANOVA was conducted to determine if attitudes towards
22
advertisements differed for the different experimental groups. This test is used as there were three levels of independent variables. Averages and range measure both brand attitudes and attitudes toward advertisements from 1 to 5. The two experimental and control groups function as independent variables.
There was a statistically significant difference between groups, as determined by one- way ANOVA (F (2,234) = 6.65, p = .002). A Kruskal-Wallis H test further demonstrated a statistically significant difference in the attitude toward advertisements between the three groups, χ2(2) = 14.65, p < .001. Finally, a Tukey posthoc test revealed that participants were statistically significantly more positive about the advertisements in the greenwashing group compared to the control group (0.41 ± 0.12 average attitude, p = .001).
Next, participants in the green marketing group were found to be statistically significantly more positive about the advertisements than those in the control group (0.30 ± 0.11 average attitude, p = .026). However, there was no statistically significant difference between the greenwashing and green marketing groups (-0.11 ± 0.11, p = .590). These findings suggest that the greenwashing and green marketing treatments yield strong persuasive results but do not differ in their effectiveness.
4.2 Attitude toward brand
Again, there was a statistically significant difference between groups as established by a one-way ANOVA (F (2,234) = 4.66, p = .01). A Kruskal-Wallis H test further demonstrated a statistically significant difference in participants’ brand evaluation between the three
experimental conditions, χ2(2) = 9.87, p = .007). Finally, a Tukey posthoc test revealed that, more precisely, participants were statistically significantly more positive towards the brand in the greenwashing group compared to the control group (0.32 ± 0.12 average attitude, p = .023). Likewise, participants in the green marketing group were found to be statistically significantly more positive about the brand than those in the control group (0.32 ± 0.12 average attitude, p = .023). However, there was no statistically significant difference between the greenwashing and green marketing conditions (-0.004 ± 0.12, p = .999).
4.3 Structural equation model
Empirical evidence from previous studies indicates that besides the treatments, and environmental involvement, the previously mentioned control variables also influence the
23
attitude toward advertisement and brand. Therefore, an ordinary least squares (OLS) regression will be performed to determine whether these effects are present for this data.
Running a robust OLS regression shows that both the greenwashing advertisement (b
= 0.40, p < 0.001) and the green marketing advertisement (b = 0.30, p = 0.008) significantly affect the attitude toward the advertisement (See table 2). As expected, the output is shown in Table 2 column (3), which reveals that the attitude towards advertisement significantly
explains the change in attitude towards brands (b = 0.73, p < 0.001). Moreover, the regression output shows that adding the attitude towards advertisement to the regression removes the effect of either treatment on brand attitude compared to table 2 column (2). This strongly suggests that both the greenwashing and green marketing advertisements influence brand attitudes through their effect on the evaluation of advertisement.
This reasoning is reinforced by conducting a mediation analysis. The indirect effects of the greenwashing advertisement (b = 0.29, p < 0.001) on brand attitudes mediated by advertisement evaluations are statistically significant. The indirect effects of the green marketing advertisement are also statistically significant (b = 0.20, p = 0.01). Neither greenwashing (b = -0.01, p = 0.95) nor green marketing appeals (b = 0.10, p = 0.24) in advertisements have a statistically significant direct effect on brand evaluation.
These findings imply that hypothesis one stating that “Green marketing
advertisements have a positive impact on brand attitudes by shaping ad attitudes,” and hypothesis two states that “Greenwashing advertisements have a positive impact on brand attitudes by shaping ad attitudes” cannot be rejected. More explicitly, the results show that green marketing and greenwashing appeals in advertisements statistically significantly positively influence attitude towards a brand indirectly by shaping participants’ attitudes toward advertisements.
4.3.1 Environmental involvement
The structural equation model without interaction terms also shows whether a consumer’s level of environmental involvement impacts their attitude toward the
advertisement. Running a robust regression with the average environmental involvement (b = 0.27, p = 0.002) of a participant shows that green involvement has a statistically significant positive effect on someone’s evaluation of an advertisement.
24
To determine which dimensions of green involvement cause this effect specifically, a robust regression is run with the key determinants of green involvement. Regarding the three different dimensions of environmental involvement, only a consumer’s attitude towards green products is found to be statistically significantly related to their attitude towards the
advertisement (b = 0.35, p < .001). In line with prior research, no significant relationship between environmental concern (b = 0.08, p = 0.40) or green purchasing behavior (-0.07, p = 0.26) and advertisement evaluation was found.
Next, to determine whether environmental involvement affects brand evaluations, another robust regression is run. In this case, environmental involvement (b = -0.07, p = 0.33) is not found to have a statistically significant effect on brand evaluations. Moreover, none of the dimensions of environmental involvement: environmental concern (b = -0.11, p = 0.17), attitude toward green products (b = 0.04, p = 0.52), and green purchasing behavior (b = -0.03, p = 0.58) exhibit a direct effect on brand evaluations.
The summary of these results can be found in table 2. In total, the independent variables explain 25.2% of the variance of the attitude toward the ad and 51.3% of the variance of attitude toward the brand.
Table 2. Structural equation modeling unstandardized path coefficients.
Independent variables Dependent variables Attitude toward the advertisement
Attitude toward the brand
(1) b
(2) b
(3) b
Greenwashing advertisement 0.38***
(0.11)
0.28**
(0.12)
-0.01 (0.10)
Green marketing advertisement 0.26**
(0.10)
0.28**
(0.11)
0.10 (0.09)
Green involvement 0.27***
(0.09)
0.12 (0.14)
-0.07 (0.07)
Environmental concern 0.08
(0.09)
-0.05 (0.08)
-0.11 (0.08) Attitudes toward green products 0.35***
(0.07)
0.28***
(0.09)
0.04 (0.06)
Green purchase behavior -0.08
(0.06)
-0.06 (0.06)
-0.03 (0.05)
25
Perceived consumer effectiveness 0.08*
(0.04)
-0.01 (0.04)
-0.05 (0.03) Negative sentiment toward marketing 0.01
(0.05)
0.03 (0.05)
0.04 (0.04)
Age 0.01
(0.003)
0.002 (0.004)
0.0004 (0.004)
Woman 0.11
(0.09)
0.34***
(0.11)
0.25***
(0.09)
Other genders -0.18
(0.24)
-0.10 (0.30)
0.03 (0.19)
Education -0.12***
(0.03)
-0.08**
(0.04)
0.01 (0.03)
Attitude toward the advertisement 0.73***
(0.06) Notes. This table shows a robust OLS of the dependent variables on both the key variables and control variables (1) and (2) and additionally on the attitude toward advertisement (3). Regressions with green involvement components are used for other independent variable results. See appendix 5 for all regression output.
Other genders include both non-binary participants and those who preferred not to state their gender.
⁎⁎⁎ p < .01
⁎⁎ p < .05
⁎ p < .10
Next, a mediation analysis is conducted to determine whether environmental involvement and attitude toward green products affect brand evaluation through
advertisement evaluation. From the mediation analysis, one can see that the total effect of a person’s attitude towards green products (0.30, p < 0.001) on brand evaluation comes from the indirect effect of a person’s attitude towards green products (0.26, p < 0.001) through how they evaluate the advertisements. However, green involvement (b = 0.11, p = 0.21) is not found to have a statistically significant total effect on brand evaluation, but green
involvement (b = 0.20, p = 0.002) does have a statistically significant indirect effect on brand evaluation. A summary of the disaggregated results of the mediation analysis can be found in Table 3 below.
26
Table 3. Disaggregated results of mediation analysis for key variables
Independent variables
b’
Direct effect ac
Indirect effect
b
Total effect Mediation
Greenwashing 0.004
(0.01)
0.27***
(0.08)
0.28**
(0.12)
96.43%
Green Marketing 0.10
(0.09)
0.18**
(0.08)
0.28**
(0.11)
64.29%
Green involvement -0.07 (0.07)
0.20***
(0.06)
0.12 (0.08)
74.07%
Attitudes toward green products
0.04 (0.06)
0.25***
(0.06)
0.28***
(0.09)
89.29%
Notes. This table shows the disaggregated results of the mediation analysis for the statistically significant key variables with attitude toward advertisement as mediator and attitude toward the brand as the independent variable.
⁎⁎⁎ p < .01
⁎⁎ p < .05
⁎ p < .10
4.4 Interaction effects
Hypotheses 3 and 4 assume that the effects of green marketing versus greenwashing
advertisements would depend on consumers’ green involvement. To determine whether this is the case, interaction terms are added to the previous model.
Running a robust regression shows that contrary to expectations, neither greenwashing (b = 0.02, p = 0.92) nor green marketing appeals (b = 0.04, p = 0.85) in combination with a person’s environmental involvement statistically significantly impact advertisement evaluation. By looking at the different components of environmental
involvement, one can see that none of the determinants have a statistically significant impact on the evaluation of an advertisement (See table 4).
The next step looks at whether the interaction terms affect a person’s attitude toward the brand. The analysis does reveal a marginally statistically significant interaction between the effect of green marketing advertisement exposure and a person’s total green involvement products (b = -0.24, p = 0.08). The negative sign of this interaction signals that the effect of exposure to green marketing advertisements decreases with rising levels of green
27
involvement. This contradicts hypothesis 4, which states: “The effects of green marketing will be stronger for consumers high in green involvement.”
Running a robust regression with the different components of green involvement shows that a person’s attitude toward green products in combination with green marketing appeals (b = -0.34, p = 0.01) has a statistically significant effect on a person’s attitude toward the brand. The negative sign of this interaction signals that the effect of exposure to green marketing advertisements on brand evaluation decreases when a person has a more positive attitude toward green products. This indicates that when people view green products more positively, they dislike the green marketing advertisement compared to the control
advertisement. This component seems to explain the contradiction with hypothesis 4.
Contrary to expectations, no statistically significant difference was found for any element for the advertisements with greenwashing appeals.
28
Table 4. Structural equation modeling unstandardized path coefficients for interaction terms Independent variables Dependent variables
Attitude toward the advertisement
Attitude toward the brand
(1) b
(2) b
(3) b Greenwashing advertisement x green
involvement
0.02 (0.21)
-0.004 (0.21)
-0.02 (0.15) Green marketing advertisement x
green involvement
0.04 (0.21)
-0.21 (0.18)
-0.24*
(0.14) Greenwashing advertisement x
environmental concern
-0.07 (0.21)
0.02 (0.20)
0.07 (0.16) Greenwashing advertisement x attitude
toward green products
0.28 (0.23)
-0.03 (0.22)
-0.23 (0.15) Greenwashing advertisement x green
purchase behavior
-0.12 (0.16)
0.02 (0.15)
0.11 (0.12) Green marketing advertisement x
environmental concern
0.30 (0.20)
0.20 (0.19)
-0.01 (0.19) Green marketing advertisement x
attitude toward green products
0.03 (0.21)
-0.32*
(0.19)
-0.34**
(0.13) Green marketing advertisement x
green purchase behavior
-0.18 (0.13)
-0.07 (0.13)
0.06 (0.11)
Attitude toward advertisement 0.73***
(0.06) Note. This table shows a robust OLS of the dependent variables on the interaction terms, key, and control variables (1) and (2) and additionally on the attitude toward advertisement (3). Only interaction terms are given.
See Appendix 6 for all regression output.
⁎⁎⁎ p < .01
⁎⁎ p < .05
⁎ p < .10
To interpret the interaction effect, an additional analysis demonstrates the effect of green marketing appeals at low (one standard deviation below the mean), moderate (at the mean) and high (one standard above the mean) levels of green involvement and attitude toward green products using manifest variables. The greenwashing treatment was excluded as the interaction effects were insignificant. Figure 3 and figure 4 visualize this. It shows that
29
those with more positive attitudes toward green products prefer the control over the green marketing advertisement.
Figure 3. Attitude toward the brand as a function of experimental group and green involvement
Note. The interaction effect with the greenwashing group was statistically insignificant and therefore not included.
Figure 4. Attitude toward the brand as a function of experimental group and attitude toward green products
Note. The interaction effect with the greenwashing group was statistically insignificant and therefore not included.
A mediation analysis further shows that the effect of the interaction terms come from the direct effect the interaction terms have on the attitude toward the brand and that the
