Unwrapping the food packaging industry
Is the Dutch market ready
for a package-free supermarket?
Nomi Meijs (11402156) Master Thesis
MSc in Business Administration
Entrepreneurship and Innovation track University of Amsterdam
Final submission date: 20-06-2018 Thesis supervisor: dr. Joeri Sol
Statement of originality
This document is written by Nomi Meijs who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Table of content
1. Introduction 6
2. Literature review 8
2.1 Changing consumer behavior 8
2.1.1 Increased ecological consciousness among consumers 8 2.1.2 The influence of mass media on consumers 9 2.1.3 Consumers’ expectations of businesses and the government 10
2.1.4 Companies’ reactions 11
2.1.5 Willingness to pay for eco-friendly products 12
2.2 Zooming in on packaging 13
2.2.1 The importance of packaging for consumers’ trust 13 2.2.2 The role of packaging in a sustainable food industry 15
2.3 Research question 18 2.4 Conceptual framework 18 3. Method 20 3.1 Research design 20 3.2 Questionnaire design 20 3.3 Data collection 21 3.4 Sample 22 3.5 Variables 24
3.5.1 Willingness to pay for eco-friendly products 24
3.5.2 Ecological consciousness 24 3.5.3 Packaging preferences 25 3.5.4 Product type 25 3.5.5 Gender 26 3.5.6 Age category 26 3.5.7 Educational level 26 3.5.8 Income category 26 3.5.9 Living environment 26
4. Analysis and results 28
4.1 Preliminary tests 28
4.1.1 Missing values 28
4.1.3 Descriptive statistics 29 4.1.4 Reliability test 32 4.1.5 Correlation analysis 33 4.2 Results 34 4.2.1 Independent t-test 34 4.2.2 Regression analysis 35 4.2.3 Mediation analysis 37
4.2.4 Ranking favorite packaging type per product 40
5. Discussion 42
5.1 General discussion and implications 42 5.2 Limitations and future research 46
6. Conclusion 49
Bibliography 50
Appendix A: Questionnaire English version 58
Appendix B: Questionnaire Dutch version 61
Appendix C: Histograms of the distribution 64
Abstract
This research takes a closer look at the relationship between ecological consciousness and willingness to pay for eco-friendly products and investigates which role packaging preferences play for this relationship. Additionally, it aims to create an overview outlining for which product types consumers are acceptant towards no packaging or more sustainable packaging and ultimately investigates if the Dutch market is ready for a package-free supermarket.
Results derived from the performed cross-sectional experiment show that ecological consciousness positively affects willingness to pay for eco-friendly products and that packaging preferences mediate that relationship. There is no effect detected of exposure to media output concerning eco-friendly packaging alternatives before measuring levels of ecological consciousness and willingness to pay for eco-friendly products. The Dutch market does not seem ready yet for a fully package-free supermarket but appears acceptant towards less packaging or more environmentally friendly packaging materials. This study highlights the need for making more sustainable packaging choices in the food industry and offers guidelines for food producers and retailers concerning which products show potential for packaging improvements.
Keywords: ecological consciousness, willingness to pay, package-free supermarket, food
1. Introduction
The topic of food is one that concerns everyone. While a segment of activists and conscious consumers already has been aware of the impacts food choices have on personal health, animal and human welfare and environmental sustainability, the ‘regular consumer’ is becoming increasingly aware of the effects of his or her food choices too. Scientists emphasize how important it is for consumers to shift their dietary patterns towards one that meets sustainability criteria, resulting in shocking articles and documentaries unraveling the dark side of the food supply chain (Aschemann-Witzel, 2015). Examples like ‘What the Health’ and ‘Cowspiracy’ are playing a part in this and help consumers realize the high impact of the way food is produced and processed. This growing sense of responsibility affects food companies as well because consumers prefer to purchase from companies which represent the same values and beliefs as they have themselves (Cone Communications, 2017). As a result, food companies feel the pressure to make adjustments and improvements in their production processes and take a critical look at their image and actions.
Some of the environmental consequences of the current food supply chain are pollution and food contamination, depletion and contamination of natural resources, climate change and diminution of food safety (Krishnan, 2009). Looking at the European Union (EU) specifically shows that the food supply chain is responsible for 17% of direct greenhouse gas emission and 28% of the consumption of material resources (European Commission, 2011). The same report by the European Commission reveals that in the EU alone an average of 180kg food per person is wasted yearly, of which the biggest part is still suitable for consumption. However, food waste is not only a matter of concern for households: before food reaches the consumers, one third of produced food globally is already wasted in the food supply chain (Steinfeld et al., 2006).
Besides food waste, material waste is a matter of concern in the food industry as well (Weber et al., 2002). Of all food and beverages, 99.8% is packed at least once in its lifespan. Packaging is essential for otherwise non-solid food and protects it from external influences like sunlight, heat, insects, dirt, and so on. It extends shelf-life and maintains food safety and quality (Restuccia et al., 2010). Besides the practical application,
packaging is also used for convenience, communication, and marketing. Clearly, packaging is important and can not be left out, but it also has its negative side. On average, 25 million tons of plastics are used yearly throughout the world of which 25% is used for packaging purposes (Weber et al., 2002). For the last twenty years, the focus of the negative side lay on the waste that came after unpacking. Except for paper and board, packaging materials normally used nowadays are non-renewable, meaning that alternatives need to be found. There is a growing awareness among companies concerning this issue and researchers are looking for solutions and ways to develop packaging materials based on renewable resources (Mohanty, 2002). An example of a strong reaction to the changing consumer behavior comes from Coca-Cola, which announced in the beginning of 2018 that the company wants to help collect and recycle all cans and bottles sold by 2030 because they recognize their responsibility when it comes to environmental harm caused by packaging material. In addition, Coca-Cola’s R&D department works on creating better bottles by working with alternative materials to reduce the amount of plastic (LaVito, 2018).
It is clear that change in the food supply change is needed to obtain and retain a sustainable food industry and to counteract the negative environmental consequences that come from the food industry. Therefore, this research focuses on the food supply change and looks for potential adjustments which can be made in order to promote sustainability. As the food supply chain as a whole is too broad to research, it is not possible to dig deep into every aspect concerning it. Therefore this study zooms in on packaging and ultimately aims to find out if the Dutch market is ready for a package-free supermarket.
The first section of this thesis is the literature review in which the context of the research field is described and an introduction to the research question and hypotheses are provided. After that, the methodology section outlines the data collection process and provides further information about the variables. The next chapter describes the ran analyses and gives the results. The research ends with the discussion and conclusion.
2. Literature review
This chapter discusses findings from existing literature that are essential for research on the ecological behavior of consumers in the food industry. First, changing consumer behavior is examined. This covers the observation of increased ecological consciousness among consumers, the role of media, consumers’ expectations of businesses and the government, the way companies react to the changing consumer expectations and demands, and lastly consumers’ willingness to pay for eco-friendly products. After this, food packaging and consumer preferences are discussed, covering the importance of consumer trust and the role packaging plays in a sustainable food industry. This section ends with a visual overview of the research model and the three hypotheses that lead the study.
2.1 Changing consumer behavior
Consumer behavior is an ever-changing concept. Consumers nowadays have become more and more selective in their product choices, they have higher demands and buy more than ever. This results in shortened product life cycles and intensified competition among companies (Trittico, 2015). Understanding how consumer behavior changes and which new demands arise is essential for companies, especially if they want to keep up with the fastened pace of change.
2.1.1 Increased ecological consciousness among consumers
A study by the European Commission Joint Research Centre (2006), aiming to identify which products cause greatest environmental problems, found that the category ‘food and drink’ is responsible for 20-30% of environmental impacts of total consumption. Environmental damage caused by the food industry includes greenhouse gases, soil and water damage, global warming, agricultural issues like chemicals, waste disposal, water pollution and deforestation (Maloni and Brown, 2006; Hamerschlag, 2011). Greenhouse gas emissions are a major source of problems and originate from essential parts of the food supply chain: production, transportation, storage, cooling, energy and water consumption, wastage of food and packaging (Scarborough et al., 2013; Roy et al., 2014). Research by the United Nations Environment Program, shortly UNEP, shows that environmental problems do not stop once food and beverages leave the retailer: household consumption is also identified as a significant source of environmental
problems (Hertwich et al., 2010). The study revealed that household consumption in most countries accounts for over 60% of all environmental impacts directly coming from consumption.
Antonetti and Maklan (2014) found that experienced guilt and pride after a consumption influence consumers in a way where both emotions regulate the decision of buying sustainable alternatives or not. This results in consumers behaving more sustainably when they believe that they can affect social and environmental issues with the choices they make. Previous research by Roberts (1996) gives further explanation as it appeared that consumers are more willing to change their behavior when they have the sense that they can make a change individually. Even though a trend is detected of consumers having an increasing responsibility sense together with a slow adjustment in consumption patterns from ‘normal products’ towards organic or locally produced products, the current market shares of eco-friendly products still shows to be fairly low: 1-6% (Nielsen, 2011). However, this share is expected to keep rising. According to a study by Forbes, every generation influences society. The research revealed that the biggest current workforce exists of millennials, and those millennials pursue a good, healthy food pattern that is eco-friendly (Williams, 2016; Fry, 2018).
2.1.2 The influence of mass media on consumers
Several environmental disasters like the Exxon Valdez oil spill or the Great Pacific garbage patch have awakened consumers’ sense of ecological consciousness. When such environmental incidents happen, all kinds of media outlets communicate heavily about it. According to Mutz (1992), mass media strongly influences the mass opinion because people are sensitive to opinions, attitudes, and beliefs of ‘anonymous others’. This influencing power increases environmental consciousness even more. Luedecke and Boyoff (2017, p.1) give the following explanation for this: 'In all media, actors such as
publishers, editors, journalists, and other content producers such as online bloggers generate, interpret, and communicate images, information, and imaginaries for varied forms of consumption. These “media representations” are therefore critical inputs to what becomes public discourse on today’s environmental issues.’. The progressive increase in
become a necessity and adjustments have to be made in all areas possible (Tantawi et al., 2009).
2.1.3 Consumers’ expectations of businesses and the government
The detected changing consumer behavior impacts businesses and governments, as consumers expect those parties to take responsibility (Fahlquist, 2009). A study performed by Cone Communications in 2017 revealed that 87% of the consumers would purchase a product because the selling company advocated a cause that the consumer strongly supports. Au contraire, 76% of the consumers refuses to purchase from a company that supports the opposite of their own beliefs. Hume (2010) also researched Generation Y - the most consumption-oriented generation ever - and sustainable consumption. Her study revealed that consumers think that the governments should focus their political agendas more on environmental issues. Two of the striking outcomes of the research were that consumers want the government to take responsibility for reducing packaging and waste, and that the government has to perform societal marketing actions to promote sustainable consumption and respectful behavior towards the planet and inclusive societies. These insights highlight the need for governments and businesses to step up and include environmental activities in their practices.
Sen and Bhattacharya (2013) did a study on environmental and societal activities by businesses by researching consumers’ reactions to Corporate Social Responsibility (CSR). Firstly, it is important to define the concept of CSR. One rather broad interpretation comes from Davis and Blomstrom (1975, p.6): ‘The managerial obligation to take action to protect
and improve both the welfare of society as a whole and the interest of organizations’.
Other researchers chose to focus purely on the economic point of view, like how CSR can maximize returns to shareholders (Friedman, 1970; Zenisek, 1979), or a proactive social responsiveness view which looks at a company’s long-term role in the social system (McGee, 1998). For this report, Brown and Dacin’s (1997, p.68) larger, societal definition of CSR is used, being: ‘status and activities with respect to [i.e., responsiveness to] its
perceived societal obligations’. Dawkins and Lewis (2003, p.185) specified four CSR
areas, which are ‘treatment of employees, community involvement, ethical issues and
environmental issues’. Sen and Bhattacharya’s research shows that overall, consumers
decrease consumers’ purchasing intentions. When negative CSR information is retrieved, this has stronger negative effects than positive CSR information shows positive effects, meaning that managers must be strongly aware of the effects of CSR and the danger of behaving socially irresponsible. The findings also show that consumers’ responses on CSR account for two areas: company-specific factors (for example product quality or chosen CSR-activities) and individual-specific factors (like consumers’ personal interest in the CSR-activities and their general norms and values). Governments also play a significant role in CSR because they have the power to facilitate CSR by implementing policies and regulations and providing subsidies. A common form of government regulation concerning CSR is tax incentives (Moon et al., 2010).
2.1.4 Companies’ reactions
Literature shows that consumer behavior has been changing throughout the years and reached a level where companies need to react. Environmental conservation and increasing demands for natural resources are forcing companies to rethink both their business models and their supply chain operations (Wu and Pagell, 2010). Several visible environmental disasters in the 60’s and the 70’s (e.g., Three Mile Island, Love Canal, and the previously mentioned Exxon Valdez) exposed the need to improve companies’ environmental strategies (Walton et al., 1998). This strategy should not only be deployed in the top management but in every layer of the organization, including all steps of the supply chain. Walton, Handfield and Melnyk (1998, p.3) studied the furniture supply chain and focused on how to make it greener, using the words ‘Greening the supply chain’. While the research is somewhat outdated, it shows the topic’s relevance and the fact that a solution remains to be found. Twenty years ago, governments and consumers were already pushing organizations to think twice about their environmental responsibility. At the same time, organizations were looking for ways to create cheaper processes and improve customer service. This study demonstrates that purchasing and supply chain managers need to integrate suppliers into their environmental initiatives. It also explains that companies were starting to realize the competitive advantage that environmental awareness could bring. Besides the competitive advantage and meeting consumers’ and governments’ demands, proactive environmental strategies could bring waste and cost reduction, improving company results as well.
The primary challenge for companies nowadays becomes running a viable business while not jeopardizing natural environment in the future. According to Wu and Pagell (2010) this is a difficult challenge for companies considering that there is high information uncertainty when it comes to how and to which degree their supply chains have an environmental impact. The rise of environmental concern created a new type of consumers: the ‘green or
ecological consumers’ (Finisterra do Paço and Barata Raposo, 2010, p.429). This offers
possibilities for ‘green marketing’ which can lead to competitive advantage. Wu and Pagell (2010) state that organizations can differentiate themselves by proactively adopting and promoting environmental initiatives, which can result in competitive advantage. This accounts for every B2C supply chain. Motivated workers can be of great help in tackling the challenge of making the supply chain greener; they have a clear vision of what is going on in the supply chain and where possibilities for improvement are found. As the authors say: ‘Each decision informs subsequent decisions, creating a feedback loop of knowledge,
practices and routines that moves the organization along their unique trajectory’ (Wu and
Pagell, 2010, p.588). Step by step with continuous improvement the existing supply chain becomes greener.
2.1.5 Willingness to pay for eco-friendly products
The four most important buying criteria are quality, value, convenience, and price, and especially the latter plays an essential role in the market for eco-friendly products. (Roberts, 1996). The majority of consumers is willing to purchase green products, as long as there is no price difference with regular products (Ishaswini and Datta, 2011). Krystallis et al. (2005) researched the relationship between ‘Food Category’ and ‘Willingness to Purchase’ (WTP). In their research among Greek consumers they considered sixteen different food products and the research showed that WTP indeed differed among food categories. Consumers are most willing to purchase and pay extra for organic fruits and vegetables, and tinned food and biscuits scored the lowest on WTP. The difference is explained by different perceptions per food category on the areas of quality and security of food, the trustworthiness of certification and brand names. Haws et al. (2014) also show the consideration of costs in the case of environmentally friendly products. They state that consumers tend to be reluctant in buying eco-friendly products because they have the perception or assumption that these are more expensive or less effective. This is not only true for consumers that never had the intention to purchase environmentally friendly
products, but this also accounts for consumers with high ecological consciousness. Greener consumers are reluctant to purchase a product that is marked as eco-friendly if they feel like the product is of lower value for the money (and therefore a waste of financial resources), or if the product is considered less effective (if the consumer is required to use more of it, which is a waste of physical resources). Making consumers more aware of the consequences of their buying choices can increase sales of eco-friendly products. Increased consumer demand can cause price reductions in eco-friendly products, which causes the snowball effect where consumer demand increases which in turn leads to price reductions (Ishaswini and Datta, 2011). Hence, it can be said that a positive correlation between ecological consciousness and willingness to pay for eco-friendly products exist. This first section of the literature review leads to the first two hypotheses:
H1: There is a positive relationship between ecological consciousness and willingness to
pay for eco-friendly products, meaning that the more ecologically conscious someone is, the more he/she will be willing to pay for eco-friendly products.
H2: Consumers exposed to media output concerning packaging alternatives are more
willing to pay for eco-friendly products.
2.2 Zooming in on packaging
The following part of the literature review covers the role that packaging plays for consumers and why more sustainable packaging alternatives are necessary.
2.2.1 The importance of packaging for consumers’ trust
Maloni and Brown (2006) examined the application of CSR in the supply chain of the food industry, or the food supply chain. Their research focuses on eight areas of CSR, which are: animal welfare, biotechnology, health and safety, fair trade, labor and human rights, procurement, community and the environment. Food companies are the prime target for social concern because food is tangible, applicable to everybody and effects of production processes are visible. Therefore food supply chain deficiencies come with great costs for an organization, both financially and reputation wise. But when used well, CSR can bring great benefits for an organization. According to the authors, brand image is one of the
main reasons for companies to pursue CSR strategies and is retrieved by innovation, marketing, and publicity, resulting in competitive positioning. These reasons underline that living up to a company’s ethical and social responsibility is not only important because of the preservation of the globe but also for a company’s financial results.
Nonetheless, food producers should think beyond the physical end product when it comes to environmental improvements. Many steps throughout the supply chain provide room for improvement. Roth et al. (2008) studied the food supply chain with a focus on the Chinese market and the topic of food safety. The research looks at trends in the food supply chain, starting a couple of decades ago. The average supply chain was fairly short and inefficient in the past because of the mainly local and regional reach, but this has changed throughout the years. Globalization offered new solutions to retailers which made the production process cheaper and provided new international selling opportunities. These new supply chain practices also bring along some potential problems. When the supply chain becomes longer, the number of suppliers increases. The higher the number of suppliers and the newer the relationship with them, the higher the chances that something goes wrong. Therefore, instead of only looking for the cheapest suppliers, companies should invest in long-term relationships with trustworthy suppliers to safeguard quality. Trustworthiness is something that Roth et al. (2008) discuss in their conceptual framework called ‘the Six T’s of supply chain’, where the T’s stand for traceability, transparency, testability, time, trust, and training. These T’s are important among business parties within the supply chain, but also matter from a consumer’s point of view. The Chinese food industry is a good example of the problem: many producers falsely label their products ‘organic’ (Hasimu et al., 2017), going against several of the T’s. These false labels are not only present in the Chinese market, the food industry in Europe and North America similarly features an abundance of (false) product certifications, trying to lure customers into buying those products by creating a feeling of trust and satisfaction. These kinds of unsubstantiated green claims, also known as ‘greenwashing’, and other business scandals about providing false information, have led to decreasing consumer trust in environmentally friendly products (Choi et al., 2007; Chen and Chang, 2013; Jahdi and Acikdilli, 2009). As Nuttavuthisit and Thøgersen (2017, p.323) say: ‘… many consumers
price is charged, which is usually the case. Green product attributes are credence attributes, which means that they cannot be verified by the consumer even after purchase and consumption.’. Edelman’s Trust Barometer research (Ries et al., 2018) proves this as
consumers’ trust in the food and beverage sector has dropped 4% in comparison to last year.
As mentioned before, packaging, which plays a widespread role in the food supply chain, serves many functions. One of them is providing information to the customer: information about the ingredients and their origin, but also about how eco-friendly the product is. Therefore it is important that the food industry maximizes its efforts to work in a more environmentally friendly way, to communicate this to the customers (including through packaging) and to live up to promises. This leads to the company’s image becoming more trustworthy, which in turn may lead to an overall rise in consumers’ trust in the food and beverage sector. If a company lives up to its ethical and social responsibilities, this is beneficial for both the company’s competitive advantage and preservation of the globe (Maloni and Brown, 2006). Besides only making their food and beverages greener and communicating this through packaging, businesses should make an effort to adopt environmentally friendly practices throughout the supply chain. Packaging, for one, can become more sustainable by using less of it or by replacing materials (EPA, 1989).
2.2.2 The role of packaging in a sustainable food industry
Brisson (1993) divides packaging into three categories: primary, secondary and tertiary. Primary packaging, also known as sales packaging, is the direct packaging on the end product. Examples are a milk carton or cereal box. The function of secondary or additional packaging is to keep units of products together. For example, a big plastic bag containing fifteen smaller bags of single-portion crisps. Tertiary packaging, or transport packaging, defines all packaging used for protection during transport from the producer to the store, like pallets or styrofoam. As became evident in the introduction, packaging is a double-edged sword. Despite its important role for food preservation and marketing purposes, it is also a major cause of material waste. Therefore it is necessary that researchers keep looking for recycling solutions and ways to develop packaging materials based on renewable resources. Marsh and Bugusu (2007) made of a list of all packaging materials currently used in the food industry. This list is very long as every material comes in several
forms, but the main categories they discuss are glass, metal, plastics and paper and paperboard. According to Weber et al. (2002), except for paper and paperboard, all of those packaging materials are based on non-renewable materials. With their research, they emphasize the need for commercially used renewable packaging materials. In 1989, The United States Environmental Protection Agency, or EPA, formed guidelines for solid waste management. These guidelines entail recycling, composting, source reduction, combustion and landfilling. The EPA gave the following description of source reduction thirteen years later: ‘Source reduction encompasses using less packaging, designing
products to last longer, and reusing products and materials’ (EPA, 2002), and states that
this is the best solution for the impact of waste on the environment. Brutscher (s.d.) describes the trend of package-free shopping and says that more and more supermarkets abandoned the use of plastic bags, or even discard any kind of packaging. Some cities that are home to successful package-free supermarkets are Berlin, London and Austin. The Netherlands has had at least four unsuccessful attempts of package-free supermarkets spread across the country. Nonetheless, social entrepreneurs keep trying as the Netherlands currently has one physical zero-waste supermarket - which is struggling to remain open - and one fairly new online package-free supermarket (Thijs, 2018).
Besides the environmental gains that could be reached by using more eco-friendly packaging materials or by reducing the amount of packaging used, progress can also be made when packaging causes reduction of food waste. Williams and Wikström (2011, p.43) state that when packaging lives up to certain standards, it can reduce food waste by increasing convenience for the consumer. The intended characteristics are: ‘protect the
food from physical damage; it should be easy to reseal to avoid biological deterioration; it should be easy to empty completely, and it should be provided in sizes that avoid leftovers. […] It should also provide the correct information (content, best-before-date, etc.) about the food to avoid that fresh food is thrown away.’ Williams and Wikström designed a model
showing the food supply system which explains where the environmental impact of packaging is the biggest. Packaging is present between the steps of ‘Food Industry’ and ‘Consumer’, which covers the main part of the food supply chain. The model is presented in Figure 1.
A development in the packaging industry is that of ‘active and intelligent packaging’. Different from traditional packaging, this new way of packaging interacts with the surrounding environment and can monitor the packed food. Examples of what it can signal are time-temperature and riping indicators (Pereira de Abreu et al., 2012). These indicators can help increase the shelf-life of processed food and therefore decrease food waste while at the same time meeting consumers’ demands when it comes to quality, safety, and freshness. According to Pereira de Abreu et al. (2012), Asia and the US are forerunners when it comes to smart packaging, Europe has strong regulations that limit the possibilities. However, the European Union has made some adjustments lately (regulations 1935/2004/EC and 450/2009/EC) to loosen the rules when it comes to the application of certain materials and safety, as well as marketing-related requirements related to intelligent and active packaging (Restuccia et al., 2010).
Evidently, significant reduction of environmental damage can be reached when food losses are lowered. Packaging plays an essential part in food losses and therefore makes an important aspect to consider. Consumer behavior is a crucial link in trying to reduce environmental damage in the food industry. How do consumers react to, for example, printed information, refillable containers or the use of an extraordinary amount of packaging? According to previous literature, more research should be done in the field of packaging and consumer behavior (Williams and Wikström, 2011). Therefore, this
Figure 1: Schematic illustration of the food supply system. F represents the environmental impact of
agriculture, food industry, retail and consumer, T is the environmental impact of packaging and packaging waste, and W is the environmental impact of food waste. Transports are included in each subsystem. Reprinted from Environmental Impact of Packaging and Food Losses in a Life Cycle Perspective: a Comparative Analysis of Five Food Items, by Williams, H. and Wikström, F. (2011),
research aims to find out what role packaging preferences play for customers. This leads to the final hypothesis:
H3: Packaging preferences mediate the relationship between ecological consciousness
and willingness to pay for eco-friendly products, so that packaging preferences depend on ecological consciousness, which in turn has a positive relationship with willingness to pay for eco-friendly products.
2.3 Research question
Reviewing previous literature demonstrates that sustainability in the food industry is a matter considered important enough to research broadly. It became clear that the food supply chain covers many different areas and is too broad to focus on entirely. Packaging material and its connection to food waste matter considerably for sustainability in the food supply chain and previous studies on packaging and its environmental impact aroused the need to dig deeper in the area of packaging and consumer behavior. Taking this into account leads to the following question: are Dutch consumers ready for a package-free supermarket? Or the more precise research question:
To what extent do packaging preferences mediate the relationship for ecological consciousness and the willingness to pay for eco-friendly products?
This research sheds light on the possibilities of packaging when it comes to increasing sustainability in the food supply chain, and aims to find out whether or not the Dutch market is ready for a package-free supermarket.Ultimately, this thesis serves as a source of inspiration for both consumers and companies in the food industry.
2.4 Conceptual framework
This research is quantitative with an experimental approach. The starting point for this study was to look for ways to increase sustainability in the food supply chain and the interest to find out if the Dutch market is ready now for a package-free supermarket. A literature study was done to get more background information on the relevant variables. Based on the literature findings, three hypotheses are formulated to test the previously mentioned assumption. The conceptual model visualizes the key variables and the relationships between them (Figure 2).
H1: There is a positive relationship between ecological consciousness and willingness to
pay for eco-friendly products, meaning that the more ecologically conscious someone is, the more he/she will be willing to pay for eco-friendly products.
H2: Consumers exposed to media output concerning packaging alternatives are more
willing to pay for eco-friendly products.
H3: Packaging preferences mediate the relationship between ecological consciousness
and willingness to pay for eco-friendly products, so that packaging preferences depend on ecological consciousness, which in turn has a positive relationship with willingness to pay for eco-friendly products.
In addition to the three stated hypotheses, explorative research is performed to find out if respondents have different packaging preferences per product type. This information sheds light on the question if respondents are willing to use their own packaging or no packaging at all when doing groceries.
H3
Ecological
consciousness Willingness to pay for eco-friendly products Packaging preferences H1 + + +
3. Method
To test the hypotheses for this research, data was collected through online surveys. The following section starts by discussing the research design which explains how the questionnaire was set up. After this, the sample and data collection are set out. Lastly, the different variables relevant to this research are explained.
3.1 Research design
The nature of this quantitative study is between-subjects experimental design. In this kind of study, the sample is divided into two groups: the treatment group and the control group. In this research, the treatment group started the questionnaire by reading an article about a package-free supermarket, and the control group started without reading any article. The inclusion of this news article, which is assigned to respondents randomly, adds insight into the way extra information about ecological alternatives for ‘normal’ packaging influences the respondent when filling out the survey. The random assignment that is part of between-subjects experimental design is beneficial because it increases the usability of the final results (Charness et al., 2012).
In order to test consumers’ receptiveness towards a package-free supermarket in the Netherlands, or at least more eco-friendly alternatives compared to the popular plastic, an anonymous online self-report questionnaire was sent out. Using this type of questionnaires is advantageous because of several reasons: (1) possibility to collect large numbers of respondents, leading to higher statistical power, (2) respondents can fill out the questionnaire at their own pace, whenever and wherever suits them best, and (3) the standardized character of the questionnaires makes processing and analyzing less complicated for the researcher (Mann, 1998). Additionally, (4) the costs are kept low since it is not necessary to rent lab space, hire people, and so on, and (5) it is possible to reach geographically and demographically diverse respondents (Reips, 2000).
3.2 Questionnaire design
By using questionnaires (see Appendices A and B), this study aims to find out if and how consumers can be triggered to use less packaging, or even no packaging, and which requirements they demand when it comes to packaging of their food.
The survey consists of questions relating to the different variables of this research and closes off with demographic questions. It starts with questions about ecological consciousness, followed by questions on willingness to pay for eco-friendly products and packaging preferences. Thereafter, the respondent is provided with an overview of certain durable and non-durable products that are commonly used in the Netherlands on the y-ax and packaging types on the x-as. Respondents are asked to fill out the boxes of the packaging types they would accept for the specific product type in ranked order (1 = favorite, 6 = least favorite), or leave the box blank if they think a packaging type does not apply.
Most of the questions are based on a five-point Likert scale (1 = strongly agree, 5 = strongly disagree), and stem from previous research. For the sake of congruence, some questions were adjusted to conform the style of questions of the overall questionnaire. Two self-designed questions about how often the respondent uses packaged food and his/her willingness to pay extra for eco-friendly products were added. The questionnaire was designed in both Dutch and English to get as many respondents as possible.
A pre-test was done before distributing the survey to filter out uncertainties, misunderstandings, and spelling or grammar mistakes. A pilot study ensures that the respondents do not misinterpret questions so that the researcher actually researches what he or she aims to (Bryman and Bell, 2011). Five persons participated in the pilot study for both the Dutch and the English survey. After they made their suggestions, some improvements were made to the surveys before sending them out to others.
3.3 Data collection
The population of this research consists of all supermarket consumers in the Netherlands and contains both natives as non-natives, therefore a Dutch as well as an English version of the survey were constructed in the program Qualtrics. The questionnaires were spread out through the use of personal network, so non-probability convenience sampling applies. Non-probability convenience sampling is used because this type of sampling saves time and money and scores high on the simplicity of sampling (Saunders et al., 2012). However, it also brings along some disadvantages, starting with low representativeness. Gerring (2004, p.348) gives the following definition of representativeness:
‘Representativeness — the degree to which causal relationships evidenced by that single unit may be assumed to be true for a larger set of (unstudied) units.’ When using
non-probability convenience sampling, the representativeness decreases and results become less generalizable (Acharya et al., 2013). Additionally, using this type of sampling comes with the risks that the sample is biased and that outliers may occur due to self-selection of the subjects (Etikan et al., 2016). Even though this type of research comes with some downsides, Saunders et al. (2012) state that it is still possible to generalize about the population when using this type of sampling, just not on statistical grounds. Therefore, continuation of non-probability convenience sampling is supported for this research.
For the sake of analyzability of the data, the attempted number of respondents was at least 200. The average response rate from individuals is about 52.7% (Baruch and Holtom, 2008), therefore at least 400 people had to be reached. The questionnaires were posted on the social media platforms Facebook and LinkedIn, and they were sent out in personal invitations through Facebook and WhatsApp. The final sample consisted of 297 respondents of who 261 filled the survey out completely. UvA guidelines suggest 30 to 50 respondents per treatment for experiments; therefore the recorded sample size of this research is big enough to continue with the study.
3.4 Sample
The survey opened on April 19th, 2018 and closed three weeks later on May 10th. Except
for the demographics outcome, which will be reviewed in this section, all results are discussed in the next chapter. The socio-demographics are shown in Table 1.
As said, out of 297 people who started filling out the questionnaire, 261 fully completed it (response rate 87.9%). Of these 261 respondents (Mage = 2.86, SDage = 1.23, age
category: 20-29 years) the majority was female (N = 177, 67.8%) and a number of 84 men completed the survey (32.2%). Studying the educational level showed that most respondents received their HBO/WO Bachelor degree (60.5%) or WO Master degree (23.4%). Only 0.4% did not finish their high school (yet), and 3.8% stopped after high school or is still studying. 6.8% of the respondents marked MBO as the highest obtained degree, and 4.6% has an advanced degree beyond a Master’s degree. The remaining 0.8% would rather not say what their highest obtained education is.
When it comes to income, 20.3% said to earn between €34.000 and €40.000 gross per year, 18.8% earns more than €40.000 per year, and 8.8% would rather not say. Over half of the respondents reported to have an income lower than €30.000 gross per year (52.1%), which is explainable when looking at the combination of average age and education level. As previously said, 60.5% has an HBO/WO bachelor degree, which in combination with the average age category of 20-29 of the majority of respondents suggests that they are still finishing their Master’s or just started working. 23.4% has a WO master, which can also account for people who just started their career.
The last socio-demographic variable to analyze is living environment. Most respondents live in cities or big cities, as almost half of them lives in big cities of more than 200.000 inhabitants (46.5%) and 29.6% lives in cities of between 100.000 and 199.000 inhabitants. The smallest group lives in small cities of 50.000 to 99.000 inhabitants (9.6%), and the remaining 14.2% lives in villages with less than 50.000 inhabitants. The fact that almost 75% of the respondents lives in (big) cities may also be explained by the combination of age and education level: research by a branch of the Deutsche Bank shows that people in their 20s and 30s represent the biggest group of inhabitants in cities (LSE Cities, s.d.).
Table 1: Socio-demographic results
Gender Male 32.2%
Living environment
Big city (200.000> inhabitants) 46.5% Female 67.8% City (100.000-199.000 inhabitants) 29.6%
Age category
<19 1.1% Small city (50.000-99.000 inhabitants) 9.6%
20-29 54% Village (<50.000 inhabitants) 14.2%
30-39 22.6%
Education
Not a high school graduate 0.4%
40-49 7.3% High school graduate 3.8%
50-59 9.6% MBO degree 6.5%
60> 5.4% HBO/WO Bachelor degree 60.5%
Income category
<€30.000 gross p/y 52.1% WO Master degree 23.4%
€34.000-€40.000
gross p/y 20.3% Advanced degree beyond a Master’s degree 4.6%
€40.000> gross p/y 18.8% I would rather not say 0.8%
3.5 Variables
This section gives more detailed information about the variables applicable to this research which were already briefly presented in the conceptual model in Figure 2. The independent variable of this study is ecological consciousness, and willingness to pay for eco-friendly products represents the dependent variable. The model also contains a mediator, being packaging preferences. Additionally, explorative research was done for product type. The socio-demographics, which are already discussed in the previous section, were added to check for differences in the results based on differences in the area of personal characteristics.
3.5.1 Willingness to pay for eco-friendly products
The dependent variable aims to find out if and to what extent the respondent is willing to pay for eco-friendly products. This is measured with four Likest-scale questions and one open question (measurement level is ordinal). All questions in some way measure if the respondent takes environmental consequences into account when doing groceries, and if (and how much) he or she would be willing to pay extra for eco-friendly products. The questions, measured on a Likert scale ranging from 1 (strongly agree) to 5 (strongly disagree), stem from previous research by Roberts (1996; α = 0.96), Ishaswini and Datta (2011; α = 0.64) and Finisterra do Paço and Barata Raposo (2010; α = 0.93).
3.5.2 Ecological consciousness
This independent variable examines how high the level of the respondent’s ecological consciousness is, meaning how ‘green’ he or she is and lives, or how much the respondent is aware of and acts on environmental problems. Many studies stated that the typical ecologically conscious consumer has a high income, an above average educational level and a prestigious occupation (Roberts, 1996). This variable is measured with six questions on a Likert scale ranging from 1 (strongly agree) to 5 (strongly disagree) asking about their environmental concerns in general and how they act on this. Part of the respondents started these questions by reading an article about package-free supermarkets; the remaining respondents directly started with the questions. The questions to measure this variable come from previous research by Roberts (1996; α = 0.96) and Finisterra do Paço and Barata Raposo (2010; α = 0.93).
3.5.3 Packaging preferences
Packaging differs per food category and specific food item. It is assumed that packaging preferences mediate between ecological consciousness and willingness to pay for eco-friendly products. The questionnaire included six questions for this variable to determine if the respondents consume packaged food regularly, if they try to avoid plastic bags, if they check labels to see if products are environmentally safe, and if they prefer refillable containers. It also measures if respondents would be positive towards package-free supermarkets and how much extra travel time they would be willing to accept. Three of the questions measuring packaging preferences come from previous studies by Ishaswini and Datta (2011; α = 0.64) and Finisterra do Paço and Barata Raposo (2010; α = 0.93) and are measured on a Likert scale ranging from 1 (strongly agree) to 5 (strongly disagree). The fourth question comes from a qualitative study by Hume (2010). The last two questions are self-constructed.
3.5.4 Product type
An additional explorative study was done to research if respondents have different packaging preferences per product type. This variable is measured with a matrix especially designed for this study. Ten durable and non-durable food products that are commonly found in Dutch households are presented on the y-ax, being: apples, milk, beans, bread, chicken filet, muesli, pasta sauce, rice, chocolate cookies and salt. The x-ax shows six packaging options of which the majority is commonly found in Dutch supermarkets, being: no packaging, own brought packaging, glass, metal, plastic, and paper or paperboard. Respondents were asked to rank their preferred packaging material per product (from 1 = favorite to 6 = least favorite), and if they thought that a particular packaging material did not apply, they could leave it blank.
The goal of this matrix was to find out where sustainability improvement possibilities were to be found. Outcomes also show for which product types consumers are acceptant towards no packaging or own brought packaging, which is important information for deciding whether or not the Dutch market is ready for a package-free supermarket.
3.5.5 Gender
According to Laroche et al. (2001), female consumers are more likely to be willing to pay for environmentally friendly products. To check for this, gender (nominal variable) is one of the control variables.
3.5.6 Age category
Previous research suggests that youngsters are more willing to accept new and innovative ideas (Ottman et al., 2006; Tai and Tam, 1997), which could result in them being more open towards green packaging alternatives or package-free supermarkets. However, Pickett-Baker and Ozaki (2008) found that age had minimal influence on ecological beliefs. As studies are not on the same page about this subject, age (measured with categories, ordinal level) is included as a control variable.
3.5.7 Educational level
Educational level (ordinal measurement level) is interesting to add as a control variable to see if and how education influences consumers’ opinion about environmental issues and if, or to what extent, they are willing to make a change. Could it be, for example, that the higher educated someone is, the more that person is aware of environmental issues and the more he/she feels responsible for playing a role in solving these issues?
3.5.8 Income category
Researching whether or not higher income consumers have different levels of ecological consciousness and willingness to pay for eco-friendly products than lower income groups could give relevant information about possible targeted consumers and adaptiveness of eco-friendly products. Therefore income level (measured on ordinal level) is included as a control variable.
3.5.9 Living environment
Most of the respondents live in cities (23.4%) and big cities (50.6%). This is interesting information because research shows that inhabitants of big cities have a lower carbon footprint than national averages (Vaughan, 2009). This raises the question if city dwellers are also more acceptant towards eco-friendly solutions like package-free supermarkets. Regularly, cities already have a greater offer of environmentally friendlier alternatives like
organic supermarkets and marketplaces selling local products. Therefore it is possible that they are already more used to eco-friendly alternatives and more willing to pay extra for eco-friendly products. Because of this reason, living environment (ordinal measurement level) is considered a valuable control variable.
4. Analysis and results
This chapter displays the analysis conducted in the Statistical software Package for Social Sciences (SPSS) version 24, followed by its results. The modeling tool PROCESS was also used for the mediation analysis. First, some preliminary tests were run of which the results are presented in the first part of this chapter. After this, the three hypotheses were tested.
4.1 Preliminary tests
This section starts with an explanation of the missing data of this research. Usually recoding would be part of the preliminary tests too, but as the questionnaire did not include any reverse questions, recoding was not necessary. Thereafter, the outcome of the randomization function is given which divides the sample into two groups: the treatment group and the control group. This is followed by descriptive statistics testing for normality, a reliability test, and a correlation analysis.
4.1.1 Missing values
Before running frequency and descriptive tests, the data was checked for missing values. In case the researcher does not handle missing values correctly, the final results may be inaccurate. A total of 297 respondents were reported at the moment of closing the online questionnaire, but 36 subjects filled out less than half of the survey. Therefore their output was excluded from the analysis and the tests were run with the sample of 261 subjects. After this exclusion, only a few missing values (<5%) were detected. The imputation technique that was used to deal with this missing data was mean substitution.
4.1.2 Randomization
As mentioned before, the respondents were randomly divided into two groups by the use of a randomization tool in Qualtrics. The respondents were assigned to condition one (reading a news article about a package-free supermarket) or condition two (not reading the news article). Of the 261 respondents who participated in the experiment, the treatment group consisted of 113 respondents who started the questionnaire by reading a news article (43.3%), and the control group was formed of 148 respondents who started without a news article (56.7%). Further investigating the previously mentioned missing
values revealed that 91.6% comes from non-response in the control group. Survey tool Qualtrics offers the option to randomize respondents evenly, but that was not applied for this research. Therefore the division between respondents in the treatment and control group was not even.
It took an average of 103 minutes to fill out the survey. However, 28 extreme values were detected that elevate the average time excessively. Of these 28, fifteen respondents came from the treatment group and thirteen from the control group. The extreme values might be due to the respondents getting distracted and therefore taking a break from filling out the survey. Excluding these outliers brought the average back to 7 minutes. Making a distinction between the treatment group and the control group shows no significant difference in average duration time. This is surprising as the treatment group started the survey by reading an article, which generally should raise the average duration time. 4.1.3 Descriptive statistics
Table 2 presents the descriptive statistics of the variables with the following indicators: N (number of items), Range, Minimum, Maximum, Skewness and Kurtosis. The Mean and Standard Deviations are shown in the correlation matrix (Table 5).
The number of items of the variables ecological consciousness (EcCon), willingness to pay for eco-friendly products (WillPay) and packaging preferences (PackPref) is 261, referring to the number of respondents. The range, which is the difference between the largest (maximum of 5.00) and the smallest (minimum of 1.00) values of the numeric variables, is 4.00.
Table 2: Descriptive statistics
N Range Minimum Maximum
Skewness Kurtosis Statistic Std. Error Statistic Std. Error EcCon 261 4.00 1.00 5.00 .49 .15 -.16 .30 WillPay 261 4.00 1.00 5.00 .32 .15 -.15 .30 PackPref 261 4.00 1.00 5.00 .30 .15 -.29 .30
Skewness and kurtosis are calculated to test if the distribution deviates from normal. The skewness gives information about the (lack of) symmetry and kurtosis about the ‘pointiness' of the distribution (Field, 2012). EcCon, WillPay and PackPref show to be moderately skewed right for both samples. When looking at kurtosis, the samples show to be slightly platykurtic, as the tails are a bit thinner than in the case of normal distribution. As the values of skewness are close to 0 and the values of kurtosis smaller than 3, there is no sign of outliers. The fact that there are no extreme values detected in skewness and kurtosis is due to the use of Likert scale in the questionnaire, which restricts a set range between 1 and 5.
Figure 3 presents six histograms showing the distribution of the data. The black curve represents what normal distribution looks like; the bars represent the data. A distinction between the treatment and control group is made to research if the two differ. The histograms confirm that, as previously stated, the distributions slightly skew to the right. The distribution of the treatment group does not seem to differ much from the distribution of the control group, except for a small difference for the variable willingness to pay for eco-friendly products. Willingness to pay for eco-friendly products is more normally distributed in the control group than in the treatment group. Appendix C offers larger representations of the histograms, making them easier to read.
Figure 3: Histograms of distribution
Other ran tests to check for normal distribution are Kolmogorov-Smirnov and Shapiro-Wilk (see the output in Table 3). When these tests are significant (p<.05) this means that the null hypothesis of normal distribution is rejected. Since the output of the data shows that all three variables have p<.05, the null hypothesis is rejected and non-normal distribution is assumed. Non-normal distribution means that the estimated coefficients in the Ordinary Least Scales (OLS) regression are not the best unbiased linear estimators. Because of this, a Logit analysis was done next to the OLS analysis for comparison in order to check if the results are qualitatively equal (see paragraph 4.2.2.).
4.1.4 Reliability test
To test for reliability, Chronbach’s Alpha (α) was analyzed for the items ecological consciousness, willingness to pay for eco-friendly products and packaging preferences. It is important to check the level of reliability of the variables to see if the data generates consistent findings. The test checked for consistency of measurements and whether or not all questions should be included in the analysis. The results are shown in Table 4. According to George and Mallery (2003, p. 231), the following rules of thumb can be used when interpreting Chronbach’s Alpha: ‘_ > .9 – Excellent, _ > .8 – Good, _ > .7 –
Acceptable, _ > .6 – Questionable, _ > .5 – Poor, and _ < .4 – Unacceptable’.
The test showed that ecological consciousness and willingness to pay for eco-friendly products have a Chronbach’s Alpha of >.7, indicating high level of internal consistency. Packaging preferences, however, initially scored .544. After deleting one item (I consume packaged food... daily/weekly/monthly) the Chronbach’s Alpha was brought up to .677 (justified because the difference is >.10), which is close enough to the suggested .7 to say that it is reliable. After deleting one item, the corrected item-total correlations shows that all
Table 3: Descriptive statistics: Kolmogorov-Smirnova & Shapiro-Wilk
Kolmogorov-Smirnova Shapiro-Wilk
Statistic df Sig. Statistic df Sig. EcCon .10 261 .000 .97 261 .000 WillPay .10 261 .000 .98 261 .001 PackPref .09 261 .000 .97 261 .000 a. Lilliefors Significance Correction
items have good correlation with the total score of the scale as they are all above the threshold level of >.30.
4.1.5 Correlation analysis
Table 5 presents an overview of the descriptive statistics, correlations and scale reliabilities. Pearson’s correlation coefficient (r) is useful as it measures the correlation between variables and detects multicollinearity problems (Duzan and Shariff, 2015). The coefficient reaches between -1 and +1 and gives information about the strength and direction of relationships between variables. The table shows that there is a strong positive relationship between ecological consciousness and willingness to pay for eco-friendly products (r = .74, p < .01), and also between ecological consciousness and packaging preferences (r = .73, p < .01). Moreover, there also appears to be a strong positive relationship between willingness to pay for eco-friendly products and packaging preferences (r = .66, p < .01).
Looking at the control variables reveals that age category has weak negative relationships with all three variables ecological consciousness, willingness to pay for eco-friendly products and packaging preferences (r = -.35, p < .01; r = -.35, p < .01; r = -.33, p < .01 respectively), meaning that if the respondent is younger, he/she has higher ecological consciousness, higher packaging preferences and is willing to pay more for eco-friendly products. Additionally, weak to very weak negative relationships between income and ecological consciousness, willingness to pay for eco-friendly products and packaging preferences are detected (r = -.21, p < .01; r = -.28, p < .01; r = -.16, p < .05 respectively). This stated correlation indicates that the lower the respondent’s income is, the higher the ecological consciousness and packaging preferences are and the more he/she is willing to pay for eco-friendly products. The last control variable showing correlation with the three variables is living environment, which has a very weak negative relationship with all
Table 4: Reliability test
Variable Chronbach’s Alpha
EcCon 0.819 WillPay 0.742 PackPref 0.677
variables (r = -.15, p < .05; r = -.12, p < .05; r = -.17, p < .01 respectively). This means that the respondents living in smaller living environments (1 = big cities with >500.000 inhabitants, 4 = village with <50.000 inhabitants) are less ecologically conscious, have lower packaging preferences and are less willing to pay for eco-friendly products.
4.2 Results
While the previous section covered the preliminary tests, this section focuses on testing the three hypotheses. This is done by performing an independent t-test, a regression analysis and a mediating analysis with PROCESS Model 4. Additionally, the product type -packaging preferences matrix is analyzed.
4.2.1 Independent t-test
To compare the average outcomes of the three variables for the treatment group and the control group, the independent t-test is performed. Besides giving information about how
Table 5: Means, Standard Deviations, Correlations and Reliabilities
Variables Number of items M SD 1 2 3 4 5 6 7 8
1. Gender (0=female, 1=male) 1 0.68 .47 -2. Age category 1 2.86 1.23 .07 -3. Education 1 4.20 .84 -.02 -.01 -4. Income 1 1.84 1.02 .09 .44** .13* -5. Living environment 1 1.92 1.06 .03 .35** -.11 .26** -6. Ecological consciousness 6 2.40 .74 -.02 -.35** -.04 -.21** -.15* (.82) 7. Willingness to
pay for eco-friendly products
5 2.53 .82 -.08 -.35** -.11 -.28** -.12* .74** (.74)
8. Packaging
preferences 6 2.68 .90 -.11 -.33** -.03 -.16* -.17** .73** .66** (.68)
**. Correlation is significant at the 0.01 level (2-tailed) *. Correlation is significant at the 0.05 level (2-tailed)
the two groups differ, the test also gives information about how significant those differences are. Table 6 shows the difference in means for the three variables. Minimal, negligible differences between the treatment group and the control group came forward. In other words, exposing consumers to media output concerning packaging alternatives does not make them more willing to pay for eco-friendly products. Therefore, H2 is rejected.
4.2.2 Regression analysis
A hierarchical multiple regression was performed to examine the ability of ecological consciousness to predict levels of willingness to pay for eco-friendly products, after controlling for gender, age category, education level, income category and living situation. Additionally, the variable packaging preferences was included to check for a mediating effect. The results are shown in Table 7.
In the first step of the hierarchical multiple regression, five predictors were entered: the control variables. The model was statistically significant F (5, 254) = 9.68; p < .001 and explained 16% of the variance in willingness to pay for eco-friendly products. After the introduction of ecological consciousness at Step 2, the total variance explained by the model as a whole became 56%, F (6, 253) = 53.93; p < .001. This introduction explained additional 55% variance in willingness to pay for eco-friendly products, after controlling for the five control variables (R2 Change = .55; F (1, 253) = 231.24; p < .001). Three predictor variables were statistically significant in the second model, with ecological consciousness recording a higher Beta value (β = .68, p < .001) than gender (β = -.07, p < .01) and income (β = -.11, p < .01). When someone’s ecological consciousness increases by one, willingness to pay for eco-friendly products increases by 0.68. Additionally, when a
Table 6: Independent t-test
Sample group N Mean
EcCon Treatment 113 2.41 Control 148 2.40 WillPay Treatment 113 2.56 Control 148 2.51 PackPref Treatment 113 2.68 Control 148 2.67
