A price increase as a means to reduce the negative consequences of meat consumption : does it work?

A price increase as a means to reduce the negative

consequences of meat consumption: Does it work?

2

A price increase as a means to reduce the negative

consequences of meat consumption: Does it work?

Lotus Arian | 10760016 Bachelor thesis | Final version

University | University of Amsterdam Faculty | Amsterdam Business School Date | 23th _{of June 2018}

Supervisor | Lita Napitupulu Word count | 12468

3 Statement of originality

This document is written by Student Lotus Arian 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.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

4 Acknowledgements

Although this thesis is entirely written by Arian, acknowledgements are in place. First of all, miss Lita Napitupulu must be thanked for her supervision and guidance during the process of writing this thesis. Second of all, the research would not have been conducted without the cooperation of the participants who filled out the survey.

5 Abstract

Climate change is a very important issue for present and future generations. Scholars and politicians are searching for solutions and ways to reduce the drivers of climate change. The meat industry is one of the key players of this problem. Meat consumption and production also causes numerous health issues. However, surprisingly, meat consumption is still increasing. The consumption of meat is connected to a certain lifestyle and part of culture, this makes it very difficult to change the behaviour of consumers towards sustainable products. In this thesis we examined to what extend a price increase can change this behaviour. We analysed three surveys, one with a 125 percent price increase of meat products, one with a 200 percent increase and one with a 300 percent increase. The only main effect that we found was when there is a 125 percent price increase. The willingness to buy the more expensive meat products decreased the willingness to buy a sustainable alternative. Even a price increase of 300 percent did not change consumers’ behaviour towards sustainable product choices significantly. Hence we found that in order to change the behaviour of meat consumers more factors play a role than the price.

More research should be done towards discovering which other factors determine why people wish to eat meat, despite the negative consequence concerning meat production and meat consumption.

6 Abstract

Title: Limiting climate change by reducing meat consumption.

Research question: To what extent does the price increase of meat products change consumers’ behaviour towards more sustainable product choices? Supervisor: Lita Napitupulu

Author: Lotus Yoy Arian

Background: Climate change is currently one of the most important problems for humans and should be as limited as possible, the reduction of meat consumption is a very important factor in this limitation.

Purpose: The purpose of this thesis is to find out if a price increase can change Consumers’ behaviour towards more sustainable choices.

Method: This thesis has a deductive research approach with a quantitative

method and an experimental research design. The data of this thesis has been be conducted through surveys.

Approach: Collecting 180 filled out surveys and analysing them through the statistical program of SPSS.

Conclusion: Even a 300 percent price increase of meat products has not enough effect, in order to change the behaviour of meat eating consumers towards sustainable alternatives.

Recommendations: More research must be done to find factors that can change behaviour of meat eating consumers. More specifically research concerning an information campaign, to make consumers aware of the consequences of meat.

Keywords: Meat consumption, price elasticity, willingness to buy, willingness to switch.

7 Table of content Statement of originality P.3 Acknowledgements P.4 Abstract P.5 Chapter 1 Introduction P.8

Chapter 2 Theoretical Framework P.13

2.1 Meat industry P.13 2.2 Price elasticity P.13 2.3 Consumer behaviour P.14 2.3.1 Sustainable incentives P.14 2.3.2 Willingness to buy P.14 2.3.3 Willingness to switch P.14 2.3.4 Gender P.16 2.3.5 Income P.16 2.3.6 Education level P.16 2.3.7 Pro-sustainability P.17

2.3.8 Emotional switching costs P.17

2.4 Conceptual model P.18

Chapter 3 Research design P.19

3.1 Variables P.19

3.2 Survey P.20

3.3 Analysis P.21

3.4 Strengths and limitations P.21

Chapter 4 Results P.21 4.1 125 percent P.21 4.2 200 percent P.25 4.3 300 percent P.28 Chapter 5 Discussion P.31 5.1 Limitations P.31

5.2 Suggestions for further research P.31

5.3 Tax policy P.32

Chapter 6 Conclusion P.33

6.1 Summary P.33

6.2 Contribution of the study P.34

Chapter 7 References P.35

7.1 References P.35

8

Chapter 1 Introduction

Climate change has become the most important problem for the human race. The increasing temperatures, melting glaciers and ice caps, rising sea levels, shifting weather patterns and ocean currents threaten the very existence of us humans. The livestock sector plays an important part, since it is generating 18 percent of the CO2 equivalents, which make this sector a

bigger polluter than the transport sector (Steinfeld et al., 2006). The World Wildlife Fund predicts that if every person in the world consumes the same as the average European, we need the resources of three Earths in order to satisfy the demand (UNEP, 2005).

Worldwide demand for meat and substitute products such as milk and eggs increased substantially over the past decades. This is caused especially by growing populations, rising incomes and urbanisation in developing countries (Marcia, 2007; Steinfeld, Gerber, Wassenaar, Castel, & De Haan, 2006). We see that the per capita meat consumption has doubled in developing countries from 14 kg in 1980 to 28 kg in 2002. Naturally, the developed countries were already consuming a substantial and increasing amount of meat which caused the total meat supply to increase from 47 million tonnes of meat in 1980 to 137 million tonnes in 2002 and is expected to increase to 465 million tonnes in 2050 (Steinfeld et al., 2006). The highest amount of meat consumed every year comes from pork (15.8 kg per capita), second highest is poultry (13.6 kg per capita) and beef is third highest (9.6 kg per capita) (FAOSTAT, 2014). All these animals need to be fed, which uses a lot of resources. In order to produce one kilo of broiler chicken, 3500 litres of water is needed, mostly to produce their food. Cows need even more. For one kilo of beef, 11000 litres of water is needed, while only 500 liters is needed for a kilo potatoes (Tarrant, 1998). Europeans have been increasingly worried about environmental and health problems since the mid-1980s (Krystallis, & Chryssohoidis, 2005).

Consumers in the Western world have an increasingly large variety of products they can choose from. Meat remains a popular product. This is notable considering that the negative consequences for human welfare, animal welfare and the environment are increasingly well known. Politicians are also hesitant to steer people’s consumption patterns, because of the value of privacy and individual autonomy (Nordgren, 2012). Consumer’s unwillingness to change could be explained through the discrepancy of associations during the shopping process. On the one hand, people around the world are increasingly concerned about animal welfare (Grandin, 2010; Seng, & Laporte, 2005). On the other hand, when eating, preparing or buying food, consumers tend to think about price, health or their preferences. They often do not think about the living conditions of the animal (Korzen & Lassen, 2010; Latvala et al., 2012).

Over the past few decades there have been a couple of major scandals concerning the production and consumption of meat. Escherichia coli (or E. coli), illegal growth promoters, and variants of Creutzfeldt Jakob disease have been found in beef. Antibiotics are found in pork and salmonella is found in chickens (Tarrant, 1998). These medicine and growth enhancers, along with other illnesses caused by food, kill 9000 Americans every year (Tarrant, 1998). Moreover, one third of human protein intake is provided by livestock products which is contributing to obesity (Steinfeld et al., 2006) and other so called prosperity illnesses such as a stroke, bowel cancer, lung cancer (Poppe et al., 2018), kidney cancer (Faramawi, Johnson, Fry, Sall, & Yi, 2007) diabetes mellitus and coronary heart disease (Micha, Wallace, & Mozaffarian, 2010). In total, food related cancer accounts for 30 percent of all cancers in developed countries (UNEP, 2005). Meat products were strongly associated with a healthy diet, in spite of the relation with serious illnesses and alternative advise of nutritionists (Vines, 1997; Tarrant, 1998). It seems that that consumers find the nutritional properties of meat more important than the connected safety

9 concerns (Font-i-Furnols, & Guerrero, 2014; Verbeke, Pérez-Cueto, de Barcellos, Krystallis, & Grunert, 2010).

Furthermore, animal products are the cause for certain health issues (Eijsackers et al., 2010). Meat consumption does not only affect the consumers negatively but, like second-hand smoking, also influences the surroundings of the animal farms negatively. Such as the Q- fever, which is responsible for 74 deaths in the Netherlands between 2007 and 2010 (RIVM, 2017). The societal costs of this Q-fever is estimated between 250 million and 600 million euro (Tempelman, Prins, & Koopmans, 2011). Unsurprisingly, health benefits is the most common reasons for the deviating consumers who reduce their meat consumption (Beardsworth & Keil, 1991; Richardson, MacFie, & Shepherd, 1994).

We see that health problems are not only a social problem, but that it is also an economic disadvantage. In the United States alone, more than $400 billion is spend on medical costs for chronic diseases that are linked to human diet (diabetes, obesity and cardiovascular disease). A healthy diet can be promoted by producing more functional foods and thus prevent a lot of human discomfort and health costs (Decker, & Park, 2010). Functional foods are products that increase your health. The market value of functional foods in Europe was approximately 15 billion dollars in 2006 (Zhang, Xiao, Samaraweera, Lee, & Ahn, 2010). Additionally, the Netherlands is one of the major countries for the functional food market.

Apart from livestock’s negative impact on humans, it has a considerable impact on the world’s land, biodiversity and water resources. Moreover, it contributes to climate change. Since approximately 30 percent of the ice-free terrestrial surface on the earth is needed for the animals and for the production of their food (Steinfeld et al., 2006).

Moreover, the land that is used for meat production is at the same time polluted by emitted nutrients, pathogens, organic matter, and drug residents, which infiltrate the ground water, which then pollutes rivers and seas (Poppe et al., 2018; Steinfeld et al., 2006). Hence, meat is the most ecologically burdensome and energy intensive food (Dagevos & Voordouw, 2013).

It is expected that in 2025, 64 percent of the human population lives in water-stressed basins. The livestock sector accounts for eight percent of the human water use globally, it is thus a key player according to Steinfeld et al. (2006).

Meat production is also responsible for the emission of greenhouse gases that cause climate change. Anthropogenic methane has 23 times the equivalent of CO2’s global warming

potential. The agricultural sector emits 37 percent of this methane. Furthermore, anthropogenic nitrous oxide has 296 times the equivalent of CO2’s global warming potential. The agricultural

sector emits 65 percent of this nitrous oxide, mainly from animal manure (Steinfeld et al., 2006). Considering all the negative consequences of meat production and consumption, We would like to discuss a number of potential theories for people’s increasing consumption of meat. Firstly, a dispersion of time between cause and effect. The negative effects of climate change caused by current generations are experienced by future generations. Thus the incentive is lacking for current generations to change their behaviour. Secondly, a dispersion of culture. If younger generations believe that consumers should behave responsible and in a sustainable manner it does not necessarily mean that older generations feel the same pressure. Thirdly, a lack of governance. The world lacks a global government that protects the rights and interests of future generations (Gardiner, 2006; Nordgren, 2012).

A big problem is the discrepancy between individual rationalism and collective rationalism. One way to overcome this problem is to change the rules through the means of

10 political steering (Gardiner, 2006; Singer, 2006). One way to bring this into practise is to raise the taxation on meat products. However, it is a challenge to convince people to decrease their animal based consumption. Because consumers are fixed on traditional values such as price, status and convenience (UNEP, 2005). In addition, many people find it acceptable to eat meat, and link it to a high standard of living. Moreover, if consumers become wealthier they consume more meat (Dagevos & Voordouw, 2013). This link could be partially based on the advertisements, which have been blamed to spread the Western lifestyle and thus substantial consumption in developing countries (UNEP, 2005).

Consumers buy increasingly more meat (Marcia, 2007; Steinfeld, Gerber, Wassenaar, Castel, & De Haan, 2006; Steinfeld et al., 2006), even though buying meat in public lowers your status, which we will elaborate on later (Griskevicius, Tybur & Van den Bergh, 2010), and there being increasingly more sustainable alternatives (Dietz, Ostrom, & Stern, 2003; Griskevicius, Tybur, & Van den Bergh, 2010). There is also still contradictory literature concerning meat. For instance, scholars have not yet decided if the price of meat is elastic or inelastic (Andreyeva, Long, & Brownell, 2010; Byrne et al., 1995; Chavas, 1983; Gallet, 2010; Karagiannis, Katranidis, & Velentzas, 1996; Mazzocchi, 2006; Purcell & Raunikar, 1971). The emotional switching costs of meat must also be taken into account. Consumers are not easily persuaded to switch, since people have more positive attitudes about their current behaviour and have negative attitudes about other types of behaviour that differs from their own (Font-i-Furnols & Guerrero, 2014; Povey, Wellens, & Conner, 2001). This makes it difficult for meat eating consumers to switch to sustainable alternatives.

The negative effects of the consumption of meat on the environment and the goal of the Paris climate agreement to reduce the emittance of greenhouse gases has led to a Dutch report from the Governmental Council for the Environment and Infrastructure, concerning considerably decreasing the meat consumption in the Netherlands. The council is predicting that the daily consumption of Dutch citizens will contain far less meat in 2050 (Poppe et al., 2018).

Scientists and citizens are increasingly concerned about the negative effects of the Dutch food production on the environment, human welfare, animal welfare and biodiversity. Moreover, according to the Paris climate agreement the greenhouse gases must be 95 percent lower in 2050 than in 1990 (Koelemeijer et al., 2017; Tweede Kamer, 2016). Presently, the meat industry accounts for 18 megatons of CO2 equivalents and thus contributes to 10 percent

of the greenhouse gases in the Netherlands annually. If all known technical measures are taken to minimize the CO2 emissions of the meat industry, the sector will still use the total allowed

amount of greenhouse gases (10 Megaton CO2-equivalents) according to the Paris Agreement. It

is obviously illogical to assume that other industries and transportations means will produce no CO2 in 2050. This is why the aforementioned council advises that the meat production must be

reduced in order to achieve the climate goals (Poppe et al., 2018). This statement is accepted by multiple other scholars by stating that decreasing the meat consumption is absolutely essential in order to meet the future food demand in a sustainable way (Dagevos & Voordouw, 2013; Roberts, 2009; Steinfeld et al., 2006; Troy & Kerry, 2010).

This entails that the CO2 reduction does not only depend on the meat industry and the

government, but also on the behaviour of the Dutch consumer. They should prepare for a dietary transition towards less animal protein (Poppe et al., 2018) Because consumers diets and the environmental consequences of consumption are closely linked. Climate change, water, land quality and biodiversity are affected by consumers’ consumption patterns, food choices, and eating habits. Although consumers might not realize it, their consumption, particularly the

11 animal based products, are environmentally harmful and we are reaching the planet’s limits (Dagevos & Voordouw, 2013).

The council advises that the amount of animal protein is reduced to a maximum of 40 percent of the total protein consumption in 2030 (Poppe et al., 2018). This can be stimulated by the government with the use of taxation (Dagevos & Voordouw, 2013; Griskevicius, Tybur, & Van den Bergh, 2010; Grunert, Bredahl, & Brunsø, 2004; Matsukawa, Asano, & Kakimoto, 2000; Odegard & Bergsma, 2012; Poppe et al., 2018; Steinfeld et al., 2006; UNEP, 2005; Van Vugt, Meertens, & Van Lange, 1995). By increasing the price of meat products, possibilities arise for new sustainable protein products made from beans, pods or seaweed. These plant based foods are much more energy- and environmental efficient than the animal based foods (Baroni, Cenci, Tettamanti, & Berati, 2007; Dagevos & Voordouw, 2013; Duchin, 2005; Marlow, Hayes, Soret, Carter, Schwab, & Sabaté 2009; McMichael, Powles, Butler, & Uauy, 2007; Pimentel & Pimentel, 2003; Pimentel & Pimentel, 2008; Tukker et al., 2011; Micha, Wallace, & Mozaffarian, 2010). The relation between an increasing gross domestic product (GDP) and increasing meat consumption is well known. It was presumed that a high GDP causes abundant meat consumption. However this relationship could be reversed, thus more meat consumption could lead to a higher GDP (FAO, 2001; Speedy, 2003).

Organisations globally are noticing the potential of environmentally friendly strategies and products (Cronin et al., 2011). Additionally we see that businesses are not only focussing on making a profit, but also on finding a balance between the multiple consequences their decisions have (Cronin et al., 2011), such as looking at the various stakeholders that are involved (Freeman 1984). A key component in communicating that the corporation makes green efforts is marketing (Cronin et al., 2011). These green efforts will cause additional costs, which is a financial disadvantage for the firm (Aupperle, Carroll, & Hatfield, 1985; Ullmann 1985; Vance 1975 ). Others say that these implementation costs are minimal, and that they are balanced by improved productivity and morale of employees (Moskowitz 1972; Parket and Eilbirt 1975). Unfortunately mixed results have been found concerning the short- and long term financial results, which makes it unclear if social responsibility has a positive effect on the financial performance in every firm (McGuire, Sundgren, & Schneeweis, 1988; Russo & Fouts, 1997; Waddock & Graves, 1997; Wright & Ferris, 1997). However, we do see that socially responsible initiatives have positive influences on the stock markets, which is an incentive for businesses to have a positive impact (Alexander & Buchholz, 1978; Cronin, Smith, Gleim, Ramirez, & Martinez, 2011).

The combination of all these theories illustrates that meat is a unique product and that it is different from other kinds of products. Therefore, a research that examines meat should regard it as a unique product instead of a regular product. Consequently the results of such research cannot be generalized beyond the meat restrictions and used for other types of products. Switching costs from meat to non-animal alternatives are not covered by the current literature. These switching costs are very important for the transition from meat consumption to meatless consumption. However, little is known about the effect of the emotional switching costs on meat, and how consumers can overcome them.

Surprisingly, all the new information about green products and the increasing awareness of consumers has not been followed by the marketing literature. The marketing literature has paid, peculiarly, little attention to green market strategies. This entails doing business while focusing on the effect of the strategy on: people, planet and, of course, profits

12 (Cronin, Smith, Gleim, Ramirez, & Martinez, 2011). Even though 75 percent of consumers state that they prefer environmental friendly products (Saad, 2006), and 40 percent of consumers are actually willing to buy sustainable products, green products are less than 4 percent of the global market share (UNEP, 2005; Luchs & Mooradian, 2012). This makes it clear that more marketing research is needed to be able to produce green products that consumers want to buy (Cronin, Smith, Gleim, Ramirez, & Martinez, 2011). Marketing can play an important role in developing the coordination between the consumer’s demands and the supply of the market (Hurley and Hult, 1998; Jaworski and Kohli, 1993; Slater and Narver, 2000). This dissonance between statements and behaviour is problematic for all research that uses surveys. Moreover, it shows that more incentives are necessary for consumers to actually buy green products, such as more information (UNEP, 2005) or a price increase. However, a price increase has no effect if the price of meat is inelastic. Unfortunately, scholars do not agree if the meat price is elastic or inelastic. We hope that our research provides new insights.

Moreover, no research has been done concerning the potential difference between the behavioural change for different kinds of meat, such as beef, pork and chicken. In addition, no study to date has investigated the conceptual model that this thesis is focussing on. Therefore, this thesis will explore if the price change of animal products changes the behaviour of the participants in the direction of sustainable products and if this relation is moderated by: gender, income, education level, pro-sustainable behaviour or social pressure.

This leads to the research question: To what extent does a price increase of meat products change consumers’ behaviour towards more sustainable product choices.

In the theoretical framework, relevant theories and studies that are related to the meat industry are explained. Based on theories and findings of previous studies, six hypotheses are derived. In the method section it is explained how the data is collected, analyzed, and which questions were asked in the survey. Thereafter, in the result section the outcomes are explained and the hypotheses are accepted or rejected. In the discussion section the results of the findings are explained by referring back to the literature review. Furthermore, in the discussion section, points of improvements and suggestions for further research are given. In the conclusion section a short summary of the study is given and the academic contribution of the study is explained.

13

Chapter 2 Theoretical Framework

In the theoretical framework of this thesis, the existing literature will be discussed concerning relevant topics for the research question. First of all the meat industry will be discussed, because it is important to know this business when conducting research about meat products. Second of all, the price elasticity of meat is discussed. Third of all, the background of the variables of the conceptual model are discussed.

2.1 Meat industry

In this thesis meat is defined as: the food that is made from animal tissue (Forrest, Aberle, Hedrick, Judge, & Merkel, 1975). In 2015, 4 million cows, almost 13 million pigs and 100 million chickens lived in the Netherlands solely for the production of their meat (Wageningen Economic Research (WecR), 2017). There are 27000 firms that produce cows (Centraal Bureau voor de Statistiek (CBS), 2017), 4500 firms produce pigs and 2000 firms produce chickens in 2016 (WecR, 2017).

Economic motives have driven Dutch owners of animal farms to expand their businesses. Between 2000 and 2016 the amount of animals has at least doubled for most farms (Poppe et al., 2018). The total added value of the agricultural industry is only 1,77 percent of the total national income, which is about 10.5 billion euro. The meat industry contributes only about a quarter of this amount, which is about 2,2 billion euro (Poppe et al., 2018). In 2017, 9600 people worked in the 280 abattoirs (poultry not included) in the Netherlands (CBS, 2017). The abattoirs had a turnover of almost 4 billion euro in 2015.

Historically, meat has been associated with power, strength and masculinity (Fiddes, 1991; Latvala et al., 2011; Twigg, 1983). Today, meat is portrayed as traditional and cultural food. Advertisements in the West include old-fashioned meat products on wooden platters and traditional butchers (Swatland, 2010). Moreover, consumers link eating meat to a high standard of living (Dagevos & Voordouw, 2013). This is partly due to the advertisements, which have been spreading the Western lifestyle and thus substantial consumption (UNEP, 2005). The packaging of meat products is also a part of the marketing strategies. It is not merely an efficient way of preserving the quality of the product but it increases the value of the product and presents information (Han, 2005; Lee, 2010). The most important factors of the packaging are: price, size, safety, and recyclability. Additionally, convenience, utility and design must also be taken into account (Duizer, Robertson, & Han, 2009; Lee, 2010).

2.2 Price elasticity

Scholars have looked at the price elasticity of meat. However there has been a lot of disagreement (Andreyeva, Long, & Brownell, 2010; Byrne et al., 1995; Gallet, 2010; Karagiannis, Katranidis, & Velentzas, 1996; Mazzocchi, 2006; Purcell & Raunikar, 1971). Price elasticity refers to the relative change in purchased quantity of the product as a result of an one percent price change. If change is below one percent then the price is inelastic. If the change is above one percent then the demand of the product is elastic (Andreyeva, Long, & Brownell, 2010). If the price is inelastic, research concerning different price increases would be irrelevant. Some scholars found that in case of environmental responsible purchases, the costs of these products are just a little influencing their choice or not at all (Tanner & Kast, 2003; Luchs & Mooradian, 2012). This suggests that the price for environmental products such as meat alternatives can be inelastic. Literature reports show that price elasticities for red meat are, surprisingly, both

14 inelastic and elastic (Andreyeva, Long, & Brownell, 2010; Byrne et al., 1995; Chavas, 1983; Gallet, 2010; Karagiannis, Katranidis, & Velentzas, 1996; Mazzocchi, 2006; Purcell & Raunikar, 1971). Andreyeva, Long and Brownell (2010) found that the prices of beef, pork and poultry are inelastic in the United States. However, Gallet (2010) found that the price of meat is elastic, with the price of beef and lamb are more price elastic than poultry. They thus suggest that policy makers should focus on beef taxation. The Dutch Governmental Council for the Environment and Infrastructure is determined to use taxation or price engorgement in order to change purchasing behaviour (Poppe et al., 2018). Therefore the government needs to know the price increase necessary to change Dutch consumers’ behaviour.

2.3 Consumer behaviour

Multiple public officials and scholars have stated the need to motivate consumers to engage in pro-environmental behaviour (Griskevicius, Tybur, & Van den Bergh, 2010). Meat has a negative image nowadays. This is due to the association of slaughter conditions, handling practices (Troy & Kerry, 2010), environmental issues (Povey et al., 2001), the presence of blood (Kubberød, Uelan, Rødbotten, Westad, & Risvisk, 2002) and ethical, moral, or religious concerns (Berndsen & Van der Pligt, 2005; Dwyer, 1991). However, these negative attitudes concerning meat production have little effect on the consumption behavior. An explanation is that the knowledge about meat comes from indirect sources, because the consumers rarely enter slaughterhouses or animal farms (Grunert, 2006). Consumers also underestimate the ecological consequences of the meat industry, even though they are concerned with it (Vanhonacker, Van Loo, Gellynck, & Verbeke, 2013). It is suggested that health is the real reason why some consumers change their consumption pattern, reduce their meat consumption or even stop eating meat (Latvala et al., 2012). Although others have found that not even health concerns can change consumers’ consumption patterns (Ares, Barreiro, Deliza, Giménez, & Gámbaro, 2010; Tuorila & Cardello, 2002; Verbeke, 2006).

2.3.2 Willingness to buy

Consumers make daily choices concerning which products they want to buy. Price is influencing which products and which quantities consumers buy (Andreyeva, Long, & Brownell, 2010; Epstein, Dearing, Paluch, Roemmich, & Cho, 2007; French, 2003; Herman, Harrison, Afifi, & Jenks, 2008; Horgen & Brownell, 2002; Powell & Chaloupka, 2009). Price is present in all the products of consumers and it gives information about the product and the buyer. The price of a product illustrates the amount a consumers must sacrifice in order to engage in the purchase transaction. This means that the price affects the purchase probability negatively (Lichtenstein, Ridgway, & Netemeyer, 1993). Consumers’ purchase choice is also influenced by their ethical concerns. They consider the positive qualities of the products, such as the qualities of being a green product, and the negative qualities, which can stimulate a consumer to boycott the product (De Pelsmacker, Driesen, & Rayp, 2005).

2.3.3 Willingness to switch

In this thesis the willingness to switch concerns the consumer’s willingness to switch from a meat product to a meat alternative. Willingness to switch can be separated in social and economic motives. In this section we will firstly discuss the social motives and secondly the economic motives.

15 Buying meat lowers your status when shopping in public. Contrarily, buying sustainable products or giving money to environmental charities helps to build a pro-social reputation (Griskevicius, Tybur, & Van den Bergh, 2010; Semmann, Krambeck, & Milinski, 2005; Wedekind & Braithwaite, 2002). Even though there are increasingly more sustainable options, consumers are still experiencing barriers to change (Dietz, Ostrom, & Stern, 2003; Griskevicius, Tybur, & Van den Bergh, 2010). All the negative consequences of meat should theoretically affect the meat consumption negatively but an increase in meat consumption is still visible (Steinfeld et al., 2006). A social effect may explain this behaviour, since people are hesitant to change their behaviour and switch to sustainable products (Griskevicius, Tybur, & Van den Bergh, 2010). Additionally, time pressure decreases the environmental consumption (Luchs & Mooradian, 2012; Tanner & Kast, 2003). Moreover, consumers link eating meat to a wealthy standard of living (Dagevos & Voordouw, 2013). However, the lack of information about the product and its production is seen, by some, as the biggest obstacle for consumers to buy sustainable products (UNEP, 2005). Thus an information campaign can also help consumers make more sustainable choices. Unfortunately, others suggest that this may result in little change, because even though a large number of people claim to be concerned about animal welfare, they do not seem to think about this when they are purchasing and consuming meat (Guerrero, Claret, Rodriguez, Hernández, & Dalmau, 2013).

Apart from the more powerful social motives, from an economic perspective it is suggested that economic motives are the main cause for conservation (Cone & Hayes, 1980; Geller, 1989). On of the most important marketplace cues is price. Price is a very complex construct and can have a positive effect (Bello Acebrón & Calvo Dopico, 2000; Erickson & Johansson, 1985; Tellis & Gaeth, 1990; Zeithaml, 1988). Scholars argue that price is, among others, used by consumers to indicate the quality of the product (Bello Acebrón & Calvo Dopico, 2000; Lichtenstein, Ridgway, & Netemeyer, 1993). Price can even have a direct negative effect and an indirect positive effect (via quality perception) on purchase intentions (Erickson & Johansson, 1985).

An important factor for firms is the willingness to switch. Because a firm wants customers to switch from products from competitors to their own products. However, consumers are not easily persuaded to switch. Considering that people have more positive attitudes about their own behaviour and they have negative attitudes about types of behaviour that differ from their own (Font-i-Furnols & Guerrero, 2014; Povey, Wellens, & Conner, 2001). In addition to the meat industry a new industry arose as a result of the increased production of meat alternatives. This industry is interesting for consumers who are environmentally conscious and want to apply this in their consumption pattern. 44 percent of Dutch citizens eat meat substitutes occasionally, and 12 percent at least once a week. Meat substitutes are mostly eaten by people under 40 years old (54 percent). Moreover, highly educated people eat meat substitutes more than two times as frequently than lowly educated people. All citizens say that they are open for the consumption of meat substitutes if they are tastier (60 percent) or cheaper (38 percent) (Poppe et al., 2018). Thus consumers are stating their wish for a substitute that is cheaper than a meat product. It is unclear if this price change does change the behaviour of consumers towards more sustainable choices. Others suggest that in contrary to the statement of these citizens, price makes no difference for the amount of consumption of sustainable products (Tanner & Kast, 2003; Luchs & Mooradian, 2012). Following the existing literature the first hypothesis is:

H1: If one is exposed to a higher price for animal products, one will be more willing to switch to an alternative, non-animal product.

16 2.3.4 Gender

Research on sustainable consumer behaviour has found a strong gender effect. This effect entails that it is more likely that a woman expresses her concerns about the environment and acts upon those concerns, than that a men expresses his concerns and acts upon them (Dietz, Kalof, & Stern, 2002; Dupont, 2004; Eagly, Diekman, Johannesen-Schmidt, & Koenig, 2004; Koos, 2011; Hunter, Hatch, & Johnson, 2004; Josiassen, Assaf, & Karpen, 2011; Loughland, Reid, Walker, & Petocz, 2003; Luchs & Mooradian, 2012; Mayer & Frantz, 2004; Zelezny, Chua, & Aldrich, 2000). Women are therefore more inclined to make real sustainable purchases (Luchs & Mooradian, 2012; Vitell, 2003). However, some scholars disagree, and find that gender makes no difference concerning ethical buying behaviour (MORI, 2000; Luchs & Mooradian, 2012 Sikula & Costa, 1994; Tsalikis & Ortiz-Buonafina, 1990). Thus gender is used as a moderating variable. This is why the second hypothesis is the following:

H2: When the gender is female, the willingness to switch from an animal based product to an alternative is higher.

2.3.5 Income

Scholars are not convinced whether a positive or negative relationship exists between income and sustainable products consumption. On the one hand, higher-income consumers have fewer constraints financially, in regard to what they buy (Josiassen, Assaf, & Karpen, 2011). On the other hand, high-income consumers do not have the time to consider their options. The value of time is higher for the high-income consumers than the low-income consumers (Murthi & Srinivasan, 1999). Thus they reduce the time spend on searching for information about the product. They prioritize time over choice, and this may lead to a higher degree of random buying (Shahir, 1974). Some scholars found that in case of environmental responsible purchases, the costs of these products are just a little influencing their choice or not at all (Tanner & Kast, 2003; Luchs & Mooradian, 2012). However, in general we see an increase in willingness to pay a higher price for environmental friendly products (Laroche, Bergeron, & Barbaro-Forleo, 2001). We can also see that income does not change the willingness to buy, but it does change which products are actually bought (Krystallis & Chryssohoidis, 2005). This is why the third hypothesis is the following:

H3: Different income levels moderate the relationship between the price increase and the willingness to switch from an animal based product to an alternative.

2.3.6 Education level

Multiple studies concluded that ethical consumers or consumers who believe that the outcome of their consumption should be positive, were citizens with a relatively high income, social status and a high education level (Carrigan & Attalla, 2001; De Pelsmacker, Driesen, & Rayp, 2005; Maignan & Ferrell, 2001; Roberts, 1996). In addition, internationally, education has been seen as one of the most important factors to affect the willingness to pay (Govindasamy & Italia, 1999).

Evidence exists that women who are highly educated eat meat less often than lowly educated women. For men, the opposite is true: men with a high education level eat meat more often than men who have a lower education level (Latvala et al, 2012). This is why the fourth hypothesis is the following:

17 H4: A high education level moderates the relationship between the price increase and the willingness to switch from an animal based product to an alternative.

2.3.7 Pro-sustainability

One’s status can increase by buying sustainable products, this is why pro-sustainable behaviour can be promoted by status competition (Griskevicius, Tybur, & Van den Bergh, 2010). The people who act upon their pro-environmental thoughts and purchase green products are considered to be altruistic. They care intrinsically about the well-being of the Earth and the human population. This is because green products generally cost more and have a lower quality than the conventional version. These altruists are signalling that they are willing and able to incur these costs for the benefit of others. This status encourages them to buy green products in public shops, when these products are more expensive than the products which are not green (Griskevicius, Tybur, & Van den Bergh, 2010). However, others have found that a way to make more people buy green products is to make them cheaper and more efficient (Matsukawa, Asano, & Kakimoto, 2000; Van Vugt, Meertens, & Van Lange, 1995). We choose pro-sustainability as a moderator. The fifth hypothesis is:

H5: The higher the pro-sustainable behaviour, the greater is the likelihood that one is willing to switch from animal based products to non-animal based products.

2.3.8 Emotional switching costs

Switching costs are commonly perceived as economically orientated (Morgan & Hunt, 1994; Yang & Peterson, 2004). However, switching costs can also have an emotional and psychological basis (Burnham, Frels, & Mahajan, 2003; Sharma & Patterson, 2000; Yang & Peterson, 2004). In this thesis we do not choose for the economical switching costs but for the emotional switching costs that are made when one purchases a non-animal based product instead of an animal based product. We define switching costs as: the emotional costs customers associate with the process of switching from animal based products to non-animal based products.

The switching costs for consumers include those that are learning, monetary, behavioural and search related (Burnham, Frels, & Mahajan, 2003; Yang & Peterson, 2004). These costs are rarely focussed on but they are important when consumers consider switching to an alternative (Burnham, Frels, & Mahajan, 2003). The search related costs (searching for price and quality information) are reduced because of the materialization of the internet as a marketplace (Bakos, 1997; Lynch & Ariely, 2000). Due to the use of the internet, previous theories concerning switching costs must undergo new investigations according to multiple managers (Yang & Peterson, 2004).

If the emotional switching costs are substantial, consumers are, naturally, likely to maintain their relationship with their current provider or producer (Jackson, 1985; Port, 1980). This means that if one is emotionally attached to meat, this can increase the likeliness of them remaining to consume animal products. Moreover if a consumer has a bond with an employee or a butcher shop, they will be less likely willing to switch to a non-animal based product, if we follow the reasoning of Burnham, Frels and Mahajan (2003).

Functional food has a healthy effect on the consumer as stated before (Zhang, Xiao, Samaraweera, Lee, & Ahn, 2010). We acknowledge that if the functional value is low then one can easily switch. Thus, if consumers are aware of the unhealthy qualities of meat,

18 it could increase their willingness to switch. Thus, we choose to use emotional switching costs as a moderator. Considering all of the above the sixth hypothesis is:

H6: The higher the emotional switching costs, the greater the likelihood that one is not willing to switch from animal based products to non-animal based products.

The variations in the cognitive processes of consumers are caused by important differences in, among others, income and gender (Cooil et al., 2007; Crask and Reynolds, 1978; Fisher & Dube´, 2005; Josiassen, Assaf, & Karpen, 2011; Lambert-Pandraud, Laurant, & Lapersonne, 2005; Meyers-Levy, 1988; Meyers-Levy & Maheswaran, 1991). Thus the relationship between the price increase and the willingness to switch could be moderated by income or gender (Josiassen, Assaf, & Karpen, 2011). In addition, a factor that is believed to affect willingness to pay is the education level (Carrigan & Attalla, 2001; De Pelsmacker, Driesen, & Rayp, 2005; Govindasamy & Italia, 1999; Maignan & Ferrell, 2001; Roberts, 1996). Therefore, the independent variable is willingness to buy, the dependent variable is willingness to switch, the moderators of the conceptual framework are: gender, income, education level, pro-sustainability, and emotional switching costs.

2.4 Conceptual model

In this paper willingness to switch was chosen as independent variable and willingness to buy as dependent variable. The willingness to buy will be conducted through different price increases for beef, pork and chicken. We choose gender, income, education, pro-sustainability and emotional switching costs as moderators. The different hypothesis and variables that are stated in this theoretical framework are visualized in a conceptual model (figure 1).

19

Chapter 3 Research design

The thesis started with describing the existing literature which determined the variables, this is considered a deductive approach (Bryman & Bell, 2011). The existing literature and the research gap are followed by the research question and the goal of the thesis (Berry & Otley, 2004; Saunders, Lewis, & Thornhill, 2009; Yin, 2012; Wahyuni, 2012). This research is based on a quantitative research strategy, whereby we can test hypotheses and measure variables (Håkansson, 2013). This method does need a large data set, which we had to collect, and use statistics in order to test the hypotheses (Håkansson, 2013). In addition, quantitative research must rely on validity and reliability in order to secure the generalisability and the replicability (Wahyuni, 2012). Validity concerns that the information gathered is a good reflection of the social phenomena that are being measured (Wahyuni, 2012). Moreover, reliability means that the measurements should be conducted in a consistent way (Wahyuni, 2012).

We used an experimental approach to answer our hypotheses. An experimental research method deals with variables, discovers relationships between variables, and studies the causes and effects between relationships (Håkansson, 2013). This is especially useful for this research because all variables can be kept constant while only one variable is manipulated. By doing this we can look at the effect of that specific variable (Håkansson, 2013). In this thesis the relationship that will be most important is the correlation between the independent and the dependent variables.

The research approach for this thesis is a deductive approach. This means that we proceed from theoretical data to an empirical study. Therefore we were able to test our hypotheses and falsify or verify them (Håkansson, 2013). This needed a quantitative method and a large data set, as stated before (Håkansson, 2013). The results were compared to the previous theories (Bryman & Bell, 2011). The outcome is a generalization (Håkansson, 2013).

The sample group for this research should be a good representation of the Dutch population. This means that a bias of gender, income and education level should be avoided. Because the research is about consumers’ behaviour we limited the participants by selecting just those that are able to buy their own groceries. This means that the participants must be 18 years or older. This selection is based on the criteria needed to answer the research question (Bloor & Wood, 2006; Wahyuni, 2012).

3.1 Variables

Figure 1 shows the independent variable, dependent variable and the moderators. The independent variable is willingness to buy. This variable contains the degree of willingness to buy meat products after a price increase. We used an eight-item scale to examine this. It is measured through the means of asking participants if they are accepting the price increase (125, 200 or 300 percent) of the meat products. A five-point Likert scale was used, in which the scope runs from (1) ‘I would totally not buy it’, until (5) ‘I would totally buy it’.

The dependent variable is willingness to switch. This variable contains the consumers’ willingness to switch from a meat product to a meatless substitute. We used a three-item scale to examine this. It is measured by asking the participants if they are willing to buy the cheaper meatless substitute instead of the more expensive meat product. A seven-point Likert scale was used, the scope runs from (1) ‘I would never switch a meat product for a meatless substitute product’, until (7) ‘I will buy the meatless substitute product every time’.

The moderators are: gender, income level, education level, pro-sustainability and emotional switching costs. Switching costs contain the emotional switching costs people

20 experience when switching from meat products to meatless alternatives. We used a one-item scale to examine it. Gender, education level and income level were all measured through a one-item scale. The participants were asked about their gender, education and income levels in the biographical questions via an appropriate scale. For gender, the options were: male, female and other. The scope of the income scale runs from (1) ‘Below 25000 euro annually’, until (4) ‘Above 65000 euro annually’. The scope of the education scale runs from (1) ‘Primary school’, until (8) ‘University’. Pro-sustainability consists of a person’s sustainable consciousness. We used a two-item scale to examine this. Participants were asked if they were already sustainably conscious and if they act on that by recycling their waste. Both items were measured through a 7-point Likert scale, in which the scope runs from (1) ‘Strongly agree’, until (7) ‘Strongly disagree’. Participants are asked if they feel a certain social pressure to eat meat and if they thus feel emotional switching costs to remain eating meat. A 7-point Likert scale was used, in which the scope runs from (1) ‘Strongly agree’, until (7) ‘Strongly disagree’.

3.2 Survey

The participants received an online survey which they had to fill out. This survey was developed through the use of the website www.qualtrics.com. The data was gathered between 20-5-2018 and 13-6-2018. A total of 180 participants completed the survey.

The survey starts with general questions about the participants’ demographic background: gender, income and education level. The participants were also asked if they eat a vegetarian diet. If this was indeed the case, we mentioned that we thanked them for their time and effort and asked them not to fill out the other questions. This is done because a vegetarian will answer that they will not buy any of the meat products. Furthermore, consumers are also asked if they, on personal grounds, do not eat certain types of meat, such as beef, pork, and/or chicken.

The following questions consider the participants‘ willingness to buy. The participants could choose to buy the animal product with an increased price or not. Three different surveys have been made in order to examine the effect of three different relative increases: 125 percent, 200 percent and 300 percent. According to the Food and Agriculture Organisation (2014) a price increase of 160 percent is needed to incorporate the external costs. However, Odegard and Bergsma (2012) state that in the Netherlands an increase of 106 to 121 percent is needed. This is why we started with an increase of 125 percent. The 200 percent and 300 percent are meant to stimulate consumers not to buy meat at all and are therefore also asked in the survey. The participants answered only one of the three different surveys, in order to see which relative price increase has the most effect. Additionally, the questions were separated into beef, pork and chicken products, in order to see if the willingness to buy these different animals has a different effect on the willingness to switch.

Furthermore, questions were asked concerning how often the participants eat meat, how much they enjoy it, and how much time they wish to spend on preparing their food. This last question is asked because preparing meat can take longer than preparing meatless alternatives. The following questions consider the pro-sustainable attitudes of the participants. These questions concern recycling practices as well as the participants’ pro-sustainable considerations.

The last questions concern the willingness to switch. The participants see the meat products from the willingness to buy questions with the increase price next to a new meatless substitute with the information and price. They are asked if they would switch a meat product

21 for a meatless substitute. This again is separated in different beef-, pork- and chicken- questions. All information from the participants will of course be recorded anonymously in order to stimulate the participants to answer truthfully (Wahyuni, 2012).

3.3 Analysis

After the data was collected it was first managed for it to be ready to be analysed (Boeije, 2010; Wahyuni, 2012). The data were analysed through the statistical program SPSS. SPSS was used to calculate the results from the sample and to establish the significance of these results (Håkansson, 2013). The data of the 29 participants who stated that they eat a vegetarian or vegan diet were deleted. The other data were checked for normality, reliability and validity.

3.4 Strengths and limitations

The strength of the thesis should be the quality assurance. This means that validity and reliability must be assured (Håkansson, 2013; Wahyuni, 2012). Thus, what is expected to be measured must be what is actually measured and the ways of measuring should be constant during the data collection period (Håkansson, 2013; Wahyuni, 2012). The dependability, thus the judgement about the correctness of the conclusions, will be conducted by my supervisor as is suggested by Håkansson (2013). Reporting the changes of the setting that occur during the data collection is very important for the dependability (Wahyuni, 2012). The confirmability, and thus the notion that the research has been done in good faith, and without a personal gain that could have affected the results, must be supervised by myself and my supervisor (Håkansson, 2013; Wahyuni, 2012). We made the data of this thesis transferable in order to assist other scholars with the results of my thesis (Håkansson, 2013: Lincoln & Guba, 1985; Wahyuni, 2012).

Chapter 4 Results

Three surveys were distributed, one with a price increase of 125 percent, one with a 200 percent price increase and one with a 300 percent price increase. We will describe the results in three separate segments.

4.1 125 percent price increase

The data of the 125 percent price increase survey comes from 49 participants. If there are less than 30 participants we have to make sure that the distribution is normal. Even though it is not necessary in this thesis, for we have more than 30 participants, the verification of normality is valuable. We accept that skewness outside [-1,1] indicates serious asymmetry. 22 male and 27 female participants filled out the survey. Skewness for gender is -0,212, therefore we can conclude that gender is not skewed. There are far less people who eat a vegetarian or vegan diet (only 5 participants) than there are participants who do eat meat: we see a skewness of 2,676. This is seriously skewed but this is to be expected, since the majority of Dutch citizens eat meat

(De Bakker & Dagevos, 2010). It is unfortunate that income and education are also skewed.

Income level with a skewness of 1,209 is just over the limit but education level is highly skewed with -2,129.

In conclusion, these 49 participants are evenly distributed between men and women. Next, there are only 5 participants who never eat meat. Furthermore, more than half of the participants (27) has an annual income of less than 25000 euro, this could be because many of the participants are students. Also, 90 percent (45) of the participants has an education level of

22 HBO or WO, which is to be expected when numerous students fill out the survey. The answers of the participants who eat a vegan or vegetarian diet are deleted. Therefore we have 44 surveys to analyse.

Reliability

The reliability and consistency are often calculated with the mathematical measurement Cronbach’s alpha. We analyse this for the willingness to buy (WTB) in three different parts, namely: beef, pork and chicken. We use the following rules of thumb to analyse the Cronbach’s alpha: a Cronbach’s alpha of 0,8 or higher shows good reliability, between 0,6 and 0,8 shows reasonable reliability and below 0,6 shows bad reliability.

For beef we see that the Cronbach's alpha is 0,739 for the three item scale, so this already shows reasonable reliability, and almost good reliability. However, if we delete the first question concerning a steak, the Cronbach’s alpha will be 0,809. The Cronbach's alpha is already reasonable and there is no extreme increase of the Cronbach’s alpha if we delete the first question. Furthermore, 30 percent of the information is lost if we delete this question, and this is valuable information for the research. This is why we will keep the scale at three items.

For pork we see that the Cronbach’s alpha is 0.767, which shows reasonable reliability. If we delete an item the Cronbach’s alpha will only decrease (0,692, 0,750, 0,624). So we will keep the scale at three items.

For chicken we see that the Cronbach’s alpha is 0,680, which shows reasonable reliability. Moreover, if we delete an item the Cronbach’s alpha will decrease (0,520, 0,520). So we will keep the scale at two items.

For willingness to switch (WTS) we see that the Cronbach’s alpha is 0,853, which shows good reliability. Moreover, if we delete an item, the Cronbach’s alpha will decrease (0,764, 0,806, 0,817). So we will keep the scale at three items.

Correlation

The correlation between the WTB and the WTS are shown in table 1 (descriptives and correlations), in the appendix. Unsurprisingly, we found a significant positive correlation between the items of WTB: Beef and pork r(44)= 0,608, p= 0,001, pork and chicken r(44)= 0,749, p= 0,000 and chicken and beef r(44)= 0,608, p= 0,001. Additionally, we found a moderate negative effect between WTB beef and WTS r(44)= -0,352, p= 0,072. We found a significant negative effect between WTB pork and WTS r(44)= -0,478, p= 0,012. We found a moderate negative effect between WTB chicken and WTS r(44)= -0,341, p= 0,081. And we found a significant negative effect between total WTB and WTS r(44)= -0,448, p= 0,019. Thus we have proof that WTB with a 125 percent price increase predicts a low WTS. Thus if participants are willing to pay a higher price for a meat product we predict that they are less likely willing to switch to an alternative.

Regression

For our research, a hierarchical regression was used to test our hypotheses. We created mean-centred variables for beef, pork, chicken, gender, income, education, pro-sustainability and emotional switching costs. We used these new variables to make the interaction variable, where we multiplied the WTB variables (beef, pork and chicken) with the mean-centred moderators. The mean-centred moderators and interactions were added one by one to the mean-centred WTB and mean-centred WTS in order to test the moderation effect. The results of the regression analysis of the 125 percent survey are shown in table 1 (regression).

23 Our second hypothesis concerned the moderating effect of gender on the relation between WTB and WTS. In table 1 we see that none of the animal items, beef pork or chicken, are significant. Therefore, the regression provides no support for the second hypothesis. Thus, we found no prove that gender has a moderating effect on the effect of WTB on WTS.

The third hypothesis concerned the moderating effect of the income level of the respondents. In table 1 (regression) we see that the chicken model is significant, with β = 1,280, t = 2,713, p <.05. Thus, we found prove that income level has a moderating effect on the effect of WTB chicken on WTS.

The fourth hypothesis concerned the moderating effect of the education level. In table 1 we see that the beef model is significant β = 0,471, t = 2,212, p <.05. Thus, we found prove that education level has a moderating effect on the effect of WTB beef on WTS.

The fifth hypothesis concerned the already pro-sustainable behaviour of participants and the sixth hypothesis concerned the influence of emotional switching costs. These models are not significant, which means that these models do not provide support for the fifth and sixth hypotheses. Hence, we found no prove that either pro-sustainable behaviour or emotional switching costs have a moderating effect on the effect of WTB on WTS.

25 4.2 200 percent price increase

The data of the 200 percent price increase survey comes from 56 participants. There are more than 30 participants, thus we do not have to make sure that the distribution is normal. However, the verification is valuable. As stated before, skewness outside [-1,1] indicates serious asymmetry. 27 male and 29 female participants filled out the survey. Skewness for gender is 0,053 therefor we can conclude that gender is not skewed. There are less people who eat a vegetarian or vegan diet (only 10 participants) than there are participants who do eat meat: we see a skewness of 1,697. This is seriously skewed but this is to be expected. The skewness of income is 0,891, this suggest that income is not skewed. It is unfortunate that education is again skewed, with a skewness of -2,672.

In conclusion, the 56 participants are evenly distributed between men and women. There are only 10 participants who never eat meat. Also, about half of the participants (26) has an annual income of less than 25000 euro, this could be because many of the participants are students. Furthermore, 82 percent (42) of the participants has an education level of HBO or WO, which is to be expected when numerous students fill out the survey. The answers of the participants who eat a vegan or vegetarian diet are deleted, which left us with 46 surveys to analyse.

Reliability

For the reliability we check the Cronbach’s alpha for the willingness to buy each of the products: beef, pork and chicken.

For beef we see that the Cronbach's alpha is 0,860 for the three item scale, so this already shows good reliability. If we delete an item, the Cronbach’s alpha will only decrease (0,849, 0,754, 0,800). Thus we will keep the three item scale.

For pork we see that the Cronbach’s alpha is 0,932, which is very high and shows good reliability. If we delete an item, the Cronbach’s alpha will only decrease (0,906, 0,930, 0,865). So we will keep the scale at three items.

For chicken we see that the Cronbach’s alpha is 0,674, which shows reasonable reliability. Moreover, if we delete an item, the Cronbach’s alpha will decrease (0,509, 0,509). So we will keep the scale at two items.

For willingness to switch we see that the Cronbach’s alpha is 0,907, which is excellent, and shows good reliability. Moreover, if we delete an item, the Cronbach’s alpha will decrease (0,836, 0,874, 0,890). So we will keep the scale at three items.

Correlation

The correlation between the WTB and the WTS are shown in table 2 (descriptives and correlations) in the appendix. Unsurprisingly, we found a significant positive correlation between the items of WTB beef and pork r(39)= 0,597, p= 0,000, and chicken and beef r(39)= 0,370, p= 0,021, but no significant correlation between pork and chicken r(39)= 0,297, p= 0,067, which is just a moderate correlation. Furthermore, we found a positive but no significant effect between WTB beef and WTS r(39)= 0,220, p= 0,178, which classifies as an interesting correlation. And a positive but no significant effect between WTB pork and WTS r(39)= 0,011, p= 0,946. We found a negative nonsignificant effect between WTB chicken and WTS r(39)= -0,178, p= 0,278, which classifies as an interesting correlation. We found no significant effect between total WTB and WTS r(39)= 0.031, p= 0,852. Thus we have no proof that WTB with a 200 percent price increase predicts WTS.

26 Regression

The results of the regression analysis of the 200 percent survey are shown in table 2 (regression). We see that no model is significant. Thus we have no support for the five hypotheses. Hence, we have no support that gender, income, education, pro-sustainable behaviour or emotional switching costs have a moderating effect on the effect of WTB on WTS.

However, we do see that, although the model is not significant, pro-sustainability does show significance for all three kinds of meat. We can conclude that it is difficult to change someone’s behaviour, and that a price increase of 200 percent alone will not have a significant effect. Nonetheless, behaviour perhaps can be changed by awareness and thus convince people to act in a more sustainable way.

28 4.3 300 percent price increase

The data of the 300 percent price increase survey comes from 65 participants. Since 65 is more than 30 we do not have to make sure that the distribution is normal. However, we will analyse the skewness. 21 male, 43 female and 1 other participant filled out the survey. Skewness for gender is -0,455 therefor we can conclude that gender is not skewed. There are less people who eat a vegetarian or vegan diet (14 participants) than there are participants who do eat meat: we see a skewness of 1,369. This is seriously skewed but this is to be expected. Income level is highly skewed with a skewness of 4,220. Education level is not skewed with a skewness of - 0,750.

The 65 participants are somewhat evenly distributed between men and women. There are 14 participants who never eat meat. A great amount of the participants (52) has an annual income of less than 25000 euro. Furthermore, 50 percent of the participants had a university level education and about 30 percent had a VMBO level or high school level. The answers of the participants who eat a vegan or vegetarian diet are deleted, which left us with 51 surveys to analyse.

Reliability

For the reliability we analyse the Cronbach’s alpha for the willingness to buy each of the three products: beef, pork and chicken.

For beef we see that the Cronbach's alpha is 0,776 for the three item scale, so this already shows reasonable reliability. If we delete an item, the Cronbach’s alpha will only decrease (0,773, 0,732, 0,572). Therefore we will keep the three item scale.

For pork we see that the Cronbach’s alpha is 0.918, which is very high and shows good reliability. If we delete an item, the Cronbach’s alpha will only decrease (0,880, 0,901, 0,863). So we will keep the scale at three items.

For chicken we see that the Cronbach’s alpha is 0,754, which shows reasonable reliability. Moreover, if we delete an item, the Cronbach’s alpha will decrease (0,606, 0,606). Therefore, we will keep the scale at two items.

For willingness to switch we see that the Cronbach’s alpha is 0,910, which is excellent and shows good reliability. Moreover, if we delete an item, the Cronbach’s alpha will decrease (0,843, 0,871, 0,898). So we will keep the scale at three items.

A summary of all the data from the WTB and WTS is presented in the appendix in table 1, 2 and 3 (descriptives and correlation), which includes the mean, standard deviation and Cronbach’s alpha.

Correlation

The correlation between the WTB and the WTS are shown in table 2 (descriptives and correlation) in the appendix. We found no significant correlation between the items of WTB beef and pork r(51)= 0,266, p= 0,077. However we found a significant correlation between chicken and beef r(51)= 0,307, p= 0,040, and between pork and chicken r(51)= 0,460, p= 0,001. Furthermore, we found a positive but no significant effect between WTB beef and WTS r(51)= 0,065, p= 0,670. And a positive but no significant effect between WTB pork and WTS r(51)= 0,117, p= 0,446. We found a negative non-significant effect between WTB chicken and WTS r(51)= -0,036, p= 0,814. We found no significant effect between total WTB and WTS r(51)= 0,079, p= 0,607. Thus we have no proof that WTB with a 300 percent price increase predicts WTS.

29 Regression

The results of the regression analysis of the 300 percent survey are shown in table 3. Our second hypothesis concerned the influence of gender on the relation between WTB and WTS. In table 3 (regression) we see that all the different models for the meat kinds are significant. Therefore, the regression part provides support for our second hypothesis for beef: β = 0,377, t = 2,508, p <.05, pork: β = 0,317, t = 2,151, p <.05, and chicken: β = 0,567, t = 4,027, p <.001. Thus, we have prove that gender has a moderating effect on the effect of WTB on WTS.

The third hypothesis concerned the influence of income level and the fourth hypothesis concerned the influence of the education level. These models, however, were not significant. Thus we have no prove that income level and education level have a moderating effect on the relation of WTB on WTS.

The fifth hypothesis concerned the already pro-sustainable behaviour of participants. We found that the model of beef was significant β = 0,406, t = 2,544, p <.05. The models for pork and chicken are significant with an alpha of ten percent. Thus, the model of beef provides prove for the hypothesis that pro-sustainable behaviour influences the relationship between WTB beef and WTS, with an alpha of five percent. Thus, we found prove that pro-sustainable behaviour has a moderating effect on the effect of WTB beef on WTS.

The sixth hypothesis concerned the influence of emotional switching costs. This model is not significant and therefore this model does not provide support for our hypothesis.

31

Chapter 5 Discussion

5.1 Limitations

A gap exists between what consumers say and their actual behaviour. One must be aware of this while writing a research paper, when asking participants questions though the means of a survey. This gap has been found by multiple scholars in multiple ways. Because of dynamic consumer perception, there is a difference between their perception and their behaviour (Troy & Kerry, 2010). People's values and intentions do not always correspond with their actions, and therefore there is a difference between stated preference and revealed preference (Reinders et al., 2013). This is due to habit, convenience, lack of knowledge and affordability (Poppe et al., 2018). In addition, when confronted with a choice, consumers rely on associations, emotions and rules of thumb to make a decision (Ingenbleek & Immink, 2011). This makes it difficult to draw a conclusion from any research concerning a survey. This is also visible with research that concerns environmental behaviour. An example is that 40 percent of consumers state that they are willing to buy green products, but just 4 percent is actually buying them (UNEP, 2005; Luchs & Mooradian, 2012; Makatouni, 2002). An explanation could be that even though the consumers claim that they are concerned about issues as animal welfare, they do not think about this when consuming and buying meat (Guerrero, Claret, Rodriguez, Hernández, & Dalmau, 2013). This behaviour results from the psycho-protective mechanism: intentional forgetting (MacLeod, 2013). This means that consumers are de-emphasizing on the threatening and unpleasant memories while consuming meat (Font-i-Furnols & Guerrero, 2014). This can also be observed when asking questions about organic products. This is why one must be cautious with positive conclusions of research relying on surveys (Padel & Foster, 2005).

In order to stimulate the participants to answer honestly. The results of the participants were processed anonymously. This could, however, also limit the research, since consumers can be stimulated by social pressure to buy sustainable products. Yet, this pressure is not present if the answers are given online and no one can connect them to the participants. Thus, participants could have made other choices if they were in a supermarket at the time instead of filling out an online survey.

5.2 Suggestions for further research

If consumers do not know how unfortunate the consequences of their habits are, chances are that they do not change their behaviour (Font-i-Furnols & Guerrero, 2014). Because, the lack of information about the product and its production is seen, by some, as the biggest obstacle for consumers to buy sustainable products (UNEP, 2005). Thus an information campaign can also help consumers make more sustainable choices. This means that more research must be done to determine what consumers know and how more information can be distributed in order to change consumers behaviour. In general, consumers tend to believe that meat is a healthy product and an important part of their diet (Verbeke et al., 2010). Again, more research must be done in order to examine which nutritional values of meat they believe to be healthy, and whether they know about the health issues concerning meat and that there are healthy substitute products available.

In the study of Latvala et al. (2012) they surveyed participants for the second time about their consumption pattern and they found that 52 percent of the participants stated that they changed their consumption pattern in the last few years. Perhaps a one time exposure to a price increase has no effect but a long term exposure will. It will be interesting to send the