Does counteradvertising affect the willingness to purchase animal products?

(1)

1

MSc Business Administration

Marketing Track

Does counteradvertising affect the willingness to purchase animal products?

Written by Arina Torop Submitted on August 28th_{, 2015}

Supervised and guided by Drs. A. (Anouar) el Haji

Amsterdam University Faculty of business and economics Master of Science in Business Administration

Marketing department

Student ID: 10694986

Email: arinatorop@gmail.com.

(2)

2

Statement of originality

This document is written by Arina Torop only. I declare to take full responsibility for the contents of this document. This statement is written to confirm that the content of this thesis is my own work, written between February and August of 2015 for the sole purpose of finalizing my MSc education. I declare that the intellectual content of this thesis is the product of my own work and that all the sources used in preparing this thesis have been acknowledged in the bibliography. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

Acknowledgments

This research was guided by Drs. A. (Anouar) el Haji from February until August 2015. Anouar

shared his experience and offered his support from the stage of deciding about a topic until the final phase of writing this thesis. His advice and constant feedback contributed valuable input that shaped and improved this research, and for that I am grateful.

(3)

3

Abstract

Eating animals dates back 10,000 years and is theorized to have had an evolutionary role in the cognitive development of the human species. Until this day it serves social, nutritious and hedonistic purposes across most cultures and regions. However, in recent decades eating animal products has become increasingly controversial. The rising awareness to the negative economic and humanitarian consequences of eating animal products in large quantities are causes of concern for individual consumers and governments alike. This research aims to find whether there are mechanisms that reduce willingness to buy animal products, and if so, to which extent. For this purpose, counteradvertisement is used on a package of fish to examine whether it reduces willingness to pay for it.

Key words:

Counteradvertisement, demarketing, willingness to pay, animal product consumption, health, environment, animal welfare.

(5)

5

List of variables

Independent variables IV

Type of treatment IV0

Gender IV1

Age IV2

(Meat eating) culture/region IV3

Education IV4

Psychological distance to animals: owning pets IV5

Ambivalence IV6

Psychological distance to animals : general attitudes IV7 Attitude towards eating animal products and the environment IV8 Attitude towards eating animal products and animal welfare IV9

Attitude towards eating animal products and health IV10

General attitudes

IV11 IV12 IV13

Strongest of the general attitudes IV14

Dependent variables DV

Willingness to pay before treatment DV1

Willingness to pay after treatment DV2

Total effect DV3

Control variables CV

Price sensitivity CV1

(6)

6

Chapter 1 Introduction

A study about eating habits in Great Britain published in 2013 found that 25% of British people have reduced meat consumption from one year to the subsequent year (Dibb & Fitzpatrick, 2014). A study published in 2015 by the Dutch Nutrition Centeri found that 55% of the adult Dutch population makes a choice to not eat meat for dinner for as much as 3 days a week, substituting meat with fish and eggs (Keuchenius & Lelij, 2015). The findings in both studies are surprising because of the traditional importance of eating meat in the developed world in general and in Europe in particular, as exhibited in the past 10,000 years (Smil, 2002; Westhoek, et al., 2011). The Netherlands and Great Britain are not anomalies. The phenomenon of reducing meat consumption is apparent in other western countries as well, and is becoming a frequently debated matter in the media (Bittman, 2012; Plumer, 2012; Barclay, 2012; Gunther, 2013). For example, initiatives such as “Meatless Monday” and “World Vegetarian Day” have become increasingly popular across many different countries, accompanied by celebrity endorsement and media attention (Jones, 2012). This trend is emerging in recent years after different parties such as consumers, academia, nutrition experts and governmental research organizations have casted doubt about the benefits of eating animal products in large quantities (Dibb & Fitzpatrick, 2014; Carrington, 2014; Vaughan, 2013; Wardrop, 2009; Jowit, 2008; Bailey, Froggatt, & Wellesley, 2014; Bajželj, et al., 2014; Hedenus, Wirsenius, & Johansson, 2014; Grunert, 2006; Kenyon & Barker, 1998; Verbeke & Viaene, 1999). The recurring arguments for reducing meat consumption are health concerns, the environment and animal welfare (Carrington, 2014; Bailey et al, 2014).

Health wise, the fat in animal products is a substantial contributor to elevated cholesterol levels, which in turn elevate the risk of cardiovascular complications such as stroke and heart attacks, heart and artery diseases and other cardiovascular impediments. (Johnson et al., 2006; Fung et al, 2001; Hu, et al., 2000; Friedman, 2015). Currently, cardiovascular diseases are considered the world’s largest health challenge; cardiovascular diseases have claimed 1,278,000 lives in 2008 in Europe and 17.5 million lives worldwide in 2012, creating an economic burden (prevention, treatment and loss of earnings) of 169 billion Euros spent annually in the European Union in 2006, and $393.5 billion spent in the United States annually in 2006, (Leal, et al, 2006).

(7)

7 In addition, the industrial farming supply chain has a negative impact on the environment, with 14% of global greenhouse-gas emission stemming directly from livestock production. The annual waste created in the in the US alone by the animal products manufacturing processes is 130 times the waste that the entire US population creates annually (Chellaney, 2015; Horrigan, Lawrence, & Walker, 2002). Lastly, throughout the animal product supply chain, animals are held in confinement, suffering from poor health and living conditions until their slaughter (Lindgreen & Hingley, 2003).

Previous research suggests that reducing animal product consumption by individual consumers even by a small fraction can have vast positive economic, environmental and humanitarian consequences for both the western and the developing world (Westhoek, et al., 2014; Hedenus, Wirsenius, & Johansson, 2014; Goodland, 1997; Rosell, et al, 2006; Slattery, et al., 1991; Daniel et al, 2011). In addition, tackling economic challenges through changes in consumers’ demand patterns is less politically dangerous than policy changes in worldwide food trade (Koningl, et al., 2008; McMichael, et al, 2007). To quote the conclusions of a report made by a UK food research organization: “While there is now strong evidence and growing interest and awareness of health, environmental sustainability and ethical reasons to reduce meat consumption, there is a lack of research to understand how best to achieve this.” (Dibb & Fitzpatrick, 2014, pp.28). Others wonder whether reducing meat consumption can even be done in the “meat loving western world” (Westhoek, et al., 2014, 2011).

The above mentioned motivations call for further research. That is why this research explores whether there is a way to reduce animal product consumption by individual consumers. For this matter, the tool of counteradvertisement can be used, similar to its usage in the tobacco and alcohol industries, as a mechanism to reduce animal product consumption. Meat consumption of an individual is difficult to measure and monitor, especially throughout time. Therefore the willingness to pay for animal products can be used as a measurable proxy for animal product consumption. This leads to the following research question: What effect does counteradvertisement have on the willingness to pay for animal products in a retail context?

To answer this research question, chapter 2 gives a literary review about meat consumption decisions and counteradvertising. Chapter 2 ends with a theoretical framework and hypothesizes to be tested. Chapter 3 elaborates on the method used in this research, and chapter 4 analyzes the results. The study ends with a conclusion chapter, which offers future research.

(8)

8

Chapter 2 Literature review

This chapter dissembles the conceptual themes in the research question. The first part of the literature review discusses factors that affect willingness to eat and buy animal products, while the second part reviews counteradvertisement. This chapter will present several hypothesis that will be tested in chapter 3 and 4 in order to answer the research question.

2.1 The historical context of eating animal products

Eating animal products such as beef, poultry and fish is a behavior exhibited throughout the geographic and historic span of mankind across all civilizations, ranging from 11,000 years ago until this day (Smil, 2002). Attaining high protein calories found in meat or fish is hard-wired into the human mind as a survival need (Smil, 2002). As malnutrition and famine were eradicated only by the middle of the 19th century, being in an economic and social position that enables one to attain a quality food such as meat, was, and to a certain extent still is, a sign of affluence and success in most societies (Smil, 2002; Grigg, 1999; Mintz & Bois, 2002).

In the 20th century, the world economy was growing alongside technical, technological and agricultural developments. This created a situation in which a financially abundant population demanded animal products for longer periods of time due to increased life spans. Therefore, supply was met by a sophisticated supply chain that resembles an assembly line of a factory. The result was that food in general and meat in particular became industrialized in the last 50 years (Grigg, 1999). Currently, meat consumption is at an all-time peak. Industrialized and western countries have relatively higher animal product consumption rates compared to the rest of the world. China, the US and the EU are the world’s biggest consumers of animal products (Linseisen, et al., 2009; Chellaney, 2015). In comparison, meat consumption in the United States is a 125 kilogram per person annually, while its only 3.2 kilogram per person annually in India (Chellaney, 2015). On an aggregated level, this pattern seems to persist, with a projected growth of animal product consumption, alongside a projected growth of the world population (Vasileska & Rechkoska, 2012; Aggidis, et al., 2013).

Although meat consumption is in an all-time peak, increasing concerns regarding eating large quantities of animal products have risen in recent decades as well.

(9)

9 The first point of criticism has to do with the adversary effect eating large quantities of animal fats has on people’s health. The most prominent examples are cancer and cardiovascular complications that claim millions of lives yearly, resulting in millions of dollars in health costs for governments (Kearney et al, 2005).

Secondly, the population of livestock grown for the meat industry is approximately 150 billion animals worldwide, meaning that the direct ecological footprint of the animal product supply chain and meat consumption is immense (Chellaney, 2015). This ecological challenge also translates to a humanitarian problem of using vast areas of land to feed cattle, instead of poor populations (Chellaney, 2015). These reasons, together with ethical considerations of the unfavorable living conditions and slaughter of animals, are at the core of the change in eating habits exhibited in current day society (for a thorough review, see appendix 1).

There are those who have not only reduced animal product consumption, but completely avoid consumption of animal products. Vegetarian or vegan diets represent about 12% of the western’s population diet choices (Povey, Wellens, & Conner, 2001; Kenyon & Barker, 1998). As discussed above, these diets stem from the following four causes: animal welfare consideration, health considerations, associations between meat and death, and environmental concerns. The main difference between animal welfair consideration and death related issues has to do with how the animals are treated and kept (animal welfare), and how the slaughter of animals makes one feel (death related issues are emotionally distressful). Research shows that also meat eaters are aware of animal welfare considerations and agree about the matter of cruelty towards animals, but rationelize eating animal products in other ways such as prioretizing their own well being and health (Worsley & Skrzypiec, 1997).

Although vegans and vegetarians represent only 12% of the population, the remaining meat eating population shows variety in dietary choices as well, from indistinctive meat eaters to those who eat only one type of animal product, and so on (Smil, 2002). Nevertheless, initiatives to reduce animal product consumption are a rather unexpected or even inconsistent occurrence in society, when observing human eating habits through a perspective of 11,000 years (Smil, 2002). Meat is considered an essential part of nutrition, health and of enjoying food (Smil, 2002). This suggests that consuming meat is deeply rooted into society, and therefore changing it is a challenging task.

(10)

10

2.2 What influences the decision to consume animal products?

The following section explains the process that determines willingness to consume animal products. It begins with an explanation of what shapes food preference for individuals, and ends with a conceptual model that explains eating behavior trough the theory of planned behavior.

2.2.1 Disgust and social influence

The domestication of livestock created a situation in which animals such as cows, chickens, goats, pigs, sheep and others are considered edible, while other animals such as dogs and cats are rarely considered food amid the world’s different cultures. This is an example of a paradox in eating meat: although meat is a coveted type of food, symbolizing survival, good nutrition and affluence, it is also the most tabooed type of food in civilization (Smil, 2002). Animal products have many restrictions and differentiations among meat types as to what is permitted to eat and what is not. Religion, culture, social norms and our sense of disgust have made only some types of animals deemed as right for consumption (Fessler & Navarrete, 2003).

Disgust serves as a disease prevention mechanism. For example, across almost all cultures, religions and societies, it is “disgusting” to eat rodents. That is because our cognition connects them with being dirty and therefore dangerous for consumption. This is a residue from our ancestors’ brain, as disgust is evolutionary developed to warn us from dangerous food that contains pathogens which may harm us (Fessler & Navarrete, 2003; Oaten, Stevenson, & Case, 2009). An animal’s appearance (“too ugly” or even “too cute”) is a prevailing factor that creates disgust about eating it (Ruby & Heine, 2012). Secondly, disgust is formed by internalization of social, cultural and religious norms. For example it is prohibited to eat pork in Islamic countries, because it “contaminates the soul”, therefore to “protect the body and the soul” it is thought that eating pork is disgusting, while it is seen as a delicious and healthy type of meat in North America and Europe (Rozin, 1996).

To continue this line of reasoning, the social and symbolic roles animals serve in society can create disgust at the thought of eating some types of animals. Animals which are perceived as less intelligent, non-emotional and generally remote from “being human” (beef, poultry, fish), are not tabooed to eat (Ruby & Heine, 2012).

(11)

11 On the other hand, animals that serve as pets, such as dogs and cats, are considered by western owners as family or “human-like” creatures (Fessler & Navarrete, 2003). The reason is that human beings tend to animate animals and attribute human behavior to animals, thereby eating them becomes a taboo, like canibalism. This is true especially for animals that live with people as pets. Pets are sometimes seen through the lense of “magical thinking”, for example to have a soul, feelings and thoughts, and therefore it becomes disguisting and immoral to eat them (Rozin, 1996). This means that individuals in most western societies will feel distressed at the notion of eating a pet (Fessler & Navarrete, 2003).

“Magical thinking” about animal’s emotions and intellegence can be further expanded to other animals than pets. Although the animal product supply chain is usually distant from a consumer’s mind when purchasing animal products, when contemplating about any edible animal’s psychological or emotional dimension, it resembles a human more and more, and as a consequence it becomes disgusting to eat it (Hoogland, Boer, & Boersema, 2005). This can be explained by the shrinking of psychological distance.

Psychological distance is a multidimensional construct that symbolizes the mental distance of an individual from a certain event or concept. High psychological distance makes an event more abstract, while low psychological distance makes something more concrete (Trope & Liberman, 2010; Trope, Liberman, & Wakslak, 2007). The change from high psychological distance to low psychological distance occurs when thinking and processing information about an event or concept. It can create positive or negative attitudes towards the event (Trope & Liberman, 2010). For example, one might have a positive attitude towards eating a juicy stake because the way it was made is psychologically distant and the animal used for it is considered edible (high level construct). However, the more a person contemplates about the emotional dimensions of an animal, may it exist or not, the more the animal becomes humanized. The more a person thinks about animals in human terms, the more they seem disgusting to eat (Ruby & Heine, 2012; Fessler & Navarrete, 2003). The reduction of the psychological distance between a person and an event may create emotional or cognitive distress, and in the case of eating meat, lead to disgust. In other words, when the familiar context of what is and is not considered food is challenged, mental discomfort arises. Therefore, there is a motivation to view certain animals as intelligent (dogs, cats) and others as less intelligent in order to be able to eat them (Fessler & Navarrete, 2003).

(12)

12 Because disgust requires internalization of social values, it has to do with cognitive processing (Rozin, 1996). However, it is hard to detach cognition from affect in the process of determining preferences altogether (Zajonc & Markus, 1982). Food values and attitudes are no different. Disgust is not purely cognitive when deciding which animal to eat (Fessler & Navarrete, 2003; Oaten, Stevenson, & Case, 2009). Therefore it is difficult to say which part of eating is emotion based, and which is cognitive based, as it is can both in different contexts (Rozin, 1996; Rozin & Fallen, 1987).

To summarize, evolutionary inherited disgust and social norms are the strongest predictors of which types of meat is considered edible (Fessler & Navarrete, 2003). After a certain type of meat is no longer “disgusting”, it can be considered edible. Next, factors such as taste, nutrition and affordability become aspects in food purchase decisions.

2.2.2 Determinants of decisions to purchase animal products

This section examines factors that influence an individual’s decision to purchase animal products. The literature reviewed in this section is adapted according to the consumer food demand pyramid, as presented in figure 1 (Kinsey, 2000 in Feuz & Umberger, 2003).

Different underlying factors can negatively or positively affect the decision to purchase and eat animal products. The first two factors that go hand in hand are income and affordability. A threshold amount of income enables individuals to be in the position of purchasing relatively expensive food such as meat.

Nutritious, safe, affordable Tastes good Variety

Convenient Promotes health

Living well Status / causes Income

Figure 1: the pydamid of food consumption (Source: Kinsey, 2000 in Feuz & Umberger, 2003).

Short term considerations Long term considerations

(13)

13 As income rises, the role of food in a person’s life changes: from the basic need of nutrition (short term goals) to symbols of lifestyle and values (long term goals). Secondly, food prices influence an individual’s decision whether to buy meat or not. Like most types of goods, price fluctuation is a large factor of purchasing decisions. When prices increase, average households show elasticity in demand, meaning reducing purchase of the particular animal product that has increased in price. Price elasticity is lower for affluent households (Caillavet, Fadhuile, & Nichele, 2014; Vasileska & Rechkoska, 2012).

Food safety is the next issue to influence purchase. Sensory quality cues determine if a type of food is safe for consumption: cues such as color, smell, taste and texture are crucial attributes in all food choices, and with meat in particular (Muiioz, 1998). Once quality cues are not satisfactory, for example a person does not like the texture of a particular meat, then disgust prevails in determining a preference and avoidance (Fessler & Navarrete, 2003; Oaten, Stevenson, & Case, 2009). Disgust and food safety become further intertwined in times when there are food scares: outbreaks of diseases that are contagious and deadly to humans, for example like Bovine spongiform encephalopathy, commonly known as “the mad cow disease” (Feuz & Umberger, 2003). These types of considerations about meat safety subside when the food scare period is gone. Hence, food scares have a short term negative effect on animal product consumption in general, with a specific effect on the demand for the type of animal who was the source of the disease outbreak (Verbeke, et al, 2007; Feuz & Umberger, 2003).

Nutrition, health and taste are the next considerations. Choosing to purchase or eat a type of meat depends on visual attributes such as freshness, visible fat amount, aroma, size, juiciness and more (Verbeke & Viaene, 1999; Grunert, 2006; Feuz & Umberger, 2003). In addition, the taste of the cooked product creates an impression about its quality and predicts future intention to buy (Muiioz, 1998). People that find meat healthy and an important part of nutrition, and, or, those who highly value the enjoyment of eating meat and connect positive emotions about eating meat as a means of self-indulgence, will consume more of it (Audebert, Deiss, & Rousset, 2006).

Food variety and the convenience of purchase are two more factors consumers consider when purchasing food. Eating is a repetitive process throughout the human life span, therefore convenience holds large importance for food preference. In other words, if food is (or is not) convenient for purchasing and cooking, it will (or will not) be consumed (Rozin, 1996).

(14)

14 The availability of products and the convenience of purchase is a common feature in modern retail in the western world, with online shopping that increases the convenience of food purchase even further (Feuz & Umberger, 2003).

Living in industrialized countries, especially in large cities, also positively affects meat consumptions because urban centers are affected most by globalization, meaning more access to food variety and exposure to other social influences (Yorka & Gossardb, 2004). Finally, as income increases, decision making about food can change. Higher standards of living are correlated with higher intake of animal products. On the other hand, research also finds that once reaching a certain level of financial affluence, consumers reduce animal product consumption, or choose particular types of products that are considered of higher quality, such as fish. This is due to their ability to translate their lifestyle, values, causes and status into food and eating patterns (Keyzer, et al, 2005; Yorka & Gossardb, 2004).

How do consumers shape their views about food? Beyond labels on food packages and basic health standards set forth in the developed world, consumers absorb information about food safety, quality and desirability from the media (Lusk, Roosen, & Fox, 2003; Johnson, 2015; Mathews, 2011). The media has an agenda setting role which influences public opinion (McCombs & Shaw, 1972). This means that the media can increase demand for animal products by increasing reports with positive information regarding animal products, or reduce consumption by focusing and repeating coverage that depicts the risks it has (Wassenhove, et al, 2012). Food advertisements have positive effects on consumption to the point that research finds links between food advertisement exposure and child obesity (Halford, et al, 2004).

To summarize, the thought process about eating meat starts with analyzing whether a certain type of meat is edible, thereafter recall whether it is tasty, and an evaluation of the quality cues of the current product in question (Verbeke & Viaene, 1999; Grunert, 2006; Muiioz, 1998). Whether a certain type of meat is edible is shaped by social norms. Whether a meat type is tasty and safe to eat has to do with evaluation of possible outcomes of eating it, and recall if past outcomes of eating it were positive. Once all analyzed parameters allow consumption, a consumption decision is likely to be made. If the parameters described above are not satisfactory, disgust rises and the food in question is not consumed (Fessler & Navarrete, 2003; Oaten, Stevenson, & Case, 2009). This decision making mechanism that leads to the behavior of eating animal products can be generalized under the theory of planned behavior.

(15)

15

2.3. The theory of planned behavior

As discussed in the previous section, food preferences are primarily socially and culturally based, with additional considerations such as price, nutrition and taste.

The numerous factors that are considered at the point of decision making to eat or not to eat meat suggest that it is an active cognitive process, rather than an automated decision (Thogersen & Olander, 2003; Rothgerber, 2012; Smil, 2002, Yorka & Gossardb, 2004; Brunsø, Scholderer, & Grunert, 2004b). This type of decision making can be explained by the theory of planned behavior.

The theory of planned behavior constructs the relationships between believes, attitudes and behavior (Armitage & Christian, 2003; Ajzen, 1991).

As shown in figure 2, attitudes do not directly lead to behavior. The gap between attitudes and behavior is mediated by behavioral intention. Intentions are composed by attitudes, subjective norms and perceived behavioral control. Each component is shaped by the beliefs towards it: attitudes are shaped by behavioral believes towards a behavior’s outcome (negative or positive evaluations of the behavior) and outcome expectation. This is reinforced or changed through a feedback loop after the behavior occurs. Perceived behavioral control is shaped by control beliefs, meaning how much a person believes it is attainable and within his or her power to behave in a certain way. The lower the perceived behavioral control, the less likely a positive intention will be formed.

Outcome expectation

Evaluation of outcome expectation

Opinions of others Motivation to comply

Attitude toward behavior

Subjective norm

Behavioral intention Behavior

Feedback

Figure 2: The theory of planned behavior. Based on: Ajzen, 1991, Armitage & Christian, 2003, pp. 193, Jager, 2000, pp. 82. Behavioral beliefs:

Normative beliefs:

(16)

16 Perceived behavioral control does not only influence intentions, but also directly influences behavior. Finally, subjective norm is comprised by the normative beliefs a person has: the opinions of others and their importance, measured by motivation to comply. All the mentioned above factors shape behavioral intentions, which lead to behavior. The weight that each component has when comprising intentions is different per decision areas (Armitage & Christian, 2003).

To use this theory in the context of eating animal products, beliefs that positively influence attitudes (and thereafter intentions) to eat meat are that it is healthy, convenient, tasty and so on. Perceived behavioral control is relevant in the case that meat become less attainable, which makes a person evaluate whether the behavior of eating meat is under their control. Subjective norms are shaped by what other people consider as appropriate behavior and the individuals’ own motivation to comply (for example, family’s opinions are more important than strangers’ opinions).

When regarding eating decisions, social influence is the strongest predictor for dietary choice altogether (Worsley & Skrzypiec, 1998). Some religions and cultures have dietary restrictions in terms of which type of meat is allowed for consumption, and in terms of how it was made. On the other hand, some countries have a “meat eating culture” as part of the local traditions, habits and folklore. This creates an emotional connection to eating animal products as well. This is true mostly in Northern Europe and the US (Smil, 2002; Yorka & Gossardb, 2004). In addition, part of social influence on eating habits has to do with gender roles. Research shows that men are more prone to eat meat as it is conceived as more “masculine” by society and thus by males that want to conform to social rules. This means that with all things constant, men are more likely to eat animal products than women, and are less likely to reduce consumption due to external stimuli (Rothgerber, 2012). As evidence, vegetarianism in society is generally more prevailing with females than males. One explanation is that females are socialized to be more altruistic and males are socialized to enjoy hedonistic values (Worsley & Skrzypiec, 1998; Kubberød, et al, 2002).

The theory up until this part explains intentions under the assumption that attitudes are consistent. In reality, competing attitudes exist as well. A person can have positive and negative attitudes simultaniously towards any behavior. This is called ambivalance, and it is a moderator between attitudes and behavior (Conner, et al, 2003).

(17)

17 Higher ambivalent people are more inclined to change their views when introducig new information that “tippes the scale” towards one of the competing attitudes (Trope & Liberman, 2010; Berndsen & Pligt, 2004; Kenyon & Barker, 1998; Povey, Wellens, & Conner, 2001; Rozin, 1996; Conner, et al, 2003). This information either supports a negative attitude they already had or a positive attitude they already had (Conner, et al, 2003). Lower levels of ambivalance have stronger perdictive power of the attitude – intention – behavior relationship (Conner, et al, 2003). When this conflic (which can be cognitive, affective, or both) persists, it may lead to psychological discomfort that must be resolved (Aaker, Drolet, & Griffin, 2008).

In the context of eating animal products, an example of having competing positive and negative attitudes is enjoying the taste of meat, while also beleiving it is not healthy to eat. Research shows that decisions regarding eating animal products can be seen as on a scale, with vegetarianism on one end, and indistinctive meat eaters on the other (Wellens, & Conner, 2001). The scale is categorized by ambivalence toward eating animals: the more ambivalent a person is, the higher the psychological distress she or he feels towards meat consumption. Vegans and vegetarians are low ambivalent, as they have decided not to eat meat. Those who indistinctively eat meat, on the other side of the scale, are also low on ambivalence because they do not hold strong conflicting feelings about eating meat. This means that those people who are ambivalent are somewhat the group in between those two opposites: meat avoiders that have both positive and negative attitudes towards eatig meat. Therefore they choose to restrict or avoid some aspect of eating meat such as quantety or meat type (Berndsen & Pligt, 2004; Kenyon & Barker, 1998; Povey, Wellens, & Conner, 2001).

The psychological distress of high ambivalence occurs as people’s beliefs and behavior are not aligned: those people believe eating animal products has negative aspects, but they do not always act according to the values they hold (Brunsø, Joachim, & Grunert, 2004). Ambivalance is therefore an extention of the classic theory of planned behavior (Conner, et al, 2003).

Theoretucally, when adding new information, meat avoiders make a decision about whether or not to eat meat according to the nature of the information. Positive information may lead to eating meat, and negative information may lead to not eating meat. This notion brings about the first hypothesis to be tested in this research: H1: Meat avoiders will be less willing to pay for animal products after being exposed to negative information about meat, compared to indistinctive meat eaters.

(18)

18

Will indistinctive meat eaters also be influenced by new negative information? The theory of planned behavior provides a general model of decision making as a tool of predicting behavior. However, it does not explicitly construct methods on how to change decisions or behavior. Consumer behavior literature about increasing willingness to pay deals with perceived price, perceived quality, value, satisfaction and loyalty, and the interrelations between them in order to improve consumers’ attitudes about them (for example see Jayawardhena, 2010; Gallarza, Gil-Saura, & Holbrook, 2011; Xia, Monroe, & Cox, 2004; Caceres & Paparoidamis, 2007; Tohidinia & Haghighi, 2011). However, reducing willingness to pay calls for changes in consumer’s perceptions about the products – to the worst.

To summerize, in order to change behaior, intentions must be changed. Intentions are made of subjective norms, attitudes and percieved behavioral control. Changing social influinces is almost impossible (Venkatesan, 1966). Therefore in order to change behavior, it is necessary to change the core beliefs that shape attitudes, and eventually intentions. This can done by adding information that alters core beliefs (Fishbein & Ajzen, 2011). For this purpose, a marketing tool called counteradvertising can be used.

2.4 Changing attitudes with counteradvertising

Advertising is a marketing strategy tactic. The explicit nature of advertising is aimed at creating positive consumer perceptions towards products of brands (Eisend, 2006).

Short term Long term Health Animal welfare Envirionment Nutrition Health Taste

Vegans Vegetarian Meat avoiders Meat eaters

High level of ambivalence

Low level of ambivalence: long term goals more important

Low level of ambivalence: short term goals more important

Long term vs short term goals compete

(19)

19 Although transitionally advertising is engaged in influencing consumer behaviors in order to increase willingness to buy, an opposite inclination is gaining momentum in the last decades, namely, demarketing (Achrol & Kotler, 2012; Vargo & Lusch, 2004; Gerstner, Hess, & Chu, 1993). This practice uses marketing tactics to change consumer’s behavior to reduce usage or purchase of a product. The goal of demarketing is to gain long term social and cost effectiveness benefits (Kotler & Levy, 1971). One of the tactics demarketing employs is counteradvertisement. This practice is aimed at discouraging consumers from buying products or engaging in particular activities, whether altogether or in reduced quantities.

Counteradvertisement is mainly used as a marketing tool in the health and NGO industry to tackle problems such as alcohol, tobacco and drug abuse (Agostinelli & Grube, 2002; 2003; Dixon et al, 2007) (See appendix 3 for a thorough review). Different types of techniques are used for this purpose, mostly aimed at educating consumers, warning of adverse results of using a product, or reducing a product’s availability to target groups (Agostinelli & Grube, 2002; 2003). Just as in traditional advertisement, cognitive and affective messages are used in printed ads, commercials and digital media, and just as advertisements in general, counteradvertisement has different influence and recall on different types of consumers. Therefore messages need to be “tailor made” for consumer target groups (Audrain-McGovern, et al., 2003). Another method of counteradvertisement is labeling products, for example, placing warning labels with short messages about the damages of smoking on cigarettes packages. This creates an intuitive link between cigarettes and health problems, which in theory should discourage smoking. However, research in the field does not offer a conclusive recipe for successful behavioral changes. Some counteradvertisement campaigns are effective, while others have limited success if any (Agostinelli & Grube, 2002; 2003).

As explained earlier, the theory of planned behavior assumes decision making is rational, bound by cognition and sensitive to information processing. Therefore, in this model, the most effective approach in counteradvertising is the outcome expectancies, which is a balance between the rational gains and losses associated with the behavior in question. For example, in smoking, gains are enjoyment and costs are potentially getting cancer. An evaluation of this would be “what are the chances to get lung cancer?”. This evaluation forms attitudes towards smoking (Agostinelli & Grube, 2003).

(20)

20 The content of counteradvertisement messages is therefore aimed at reducing the benefits of a behavior or increasing its implicit costs, which are damaging either the user or the user’s environment (Mosbergen, 2014). Focusing on outcome expectancies is deemed fairly effective in decreasing intention to smoke (Agostinelli & Grube, 2003). Another approach is influencing perceived social approval by creating and promiting new social norms. For example, claiming that smoking is considered unaccapted or “uncool” (Agostinelli & Grube, 2002; Ibrahim, 2010). This type of message may be effective only when there are already pre-existing negative attitudes about smoking with non-smokers.

A behavioral change of actual smokers is less evident, and may increase backlash: smokers may form even more favorable attitudes towards the behavior and increase smoking (Agostinelli & Grube, 2003).

Finally, messages that create negative self-implications are threatening to a person’s self image (for example “smoking causes impotence” may not even be processes or remembered by males). Such counteradvertisments will evokes defensiveness towards a counteradvertisment message and eventually its rejection (Agostinelli & Grube, 2003).

To summerize, non-smokers do seem to shape negative attitudes towards smoking after being influenced by counteradvertising. This means they may avoid smoking. However, actual smokers do not seem to be influenced by counteradvertisement, either due to wrong measurement methods or due to a flawed or ineffective execution of advertising techniques (Agostinelli & Grube, 2003). Changing the behavior of those who have favorable attitudes towards smoking and the behavior that matches it is still puzzling.

Behavioral change is most attainable when there are already some positive beliefs about a behavioral change and when there is an inconsistence between these beliefs and actual behavior (Dickerson, et al, 1992). If a conflict between a belief and a behavior does not (yet) exist, adding negative information about a behavior might create resistance to the message, and instead of changing core beliefs, people may strengthen them or increase the undesired behavior (Agostinelli & Grube, 2002; 2003; Dickerson, et al, 1992). Counteradvertisement is met with resistance from consumers and different stakeholders in the economy. In such cases there is a backlash of an increase in the undesired behavior when target groups are disgruntled at persuasion messages which are overly forceful (Agostinelli & Grube, 2002; 2003; Dickerson, et al, 1992).

(21)

21 For example, counteradvertising regarding smoking cigarettes has encountered strong resistance not only from tobacco companies, but also from smokers themselves (Ong & Glantz, 2001; Brownell & Warner, 2009). This type of phenomenon stems from cognitive dissonance (Aronson, 1969).

Cognitive dissonance has to do with consistancy between believes and behavior (Harmon-Jones & Mills, 1999). Believes are cognitive (and sometimes emotional) constructs. When pairs of cognitions (elements of knowledge) are relevant to one another, they can be either consonant (one follows from the other) or dissonant. When there is lack of consistancy between attitudes and behaviors, a pschological discomfort is triggered at the face of this internal conflict of competing attitudes and, or, behaviors (Harmon-Jones & Mills, 1999). The greater the dissonance, the greater the motivation to change it becomes. In order to reduce pschological discomfort and create internal consistancy once more, one of the following will occur: either the disonant cognition will be removed (by avoiding information that might increase dissonance, or reducing its importance), or a constant cognition will be added to increase the importance of current cognition. The likelyhood of a change in behavior is based on the resistance to change of the cognition, which is the interplay between the responsiveness of the cognition to reality and other constant cognitions (Harmon-Jones & Mills, 1999). Finally, it is possible that behavior might change, if the “loss” connected from the change is smaller than the “gain”, and if the satisfaction of the current cognition is larger than the potential behavioral change (Harmon-Jones & Mills, 1999).

Cognitive dissonance has been successfully applied in attitude changes in many settings and contexts. In recent years, it has also been used as a tool for persuasive communication. As mentioned before, changing behavior with cognitive dissonance is the most effective when there is a conflict between already existing values, and exhibited behavior (Dickerson, et al, 1992). When an attitude is not pre-existant, using cognitive dissonance may be futile. Also, when new information is too overwhelming, a potential danger is backlash in the form of increased undesired behavior. Therefore, new information needs to challenge current beliefs in order to directly support negative attitudes about a behavior, or reduces positive attitudes. New information can also change attitudes indirectly, by creating inconsistencies between one set of beliefs (for example: “I care about my health”) and others (“eating meat is not healthy”).

(22)

22 In the context of this research it raises a second hypothesis: H2: A counteradvertisement message on a product will reduce willingness to pay for it relative to a product that does not feature counteradvertisement. In addition, the first hypothesis is revised from “H1: Meat avoiders will be less willing to pay for animal products after being exposed to negative information about meat, compared to indistinctive meat eaters” into “H1: Meat avoiders will be less willing to pay for animal products after viewing a counteradvertisement message, compared to indistinctive meat eaters”.

The following section addresses the messages that will be presented in the counteradvertisement in this research, and the theoretical framework that explains their potential influence on willingness to pay.

2.5 Counteradvertisement messages aimed at reducing animal product consumption

Lifestyle is a general description of a person’s overall values, behavior and habits. For example, people that are health conscious will be less inclined to smoke. People that have an emotional connection to animals might contribute to animal welfare charities, and so on (Worsley & Skrzypiec, 1998). In order to change a belief, new information needs to be introduced. This has the potential to influence the belief, and create a conflict between a person’s behavior and lifestyle.

The following three lifestyles are in conflict with eating animal products, and therefore can be used as information items that create cognitive dissonance. Therefore, adding negative information to pre-existing attitudes such as attitudes towards health, animals and environment may have an effect on willingness to pay. Making a link between eating animal products and negative consequences on health, the environment and animal welfare can be especially influential to those who already have positive (and therefore conflicting) beliefs about any of those subjects.

Nutritious value and impact on health

Health is a representation of one’s self, which is a critical component in choosing food. This attribute is one of the two most important features in all food choice (Muiioz, 1998). Foods which are seen as having a positive impact on health and with a high nutritious value will be more coveted (Feuz & Umberger, 2003). This means that when people believe that animal products are healthy, they will demand and prefer them.

(23)

23 The opposite is also true: when people belive that eating many animal product is unhealthy, they will be more inclined to reducing consumption (Verbeke, 2000). Therefore, a hypothesis to be examined is H3: people who believe eating animal products is unhealthy will be willing to pay less for meat after viewing a counteradvertisement message about health, compared to those who believe it is healthy.

Animal welfare considerations

People do not like to think about slaughter of animals because thinking about it causes psychological distress (Fessler & Navarrete, 2003). To reduce this distress, one may add positive cognitions to eating animal products, or reduce the importance of the negative cognition. Therefore, a hypothesis to be examined is H4: People with strong positive attitudes towards animals will be willing to pay less for animal products after viewing a counteradvertisement message about animal welfare, compared to others that do not have strong positive attitudes towards animals.

Attitudes about the environment

Once a person possesses personal norms regarding environmentally-friendly behavior such as recycling and using less fuel, the need for consistency in all life’s decisions creates spillover effects to other domains such as consumption (Gilg, Barr, & Ford, 2005). Such people are more informed and actively seek information about environmentally friendly behaviors, and therefore may be more inclined to reduce the consumption of animal products if they believe it is bad for the environment (Thogersen & Olander, 2003). Therefore, a hypothesis to be examined is H5: People with strong positive attitudes about the environment will be less willing to pay for animal products after viewing a counteradvertisement message about the environment, than others that do not have strong positive attitudes about the environment.

The following chapter elaborates on the method to be used in chapter 4 to test the following hypotheses:

H1: Meat avoiders will be less willing to pay for animal products after viewing a counteradvertisement message, compared to indistinctive meat eaters.

H2: A counteradvertisement message on a product will reduce willingness to pay for it relative to a product that does not feature counteradvertisement.

(24)

24 H3: People who believe eating animal products is unhealthy will be willing to pay less for meat after viewing a counteradvertisement message about health, compared to those who believe it is healthy.

H4: People with strong positive attitudes towards animals will be willing to pay less for animal products after viewing a counteradvertisement message about animal welfare, compared to others that do not have strong positive attitudes towards animals.

H5: People with strong positive attitudes about the environment will be less willing to pay for animal products after viewing a counteradvertisement message about the environment, than others that do not have strong positive attitudes about the environment.

The theoretic model that this research will use is the theory of planned behavior, with a focus on behavioral beliefs:

Outcome expectation Evaluation of outcome expectation Attitude toward behavior

Behavioral intention Behavior

Feedback

Behavioral beliefs:

Ambivalence

(25)

25

Chapter 3 Methodology

This chapter describes the method used to examine the hypothesis and answer the research question. It begins with an explanation about the research method, analytical strategy and data collection. In ends with a description of the sample, to be analyzed in chapter 4.

3.1 Research method

This research examines whether counteradvertisement has an effect on animal product consumption, and if so, to which extent does it reduce willingness to pay. For this purpose, a survey will be used (see appendix 4). The survey has been disguised as a survey that measures the lifestyle of pet owners. The reason for adding unnecessary elements to the survey is to reduce hypothesis guessing bias, and reduce resistance to participate by those who are indistinctive meat eaters, because it is plausible to assume that indistinctive meat eaters will feel “attacked” once a survey is explicitly about reducing meat consumption.

The type of animal product that will be asked about in the survey is fish. Fish is neither “too cute” nor “too ugly” to create disgust (Ruby & Heine, 2012). Unlike “cute looking” animals, fish do not arouse emotional associations, nor they are perceived with human like features such as a complex social hierarchy, intelligence or emotion. In addition, eating fish is considered healthy and nutritious (Fung et al, 2001; Pan, et al., 2012). Therefore, fish will pose as a “stress test” for the hypotheses.

The package of fish used in the survey is a real product, sold in food stores in The Netherlands. The product is available for purchase in stores and online as well, for example through the food retailer Albert Heijn. The reason of using a real product is to test the hypothesis in a retail context that is as close as possible to real world conditions, when consumers need to make purchase decisions in stores or online. Because an electronic survey reduces exposure to sensory cues such as texture and smell, it is mentioned that the fish is of high quality and freshness, to eliminate disgust or resistance to purchase on that base.

The survey has a mixed factorial design. First, the between-subject model has three treatments: health, animal welfare and the environment. In addition, each participant may change his or her answer after being exposed to the condition.

(26)

26 This means a within-subject design in which each participant will be the control group to him/herself: at first the participant will establish the price one would pay for a pack of salmon and thereafter a counteradvertisement message will be presented, after which the same participant will be asked to answer how much she or he is willing to pay for the package at that point. The goal will be to analyze the difference before and after the treatment, in the hopes that willingness to pay “after” manipulation will be (statistically) significantly lower than “before” throughout all treatment types. This type of survey design has the benefits of statistical power and internal validity because it reduces errors that are made due to differences between subjects (Greenwald, 1976; Charness, Gneezy, & Kuhn, 2012). This means that results will not be distorted due to individual differences.

Each participant will only be able to see one treatment in order to reduce carryover effects to the next treatment (Greenwald, 1976; Charness et al, 2012). Each respondent was randomly assigned either a “health message”, an “animal welfare message” or “environmental message”, and thereafter was asked to fill in several measures about attitude.

1. The health condition will feature “removing fish from your diet eliminates half of all

your mercury exposure and reduces your intake of other toxins”.

2. The animal welfare condition will feature “when these fish are caught, they suffer

pain and fear until they die”.

3. The environment condition will feature “water pollution, waste and 68 percent of

fish extinction worldwide is caused because of fishing”.

Health condition

Animal welfare condition

The environment condition

How much would you pay? (Before condition)

How much would you pay? (After condition)

Between subjects

Within subjects

(27)

27 Each question about attitudes is measured on a scale of 1 (completely disagree) to 5 (completely agree). Finally, each participant has to rate which type of value is most important to them in their lives: health, animal welfare or the environment.

The amount to be paid has been established with one type of coin (dollar), but this is a relative term rather than an absolute one, because not all participants are North American. A European might find it as a nuisance to calculate how much she or he will pay in Euros and then transform it to dollars. This might create the risk of too much complexity, which could reduce response rate. Therefore, although less favorable, the following text was written next to the question: “Please indicate the amount of dollars you would pay using digits in this empty space:.. (If you are not sure, or your local currency is different, then please just guess what you would pay or what you can afford”. Adjusting for differences is simple with creating percentage decrease or increase. For a full list of questions and the variables they measure, see appendix 5.

3.2 Survey distribution and analytical strategy

Several pilot tests were initiated to make sure the survey is clear and “user friendly”. After modifications and adjustments with the advice and contribution of Drs. A. (Anouar) el Haji, the survey was launched between June and July 2015. The survey had an electronic restriction: only one respondent per survey (tracked by IP address) and each question had to be answered in order to continue to the next one.

The survey was distributed online through different types of electronic media such as social media (Facebook, Twitter) ad e-mail. The survey was posted in a few online discussion groups that were disbursed across different interests (for example, university related groups, groups of animal enthusiasts, groups of fashion bloggers, and so on), and in addition through a Twitter account with about a 1000 followers. The reason that social media was the preferred medium of distribution is twofold: reach and type of sample. First, it allowed for a large reach and disbursement across geographic areas, demographic characteristics, genders and ages. In addition, this type of distribution allowed for a random sample rather than a convenient or snowball sample, which reduces biases. Secondly, those who have an online presence and are able to take part in an electronic survey probably will poses basic enabling factors that will allow them to be in a position to purchase animal products to begin with. In other words, standards of living, urbanization, price, and convenience will not pose as problems that will create noise in the sample analysis.

(28)

28 Finally, the target audience represents those geographic areas in which animal product consumption is in a world peak, with Europe and North America as the origin of the majority of respondents

The total number of accepted responses was 281, out of a larger survey sample. Only complete responses were accepted in the sample, therefore there are no missing values. One respondent was deleted from the sample due to an obvious faulty manner of answering the survey, therefore it was removed (N = 280). The answers received in the survey program were coded with Microsoft excel. Irrelevant questions such as attitudes towards eating vegetables and other lifestyle indicators (inserted only to disguise the survey’s purpose) were deleted. The data was thereafter inserted in SPSS to be analyzed. Counter indicative items (see question 14a, appendix 4) were transformed in their rating (5 = 1, 4 = 2, 3 = 3, 2 = 4, 1 = 5) to fit into the variable “psychological distance: general attitudes”. (See appendix 5 for all the questions included in computing the variable).

3.3 Descriptive statistics

In the first question “How much would you pay to buy this pack of high quality salmon?” (Before treatment), an extreme outlier is apparent (value = 100), which makes skewness= 7.554 and Kurtosis 88.086. A closer inspection of the respondent teaches about a probable typing error, as it is a person from The Netherlands, meaning that a 100 Euros for one pack of fish is an improbable response. That case was also deleted making N = 279. Overall most results have a normal distribution as Skewness and Kurtosis are around 0 and between -1 and 1. However, normality of the first two questions is problematic due to high kurtosis. This is statistically problematic, however this makes economic sense. There is a general reference point to how much a person would pay for a pack of fish, especially in the west. Therefore it is logical to have high kurtosis in the means of “before” and “after”, as the change is stable yet with a similar dispersion around a mean.

Total effect also has high kurtosis because of the fact that overall, the change (between before and after treatment) has a mode of zero. Therefore higher kurtosis is logical as well. When looking at the frequency it becomes more evident: there are overall 161 (57.5%) people that would not change their price point between before and after. There are 81 (29%) respondents that have lowered their price after the treatment.

(29)

29 The remaining respondents have increased their price offer (about 13%). This means that high kurtosis simply represents the reality of the effect of the counter advertisement: the effect has a mode of zero.

The mean score of the statement “animals should be treated with respect” is not normally distributed. When looking further into this question, 37.9% of responses “agree” and 60.7 % “completely agree” with the statement. Therefore, this question is not normally distributed, however it does not necessarily need to be transformed, as the meaning behind it is that the majority of the sample have positive attitudes about animals: 98.6% of the respondents support the statement “animals should be treated with respect”. The statement “I care about my health” has similar response rates with 51.1% answering “agree” and 45% answering “completely agree”. Therefore it is no wonder that kurtosis is high, as most responses (96.1%) are clustered around these two answers. Therefore, even though that some questions are not normally distributed, changing them will reduce the potential interpretation of the results.

3.4 Sample characteristics

There are 121 (43%) males and 157 (56%) females in the sample, with 1 person identifying as “other”. There are 26% respondents between the age of 18 – 24, 49% of those who are between 25 – 34 years of age, and 19% of those who are between the ages 35 – 54. There are 3.9% of respondents that are older than 55 and 0.7% younger than 18. The majority of the sample has some level of higher education: 43.9% have a Master of Science degree, 24% have a bachelor’s degree, 10.4% have a college / degree and 3.9% have a PhD. The remaining 8.2% of the sample have a professional or technical diploma, 6.8% have only a high school diploma and an additional 2.5% have no schooling completed at all. The amount of respondents with higher education is a limitation for generalizing results, as it is doubtable that 67% of the general population have obtained higher education. Nationalities are disbursed across North America (16%), Continental Europe (59.25%), and Israel (17.74%). The other remaining 7% are disbursed across the world. A second limitation of the research evident by now is that there is a skewed representation of one country within Europe: The Netherlands, with 49.69% of total respondents coming from there.

129 respondents (46%) do not own pets. 33.2% have at least one animal. 20% of the sample have some sort of animal which is socially acceptable to eat (such as a rabbit, fish).

(30)

30 When dividing ambivalence into groups (1: low, 2: high), the mean is 1.17 with a 0.37 standard deviation, with a normal distribution (median and mode are 1).

3.5 Reliability

Five questions (14a-14e, see appendix 4) have been asked to measure the variable “Psychological distance to animals: general attitudes about animals”. The questions needed to be transformed into one variable by computing scale means. The first question is “What are your general perceptions about animals? I don't like animals as pets” is a counter indicative item. It was transformed by recoding it (5=1, 4=2, 3=3, 2=4, 1=5). When looking into the reliability of scale means compromising the new variable, Cronbach's Alfa is 0.798, which can be seen as a reasonable score. Deleting any of the items will not increase Cronbach's Alfa, therefore all five questions that comprise the new variable have been used to create the mean scale. The new variable has a mean of 4.26, skewness of -0.985 and kurtosis of 1.7. The standard deviation of the variable is 0.649.

(31)

31

V µ SD DV1 DV2 DV3 CV1 CV2 IV1 IV2 IV3 IV4 IV5 IV6 IV7 IV8 IV9 IV10 IV11 IV12 IV13 IV 14 DV1 6.8 4.29 - DV2 5.89 4.98 .789** - DV3 0.11 0.361 .085 -.414** - CV1 3.06 1.035 .063 .082 .044 - CV2 2.97 1.263 .057 .099 -.013 .199** - IV1 0.56 0.497 .127* .156** -.02 .168** .155** - IV2 2.99 0.802 .119* .107 .073 .53 .085 .01 - IV3 1.52 1.02 .213** .178** .06 .173** .085 .135 .163** - IV4 3.03 1.329 -.131* -.136* -.019 -.127* -.109 .01 -0.206** -.143* - IV5 0.85 1.01 .184** .117 .083 .101 .119* .224** .078 .141* -.260** - IV6 1.2 0.469 -.03 -.022 .072 -.003 .101 .083 .025 .026 -.005 .078 - IV7 4.26 0.649 .091 .119* -.08 .159** .094 .260** -.115 .167** .-129* .290** .022 (0.798) IV8 3.35 1.11 .029 .04 -.019 .152* .216** .144* .052 .006 .060 -.008 .133* .136* - IV9 3.29 1.245 .07 .123* -.065 .168** 0.307** .368** .104 .113 -.050 .079 .129* .292** .604** - IV10 2.53 1.095 .002 .027 .075 .209** .232** .202** .125* .038 -.090 .047 .221** .011 .439** .435** - IV11 4.41 0.603 -.012 -.024 -.015 .104 -.051 .195** -.132* .047 .140* -.015 -.047 .360** .082 .118* -.008 - IV12 3.01 1.344 .05 .052 -.076 -.07 .017 -.197** .01 -.137* -.029 -.102 -.119* -.181** -.147* -.166** -.154** -.406** - IV13 4.05 0.773 .122* .119* -.028 .126* .127* .264** -.016 .246** -.137* .268** .050 .585** .174** .335** .162** .369** -.423** - IV14 1.237 0.531 .004 -.005 .009 .072 -.022 .086 .014 .266** -.073 .253** .127* .077 .144* .153* .150* -.076 -.181** .223** -

** Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed).

Independent variables IV

Type of treatment IV0

Gender IV1

Age IV2

(Meat eating) culture/region IV3

Education IV4

Psychological distance to animals: owning pets IV5

Ambivalence IV6

Psychological distance to animals : general attitudes IV7 Attitude towards eating animal products and the environment IV8 Attitude towards eating animal products and animal welfare IV9 Attitude towards eating animal products and health IV10

General attitudes

IV11 IV12 IV13

Strongest of the general attitudes IV14

Dependent variables DV

Willingness to pay before DV1

Willingness to pay after treatment DV2

Total effect DV3

Control variables CV

Price sensitivity CV1

Taste preference CV2

Does counteradvertising affect the willingness to purchase animal products?

MSc Business Administration

Marketing Track

Does counteradvertising affect the willingness to purchase animal products?

Statement of originality

Acknowledgments

Table of contents

Abstract

List of variables

Chapter 1

Introduction

Chapter 2

Literature review

Chapter 3

Methodology