PHD THESIS

Sustainable Consumer Behaviour

Supervisor: Prof. Marco Frey Tutor: Prof.Francesco Testa

Pisa, September 2015.

PhD candidate: Ajla Cosic

2

©2015, Ajla Cosic. All rights reserved. Printed in Pisa, Italy. Sant’Anna School of Advanced Studies, Institute of Management. Piazza Martiri della Liberta 24, 56127 Pisa, Italy.

3

“Read! In the Name of your Lord, Who created” The Qu’ran 96:1

4

Acknowledgements In September 2013 I started my PhD journey at Scuola Superiore Sant Anna in Pisa. Last two years of my PhD journey I have spent at London School of Economics, Em Strasbourg Business School and Bilgi University learning about science. I would like to thank all people that I have met for encouraging my research and for allowing me to grow as a research scientist. Your advice on both research as well as on my career have been priceless.

I wish to express my sincere thanks to Institute of Management for providing me with all the necessary facilities for the research. I would like to express my special appreciation and thanks to my advisor Marco Frey and my tutor Francesco Testa. I had the pleasure to work on different chapters of my thesis with Fabio Iraldo, Francesco Testa, Sebastian Ille, Hana Cosic and Sihem Dekhili.

A special thanks to my family. I owe my deepest gratitude to my parents and sister. Words cannot express how grateful I am to my mother, father and my sister for all of the sacrifices that you have made on my behalf. When it was hardest you have been there, making this journey easier. I would also like to thank all of my friends and colleagues who supported me in writing, and incented me to strive towards my goal.

5

Table of Contents 1 Introduction ...........................................................................................................................9 1.1 “Attitude Behaviour Context” (ABC) theory .................................................................10 1.2 ‘Nudges’ and consumer behaviour .................................................................................11 1.2.1 ‘Nudges’ and recycling - Can Nudges Affect Students’ Green Behaviour? .......11 1.2.2 ‘Nudges’ and healthy food - Nudging Students toward Healthier Choices in a University Cafeteria ............................................................................................................12 References.............................................................................................................................13 2 Determining factors of curtailment and purchasing energy related behaviours ..........15 2.1 Introduction ....................................................................................................................16 2.2 Theoretical framework and research hypotheses...........................................................18 2.2.1 Attitudinal factors as determinant of energy-saving behaviour...........................18 2.2.2 Contextual factors: the role of trust concept in energy-saving ...........................19 2.2.2.1 Government ..........................................................................................21 2.2.2.2 Environmental NGOs ...........................................................................21 2.2.2.3 Private companies ..................................................................................22 2.2.2.4 Friends and family ................................................................................23 2.2.3 Personal capabilities ............................................................................................23 2.3 Methods ..........................................................................................................................24 2.3.1 Measurements .......................................................................................................25 2.3.1.1 Independent variables ..............................................................................25 2.3.1.1.1 Level of trust in the information provided by different entities.25 2.3.1.1.2 Personal norms ..........................................................................26 2.3.1.1.3 Personal Capabilities .................................................................27 2.4 Empirical Models ...............................................................................................................29 2.5 Results ................................................................................................................................30

6

2.6 Discussion ..........................................................................................................................33 2.7 Conclusion ..........................................................................................................................35 References ................................................................................................................................38 Appendix ..................................................................................................................................45 3 Nudges Can Affect Students’ Green Behaviour? –A Field Experiment .........................48 3.1 Introduction.....................................................................................................................49 3.2 Literature review ............................................................................................................50 3.3 Model .............................................................................................................................52 3.4 Methods ..........................................................................................................................56 3.5 Results ............................................................................................................................59 3.6 Discussion and conclusion .............................................................................................63 References ............................................................................................................................65 Appendix ..............................................................................................................................67 4 Nudging Students toward Healthier Choices in a University Cafeteria ........................68 4.1 Introduction ....................................................................................................................69 4.2 Literature review ............................................................................................................70 4.2.1 Social norms .........................................................................................................71 4.2.2 Convenience and other ‘nudges’ ..........................................................................72 4.3 Methods ..........................................................................................................................73 4.3.1 Experimental design ...........................................................................................73 4.3.2 Treatment: The role of social norm and ‘easy to choose’ nudge on healthy food purchase ..................................................................................................................................74 4.4 Results and discussion ....................................................................................................76 4.4.1 Why nudge do not always work out as planned? ..................................................79 4.6 Conclusion ......................................................................................................................81 References ............................................................................................................................84 5 Conclusions ..........................................................................................................................88

7

List of Tables and Figures Tables 2.1 Correlation matrix and descriptive statistics .....................................................................28 2.2 Results of regression analysis ...........................................................................................31 4.1 Prices and total quantity of drinks and food sold during control and treatment period ....78 4.2 T test statistics-drinks ........................................................................................................79 4.3 T test statistics-food...........................................................................................................79

Figures 2.1 Conceptual model and Hypotheses ...................................................................................29 3.1 Dynamics of the control group ........................................................................................55 3.2 Dynamics of treatment 1 ...................................................................................................55 3.3 Dynamics of treatment 2 ...................................................................................................56 3.4 Treatment 1 ........................................................................................................................58 3.5 Treatment 2 ........................................................................................................................59 3.6 Survey results ....................................................................................................................60 3.7 Percentage of recycled cups over the experimental period ...............................................61 3.8 Average of percentage of recycled cups ...........................................................................62 3.9 Treatment 2 – Share of correctly disposed recyclable and non-recyclable garbage .........62 3.10 Effects of parameter changes. ..........................................................................................67 4.1 Social norm message .........................................................................................................75 4.2 Social norm message and label ‘healthy eating’ in cafeteria.............................................75 4.3 ‘Easy to choose’ nudge - green footprints in cafeteria ......................................................76 4.4 Sales of healthy and less healthy food in cafetaria............................................................77 4.5 Sales of healthy and less healthy drinks in cafetaria .........................................................77

8

“Bismilahir-rahmanir-rahim! I call to witness the ink, the quill, and the script, which flows from the quill; I call to witness the faltering shadows of the sinking evening, the night and all she enlivens; I call to witness the moon when she waxes, and the sunrise when it dawns. I call to witness the Resurrection Day and the soul that accuses itself; I call to witness time, the beginning and end of all things - to witness that every man always suffers loss.” Mesa Selimovic, Death and the Dervish

9

1 Introduction One of the important long term social and policy challenges facing the planet is how to promote sustainable resource use and change people’s behaviour. Sustainable development requires not only technological innovations but also changes in individual and collective behaviours. In our opinion policies that ignore results of human psychology and assume that we are Homo economicus will hardly reach their aimed level of impact. Why do not we save more energy? Why do not we recycle more? Why do not we eat more healthy food? For possible explanations and answers to these questions principles of consumer behaviour can be used. Consumer behaviour is a field that combines on different disciplines such as psychology, sociology, and economics to explain the choices that consumer make. This thesis explores different approaches of consumer behaviour to management, in order to understand consumer behaviour in relation to sustainable development. Moreover we tried to use different approaches in order to see are they effective in helping people to live more sustainably (to recycle more, to save more energy and to eat healthier food). The most commonly used definition of sustainability and sustainable development comes from the 1987 Brundtland Commission report. Sustainable development is defined as “development that meets the needs of the current generation without compromising the ability of future generations to meet their needs.” (United Nations, 1987) According to Belz and Peattie (2009) sustainable consumer behaviour is consumers’ behaviours that improve social and environmental performance as well as meet their needs. Moreover it studies why and how consumers do or do not incorporate sustainability issues into their consumption behaviour and everyday life. Even though all of the progress and efforts that has been made globally toward addressing issues of sustainability, the problem of unsustainable consumption is growing. Many obstacles stand in the way of adopting sustainable behaviour whether material, financial or psychological. However small, everyday changes in people’s behaviour can have significant positive environmental impacts.

10

In literature, several models have been developed to investigate consumer behaviour. For instance, Ajzen developed the Theory of Planned Behaviour focusing on self-interest based and rational choice-based (1988; 1991). On the other hand Stern et al. (1999) has proposed the Value-Belief-Norm Theory (VBN) focusing on values and moral norms (Lopez et al., 2012). However, today it is widely accepted that consumer behaviour is the result of many factors and can be complex to understand. In fact, no single model or theory is able to provide a framework that can analyse more than a small portion of behaviour (Keirstead, 2006; Stephenson et al., 2010; Wilson and Dowlatabadi, 2007).

This thesis explores two different approaches of consumer behaviour: ‘nudge’ as a behavioural economics approach and “Attitude Behaviour Context” (ABC) theory.

1.1

“Attitude Behaviour Context” (ABC) theory

An effort to integrate different theories to predict environmental-friendly behaviour had been made by Stern (2000) and Guagnano et al. (1995) through the development of the “Attitude Behaviour Context” (ABC) theory which affirms that behaviour (B) is an interactive product of personal-sphere attitudinal variables (A) and contextual factors (C) In Chapter 2 we used “Attitude Behaviour Context” (ABC) in order to analyze the determinants behind individuals' decisions to adopt curtailment behaviour or to purchase energy saving products. Energy is a fundamental input for everyday consumer activities. Changing people’s behaviour in relation to energy consumption will be one of the most important challenges in the near future. Consumer behaviour is both complicated and difficult to change as they are influenced by a range of internal and external factors such as personal values, beliefs, norms, attitudes, and other people’s behaviour. Curtailment behaviour focuses on reduction in everyday energy use, such as lowering temperature in unused rooms or switching off the lights when leaving a room, and require either no or minimal structural adjustment (Barr et al., 2005). While behaviour based on adoption of energy efficient technologies is also called investment behaviour and is related to a purchasing decision (e.g., purchases of energy efficient light bulbs or change of insulation) (Gynther et al., 2012). Using data from 213 university students, we explored the influence of personal capabilities and moral norms, along with trust in information on energy saving actions provided by different entities on two energy saving behaviours. The results of the statistical model emphasise how

11

personal norms and trust in information provided by private companies, on the one hand, and family and friends, on the other, strongly influence the adoption of energy saving actions and curtailment behaviours.

1.2

‘Nudges’ and consumer behaviour

A growing literature on behavioural economics and psychology suggests use of non price interventions- nudges. A nudge is a ‘helping hand’ that will lead someone to make better decisions for itself and for the public interest as well. Nudges are suggested as a policy of libertarian paternalism and favoured for its simplicity, relatively low cost of implementation and its effectiveness. As suggested by (Thaler and Sunstein, 2008), 'libertarian' aspect refers to the necessity of respecting everyone's freedom to act, decide or even change their minds as it suits them. Nudges used in the field of ecology and environment saving, are called ‘green nudges’ or ‘ecological nudges’ (e.g., reducing the number of plastic bags, energy-saving). One example of ‘green nudge’ is reducing the number of plastic bags in China. Since 2008 in China stores are not providing customers with plastic bags at checkouts obliges them to ask for or even pay for them. According to Watts (2008) this measure has led to a reduction of around 40 billion plastic bags used between 2008 and 2009, representing a saving of 1.6 billion tonnes of oil. Nudges are also used to promote healthier eating habits. One example is removing the trays for people who eat at the self-service restaurant on a university campus. According to Oullier et al. (2010) this action has reduced the portions the students took for themselves and has reduced food wastage by an average of 50%. In two studies that we carried in Pisa and Strasbourg we used principles of nudges in order to see effect of nudges on consumer behaviour (Chapter 3 and Chapter 4). In the first paper we used ‘green nudges’ in order to test can nudges affect students’ green behaviour? In the second paper we used nudges to promote healthier eating habits. Moreover we tested can nudges affect healthier choices in a university cafeteria?

1.2.1 ‘Nudges’ and recycling - Can Nudges Affect Students’ Green Behaviour?

In Chapter 3 we study whether nudges are efficient in promotion of ecological behaviour-

12

recycling. Ecological behaviour is impeded both by financial and behavioural hurdles. A growing literature in behavioural economics and psychology suggests the use of non-price intervention nudges over other monetary incentives. We analyse whether nudges are indeed efficient in promoting recycling of resources among young people, and whether the combination of different types of nudges serve as better instruments. The study was performed on primary data from both a survey and field experiment conducted among university students in Pisa over a 60-day span (from October to December 2013). We collected data on 1849 instances of plastic cup recycling at a coffee vending machine at the Scuola Superiore Sant’Anna in Pisa. Recycling behaviour was measured by the number of plastic cups disposed in the proper dustbin, observed at the end of each day. Results of the experimental treatments showed a significant improvement in the amount of recyclable cups when a combination of nudges was applied. In addition to the empirical analysis, the paper further analytically replicates the results and illustrates the effect of a change in perception(awareness raising) of individuals, a shift in the social norm, as well as an ‘easy to do’ nudge.

1.2.2 ‘Nudges’ and healthy food- Nudging Students toward Healthier Choices in a University Cafeteria

Small everyday changes in people’s eating behaviour can have significant positive impact on our health. In Chapter 4 we study nudge and its effect on healthy food purchases in a university cafeteria. The study was performed on primary data; a field experiment was conducted among university students in Strasbourg. The field experiment was conducted over a 20-day span (from February to March 2014). In total, we collected data on 606 bottle of waters, 675 soft drinks, 339 fruit juice, 247 fruits, 257 salads, 227 desserts, 130 yogurts (without sugar), 193 yogurts (with sugar) in a cafeteria of School of Economics and Business School at the University of Strasbourg. Consumption of healthy food was measured by sale records of healthy food observed at the end of a day. Results of the experimental treatments showed a non significant impact on the amount of healthy food and drinks purchase.

13

References Ajzen, I. (1988). Attitudes, personality and behavior. Milton Keynes: Open University Press. Ajzen, I. (1991). The theory of planned behavior. Organizational behavior and human decision processes 50, 179-211. Barr, S., Gilg, A. W., & Ford, N. (2005). The household energy gap: examining the divide between habitual-and purchase-related conservation behaviours. Energy Policy, 33(11), 14251444. Belz, Frank-Martin & Peattie, Ken (2009) Sustainability Marketing: A Global Perspective. John Wiley & Sons, 73 Guagnano, G. A., Stern, P. C., & Dietz, T. (1995). Influences on attitude-behavior relationships a natural experiment with curbside recycling. Environment and behavior, 27(5), 699-718. Gynther, L., Mikkonen, I., & Smits, A. (2012). Evaluation of European energy behavioural change programmes. Energy Efficiency, 5(1), 67-82. Keirstead, J. (2006). Evaluating the applicability of integrated domestic energy consumption frameworks in the UK. Energy Policy, 34(17), 3065-3077. López-Mosquera, N., & Sánchez, M. (2012). Theory of Planned Behavior and the Value-BeliefNorm Theory explaining willingness to pay for a suburban park. Journal of environmental management, 113, 251-262. Oullier O., Cialdini R., Thaler R. and Mullainathan S. (2010), “Improving public health prevention with a nudge” Stern P. C., Dietz T., Abel T., Guagnano G. A., Kalof L. (1999). A value-belief-norm theory of support for social movements: The case of environmental concern. Human Ecology Review 6, 81–97. Stern, P. C. (2000). New environmental theories: toward a coherent theory of environmentally significant behavior. Journal of social issues, 56(3), 407-424. Stephenson, J., Barton, B., Carrington, G., Gnoth, D., Lawson, R., & Thorsnes, P. (2010). Energy cultures: A framework for understanding energy behaviours. Energy Policy, 38(10), 6120-6129.

14

Thaler, R. H., & Sunstein, C. R. (2008). Nudge: Improving decisions about health, wealth, and happiness. Yale University Press. United Nations. 1987. Report of the World Commission on Environment and Development, General Assembly Resolution 42/187, 11 December 1987. Retrieved: March, 2015 Watts J. (2008), “China plastic bag ban 'has saved 1.6m tonnes of oil’”, The Guardian, 22 May. Wilson, C., & Dowlatabadi, H. (2007). Models of decision making and residential energy use. Annu. Rev. Environ. Resour., 32, 169-203.

15

2 Determining factors of curtailment and purchasing energy related behaviours1 Abstract Changing people’s behaviour in relation to energy consumption will be one of the most important challenges in the near future. We analyzed the determinants behind individuals' decisions to adopt curtailment behaviour or to purchase energy saving products. Using data from 213 university students, we explored the influence of personal capabilities and moral norms, along with trust in information on energy saving actions provided by different entities on two energy saving behaviours. The results of the statistical model emphasise how personal norms and trust in information provided by private companies, on the one hand, and family and friends, on the other, strongly influence the adoption of energy saving actions and curtailment behaviours. The paper reveals the pivotal role of private companies in developing the market demand for energy-saving products by providing credible and scientifically-based information on environmental performance. The paper also contributes to strengthening the reliability of value-belief-norm theory and emphasizes the role of trust in information as a contextual factor that influences the adoption of a pro-environmental behaviour.

Keywords: energy-saving; green consumer; curtailment behaviour; personal norm; trust.

1

This is a joint project with Fabio Iraldo and Francesco Testa.

16

2.1. Introduction

One of the main challenges of the 21st century is to reduce the depletion of natural resources by human activities. Energy consumption produced by fossil resources is a principal cause of this impoverishment and a major source of carbon emissions (Tukker et al. 2006; Zhang and Cheng, 2009). The increase in income and well being in developed and emerging countries as well as the increased use and ownership of electric appliances (Soytas and Sari, 2003), has made energy efficiency a priority of policy makers. Several studies have shown that electricity consumption in private households could be substantially reduced if people paid more attention when buying more efficient electric appliances or by avoiding the unnecessary use of electricity (e.g., Gram-Hanssen et al., 2004). As the International Energy Agency concluded, there is a need for “a huge step-change in the attitudes to energy efficiency and consumer purchases by hundreds of millions of people worldwide…” (IEA, 2008). Energy consumer behaviour is, therefore, a key issue for scholars and practitioners from a wide range of scientific disciplines (Stephenson et al., 2010). Several models have been developed to investigate consumer behaviour. Ajzen developed the Theory of Planned Behaviour focusing on self-interest based and rational choice-based behaviour (1988; 1991). Stern et al. (1999) proposed the Value-Belief-Norm Theory (VBN) focusing on values and moral norms (Lopez et al., 2012). This theory is based on the principle that pro-social attitudes and personal moral norms are predictors of specific behaviour, such as environmental-friendly or energy saving behaviour (Jackson, 2005 as referenced in Martiskainen, 2007). Stern (2000) and Guagnano et al. (1995) have integrated different theories to predict environmental-friendly behaviour through the development of the “Attitude Behaviour Context” (ABC) theory, which affirms that behaviour (B) is an interactive product of personal-sphere attitudinal variables (A) and contextual factors (C). However, today it is widely accepted that consumer behaviour is complex and is the result of many factors. In fact, no single model or theory provides a framework capable of analysing more than a small portion of behaviour (Keirstead, 2006; Stephenson et al., 2010; Wilson and Dowlatabadi, 2007). Energy saving behaviour can be considered as a sub-set of more general environmentalfriendly behaviours. There are essentially two fundamental categories of behaviour: energysaving actions based on curtailment, and actions based on the adoption of energy efficient

17

technologies (Barr et al., 2005; Stern 1992; Sutterlin et al., 2011). Curtailment behaviour in the literature is also known as “habitual behaviour” (Maréchal, 2009). This type of behaviour focuses on the reduction of energy use in everyday life, such as by lowering the temperature in unused rooms or switching off lights when leaving a room, and requires no, or minimal, structural adjustment (Barr et al., 2005). Behaviour based on the adoption of energy efficient technologies on the other hand, is also called “investment behaviour” and is related to a purchasing decision (e.g., purchases of energy efficient light bulbs or change in insulation) (Gynther et al., 2012). Several studies have investigated energy-saving behaviours mainly focusing on the influence of attitudinal and personal factors on curtailment or purchasing behaviours, finding positive causal relations (Barr et al., 2005; Ek and Soderholm 2008; Gadenne et al., 2011; Hori et al., 2013; Oikonomou et al., 2009; Stern, 2000; Sutterlin et al., 2011). However, most of these studies have analyzed the predictors of curtailment or purchasing behaviours separately. Moreover, the level of trust in the source of information concerning the energy performance of products or energy-saving behaviour has been underestimated in the analysis of the contextual factors that can persuade individuals to adopt energy saving behaviours. Hence, in order to provide a valuable theoretical, policy and managerial contribution, it was investigated the role of trust, personal norms and personal capabilities (e.g. age, education, and income) in influencing both curtailment and purchasing behaviours of a sample of university students using data collected through a survey. The focus on university students in these types of studies is not uncommon in the literature. A growing literature relies on students’ responses and according to Cullis et al. (2012, p. 167) ‘there is no reason to believe that the cognitive processes of students are different from those of ‘real’ people’. Moreover, students play an important role in their family household by influencing their parents and other household members. Using data collected through questionnaires to 200 undergraduate students from a major private university in Malaysia, Chen and Chai, (2010) investigated the relationship between attitude towards the environment and green products. Their results revealed that consumer attitudes towards the government’s role and their personal norms regarding the environment, contributed significantly to their attitudes towards green products. Although the present study focuses on the same target audience (university students), similar to Stern (1999), it was extended the types of casual factors that can drive an individual to carry out two specific environmentally significant behaviours. Straughan and Roberts (1999) also collected data by distributing a questionnaire to a

18

convenience sample of 235 students attending a major university, in order to examine the dynamic nature of ecologically-conscious consumer behaviour. They focused on two elements of the VBN theory: the “self-efficacy” of consumer actions (perceived consumer effectiveness) and environmental awareness. They found that demographic criteria are not as useful a profiling method as psychographic criteria. Since the validity of VBN is largely supported in the literature (Stern, 2000), the present study focused on two important contextual factors which, as highlighted by Stern (1999), can play a significant role in determining environmentally significant behaviour: social norms and trust in sources that provide information. The paper is organized as follows. Section 2 provides an overview of the literature concerning the hypotheses of the study. Section 3 describes the data set and the estimation methodology. Section 4 then presents the statistical results and Section 5 makes some recommendations for future research and policy implications.

2.2. Theoretical framework and research hypotheses

The term “curtailment” (or “habitual”) behaviour encompasses a set of energy-saving actions that have to be performed rapidly and that are related to a change in the consumer’s everyday life, because they involve new habits in the use of energy (Aarts and Dijksterhuis, 2000; Marechal, 2009; Sutterlin et al., 2011). On the other hand, energy-saving behaviours based on energy-efficient measures (e.g. purchasing of energy efficient appliances) require a single action and occur occasionally — typically implying a change to a new technology or ”technology choice” (Stern, 1992). Purchases of energy efficient light bulbs or changes in insulation are some examples of purchase-related energy-saving behaviour. Stern (2000) divided the determinants of environmentally significant behaviour into four major categories: attitudinal factors (norms, beliefs and values), contextual forces (e.g., community expectations, advertising and government regulations), personal capabilities (sociodemographics: e.g., age or income) and habits or routines. The following sub-sections provide a brief overview of the literature and introduce the hypotheses of the study.

2.2.1 Attitudinal factors as determinant of energy-saving behaviour

Many studies have been carried out to clarify the key factors that influence energy-saving

19

behaviour (e.g., Oikonomou et al., 2009; Gadenne et al., 2011; Hori et al., 2013; Stern, 2000), highlighting that “personal moral norms are the main basis for individuals’ general predisposition for pro-environmental action” (Stern, 2000). Hori et al. (2013) carried out a survey in five major Asian cities, in order to identify factors that affect household energy-saving behaviour. Their results showed that global warming consciousness, environmental behaviour, social interaction and community-based activities significantly affected energy-saving behaviour. The results of a study carried out by Gadenne et al. (2011) showed that general environmental beliefs highly influenced norms on environmental actions, and emphasised a strong association between environmental attitudes and energy-saving behaviours. The main influence of attitudinal variables seem to be on specific stages of energy-saving behaviour. According to a review of US-based studies, attitudes are good predictors of general intentions to change residential energy use, however structural characteristics (of the residence) are better predictors of specific actions, such as weatherization (Guerin et al., 2000). Similarly, Oikonomou et al. (2009) found that people not only consider the comfort and costs of energy-saving, but also moral aspects such as environmental quality and impact on future generations. Based on the literature available, our aim was to further explore the effect of personal norms both on purchasing decisions and curtailment behaviours: H1-2: Consumers with strong personal norms related to energy-saving issues are more likely to purchase energy-saving products (1) and to adopt curtailment behaviours (2).

2.2.2 Contextual factors: the role of trust in energy-saving A second major type of causal variables is the contextual or external forces, which include interpersonal influences, community expectations, government regulations, monetary incentives and other legal and institutional factors (for an overview, see Stern, 2000). Contextual factors can impede pro-environmental personal attitudes from generating concrete actions. Although information is not directly included by Stern (2000) as a contextual factor (he explicitly mentions only the role of advertising), it can play a considerable role in supporting both curtailment and purchasing behaviour. Behaviours and actions regarding environmental protection and energy-saving are shaped not only by how individuals react to specific environmental issues, but also by information, the openness of society, and the attitudes toward the reliability of the source of information

20

(Tjernström and Tietenberg, 2008). Trustworthy information provided by external entities can make a social norm more pervasive (Stern, 1996) and compensate for a weak personal attitude towards environmental issues. Additionally, the energy and environmental attributes of a product are characterized by an asymmetrical distribution of information between the consumer and producer (Perrini et al. 2010). Therefore, how consumers perceive the reliability of information provided by companies on their product attributes, may have a significant influence on purchasing behaviours (Testa et al. 2013). The concept of trust has gradually acquired importance in both marketing and management research (Schoorman et al., 2007) and has proven to be an effective key in analyzing situations where the truster (i.e. the consumer in our case) is vulnerable (Castaldo et al., 2009). Trust can be defined as the truster’s expectation that the trustee (i.e. a producer in our case) is willing to keep promises and fulfil obligations (Hagen and Choe, 1998). The expectation is based on such variables as the level of competence, honesty, altruism, and goodwill of the trustee (Blomqvist, 1997). According to Castaldo et al. (2009) trust is multidimensional and can be applied across different levels of analysis (interpersonal, intergroup or inter-organizational). Although relationship of trust with energy-related issues has gained the interest of scholars, researchers and policy makers (Mitchell et al., 2010; Rayner, 2010), the focus on behaviours has been very limited. For instance, Rayner (2010) looked at diverse concepts and roles of trust in the fields of energy and environmental policy research: public trust in science, institutional trust in technology choices, and the idea that high-trust societies are more sustainable than those exhibiting low-trust. Numerous studies have also analyzed the importance of trust in the field of service provision (Price and Arnould, 1999; Geyskens et al., 1998) and in energy technologies (e.g., Ashworth et al., 2011). The influence of trust in the energy provider on customer loyalty has been investigated (Ibáñez et al., 2006) but mainly focusing on the effect of the perceived trust and switching costs on customer loyalty in residential energy markets. Consumers receive information regarding energy-saving from different entities: government, local authorities, EU commissions, NGOs, scientists, private companies, the media, friends and family. Trust in information received by an individual plays an important role in this process and could determine consumer responses to the energy-saving information they receive from various entities. Some research has investigated the relation between the concept of trust in information and green consumption (e.g., Bonini et al., 2008; Darnall et al., 2012). Studying a sample of more

21

than 1,200 UK residents, Darnall et al. (2012), found evidence that consumers who have greater trust in information provided by governments, environmental NGOs, and friends/family are more likely to rely on eco-labels in their product purchases. Additionally, according to Bonini et al. (2008) businesses must act on global warming and other environmental issues to narrow the trust gap between them and the public. Whereas the literature tends to focus on environment-related behaviour, this study concentrated on the trust in information on energy-saving issues provided by governments, local authorities, the European Commission, NGOs, scientists, private companies, friends and family. It is investigated the effect of trust not only on purchasing decisions, but also on the adoption of curtailment behaviours.

2.2.2.1 The governments

The government is responsible for establishing energy laws, developing environment protection policies and distributing information that directly or indirectly affects energy saving. Literature related to energy consumption and trust in the government is still not abundant, and only a few studies on the role of trust in the fields of energy and environmental policy have been conducted (e.g., Mitchell et al., 2010; Rayner, 2010). However, Margaret Walls, one of the energy experts for The Wall Street Journal, suggested that government should focus more on behavioural approaches and provide more information to energy users in order to make them to save more (Ball, 2013). Her idea is that governments should concentrate on information programs that include product labels, such as the "Energy Guide" on appliances; voluntary certification programs such as Energy Star; energy audits; and other programs focusing on making energy uses and costs more transparent (Ball, 2013). This leads us to formulate the following two hypotheses: H3-H4: Consumers with greater trust in information on energy saving actions provided by governments are more likely to purchase energy-saving products (3) and to adopt curtailment behaviours (4)

2.2.2.2 Environmental NGOs

Environmental NGOs play an important role in energy-saving and environmental activities. NGOs have established different working relationships in order to exchange information and collaborate on issues related to energy-saving and environmental protection (Gan, 2000).

22

Through formal and informal networks, NGOs shape the attitudes and operations of other social institutions (Gan, 2000). Environmental NGOs help consumers by protesting publicly against labels that fall short of environmental expectations (Rivera and de Leon, 2004). Like governments, environmental NGOs also help to protect customers from false market claims, e.g. by developing eco-labels and eco-label guidelines (Rex and Baumann, 2007). This leads us to formulate the following hypotheses: H5-6: Consumers with greater trust in information on energy saving actions provided by NGOs are more likely to purchase energy-saving products (5) and to adopt curtailment behaviours (6)

2.2.2.3 Private companies

Companies can differentiate themselves from their competitors by acting on environmental and other social issues, which can help them to build trust among their consumers. However, issues connected with “greenwashing“ make customers confused and disoriented regarding the environmental claims that companies provide (Mayer et al., 1993). Although companies increasingly make use of green claims in advertising their products (Testa et al., 2011), consumers often believe that these claims are not reliable and thus do not orient their purchasing decisions towards greener products. Greenwashing has increased consumer distrust and reduced consumers’ willingness to “buy green” (Peattie and Crane, 2005), and has created barriers towards encouraging a broader societal change (Knott et al., 2008). Based on a study by McKinsey (Bonini et al. 2008), awareness promotion is critical for companies, insofar as consumers are increasingly willing to “do business” with companies only if they trust them to perform well in terms of societal and environment issues. In other words, performing concrete actions towards sustainability increases the corporate reputation and the level of trust by consumers, as well as their propensity to buy green products. Using an extensive dataset of consumer choices Testa et al., (2013) found that some ecolabels are able to provide reliable messages to consumers and encourage them to make environmental friendly purchasing behaviours. In order to contribute to the current debate on the role of trust in information provided by private companies on energy- savings behaviours, the following hypotheses are formulated: H7-8: Consumers with greater trust in private companies who provide information on the energy efficiency of their products are more likely to purchase energy- saving products (7) and to adopt curtailment behaviours (8)

23

2.2.2.4 Friends and family

A number of studies have investigated the role of other social actors on an individual’s choice to adopt energy-saving behaviour. For instance, Ek and Soderholm (2008) found that perception regarding the behaviour of others in general affects individual moral norms and ultimately contributes to determine a specific behaviour. Friends and family are the most trusted individuals in our social network; at the same time they are very frequently reported as trusted sources of motivation for green purchasing (Lee 2008; Young et al., 2010). The literature suggests that consumers are favourably influenced by the opinions and actions of their family and friends (Pickett-Baker and Ozaki, 2008; Sidiras and Koukios, 2004). This leads us to formulate the following hypotheses: H9-10: Consumers with greater trust in information on energy saving actions provided by friends and family are more likely to purchase energy-saving products (9) and to adopt curtailment behaviours (10)

2.2.3 Personal capabilities

Personal capabilities include both the knowledge and skills required for specific actions and the more general capabilities and resources (such as money). Personal capabilities are usually measured by means of sociodemographic variables such as age, education, and income (Stern, 2000). Many studies have investigated the role of sociodemographic variables as predictors of environmental behaviours, and have found contrasting results. A few studies identify the typical “energy saver” as young, female, with high level of education, and wealthy, (Roberts, 1996; Sardianou, 2007). A number of past studies (Roberts, 1995; 1996; Zimmer et al., 1994) have shown that younger individuals are more likely to be sensitive to environmental issues. Conversely, results from other studies (Stern, 1999, Testa et al., 2013) show that demographic criteria were found to be unrelated and not useful for profiling college students based upon ecologically-conscious consumer behaviour. Income is generally thought to be positively related to energy-saving behaviour. Numerous studies have addressed the role of income as a predictor for ecologically conscious consumer behaviour (Zimmer et al., 1994), whereas fewer studies have found a negative relation between income and environmental concerns (Roberts 1995; 1996). The level of education is another demographic variable that has been related to energy-saving behaviour (Roberts 1995; 1996).

24

Hence, in order to contribute to the current debate, our aim was to further investigate energysaving behaviours, and focus on the role of personal capabilities in individual choices. H11-12: Consumers with higher personal capabilities are more likely to purchase energysaving products (11) and to adopt curtailment behaviours (12)

2.3. Methods

The study was performed on primary data from a survey conducted among university students in Pisa, one of the most important university cities in Italy. The study instrument was a 3-page questionnaire that posed questions concerning energy-saving behaviour. The questionnaire was composed of four sections. Section I assessed the participants’ energy-saving behaviour (curtailment). Section II measured different energy related beliefs, including personal norms, awareness of consequences and ascription of responsibility. In Section III behaviours toward the purchasing of energy-saving appliances were assessed. Section IV invited participants to answer questions about socio-demographics (See Appendix for more details). Data were collected between May and June 2013. Prior to the final submission, a pre-test was administrated to 30 students during the month of May. This test was developed to reveal any possible weaknesses and misunderstandings arising from the text. Consequently, the final questionnaire was prepared adjusting the pre-test drawbacks, summarizing and changing the statements of some of the questions, and eliminating some questions. Using the mailing list provided by the university administrative departments, 450 emails were sent to university students in Pisa including the survey link and a description of the aim of the study. The response rate after two reminders was 47%. The group of participants included 120 males and 86 females. Over 45% of participants were graduates. The highest percentage of students (around 40%) was in the range of 26-29 years. Approximately 50% of the sample were living in apartments, while 43% were renting a room, and the rest were in university dorms. In order to overcome methodological biases based on survey techniques, several procedural remedies were adopted. Because many researchers have highlighted social desirability as one of the most common sources of bias affecting the validity of experimental and survey findings (King and Bruner, 2000; Tourangeau and Yan, 2007) anonymity of respondents was guaranteed. It was also investigated common method variance by performing Harman’s single-factor test, which included all the variables in an exploratory factor analysis. A single factor accounting for the majority of covariance among the variables indicates the common

25

method variance. The test revealed that no single factor accounted for the majority of variance in the variables.

2.3.1 Measurements

For the purposes of our study, the energy-saving behaviour was measured from a twofold perspective, in accordance with Barr et al. (2005). First, purchasing decisions were measured using four different questions able to reflect the attitude or behaviour of an individual towards energy- saving products. The students were asked the following questions: 1) When buying electrical appliances, if I could choose between energy-saving and conventional products, I would prefer energy-saving products; 2) I try to buy products that save more energy; 3) I buy high efficiency light bulbs to save energy. For each of these three behaviours, respondents reported “Always”=5, “Almost always”=4, “Often”=3, “Rarely”=2, or “Never”=0 (See Appendix for more details). The responses of the three purchasing behaviours were entered into a common factor analysis and a reliable factor emerged to account for purchasing behaviour (Cronbach’s Alpha = 0.8618). Second, energy-saving actions based on curtailment behaviour were measured as in Sutterlin et al. (2011). Three everyday actions were listed and participants were asked how often they carry out the following activities: 1) I turn off the light upon leaving a room; 2) I adjust room temperature according to room usage; 3) I turn off standby appliances (e.g., TV, PC). For each of these three actions, respondents reported “Always”=5, “Almost always”=4 “Often”=3, “Rarely”=2, or “Never”=1. The responses were summed to obtain an overall curtailment behaviour index, which accounted for both the frequency and amplitude of an individual’s energy-saving behaviour.

2.3.1.1 Independent variables 2.3.1.1.1 Level of trust in the information provided by different entities

Trust is a complex and multidimensional concept that can be applied across different levels of analysis in the field of energy consumption and, as a consequence, measured in several ways (Price and Arnould, 1999; Geyskens et al., 1998, Ashworth et al. 2011). For instance, the importance of trust in the service provider was investigated by Price and Arnould (1999). Hartmann and Ibanez (2007) investigated the impact of energy branding using two constructs:

26

familiarity with brand and its trustworthiness. Ashworth et al. 2011 investigated public trust in energy technologies. While some studies ask general ‘trust’ questions using various methods such as experiments, interviews (Glaeser et al. 2000), some go beyond the general and focus on specific ‘trust’ behaviours. Similarly Rahbar and Abdul Wahid (2011) measured trust in eco-labels and ecobrands by asking the following: “I am doubtful about the above logo” and

“I am doubtful

about the eco-brand”. According to Rahbar and Abdul Wahid (2011) customer trust in ecolabels and ecobrands and their perception of ecobrands show a positive and significant impact on their actual purchase behaviour. Similarly, the concept of trust has been analysed from a different perspective, that is the reliability of information provided by different entities that are directly and indirectly related to energy-saving issues. According to Sayogo et al. (2014), trust in the information regarding product and certification is crucial for the adoption and use of smart disclosure tools that make use of such information. They investigated the determinants of trust in sustainable product information through a survey administered in Mexico and the United States, and found that the reputation of brands and certificates are important in developing trust. Following Darnall et al. (2012), the level of trust was measured by asking: “How much do you trust the following bodies in providing you with reliable information on energy-saving actions”. Respondents indicated the level of trust in local authorities, national governments, the European Commission, environmental NGOs, scientists and friends/family using a 5 point Likert scale (“No trust at all”=1, “Little trust”=2, “Neither”=3, “Trust a little”=4, “Trust wholly”=5). The responses in the three public institutions were entered into a common factor analysis and one reliable factor emerged to account for trust in public institutions (Cronbach’s Alpha =0.8276). This factor measures the extent to which the information provided by several entities are perceived as credible and reliable by interviewees. Additionally, the trust in information provided by private companies was measured by asking “How much do you trust private companies that provide information on the energy efficiency of specific appliances”. Respondents replied using the above mentioned Likert scale.

2.3.1.1.2 Personal norms

Subjective norms are widely considered as a relevant predictor of environmental behaviours. Values, norms, and beliefs play a significant role in determining the actions of an individual regarding energy-saving. Since there is a causal order between value, belief and personal

27

norms (Stern, 2000), and many studies have empirically demonstrated the reliability of VBN theory (Stern et al. 1999), this study focused on the personal norms that influence the adoption of an environmentally significant behaviour. Personal norms were measured by asking respondents to express their level of agreement with the following four assertions: i) I pay attention to energy consumption because I care about the environment; ii) I have a responsibility to contribute to environmental preservation by using energy-saving products; iii) I do not feel good when energy is consumed unnecessarily in the household (e.g. leaving lights on in an unused room); iv) I feel personally obligated to avoid unnecessary energy consumption wherever possible. For each of these four assertions, respondents reported “Strongly agree”=5, “Agree” =4 “Neutral”=3, “Disagree”=2, or “Strongly disagree”=1. The responses of the four assertions were entered into a common factor analysis and one reliable factor emerged to account for purchasing behaviour (Cronbach’s Alpha =0.7953).

2.3.1.1.3 Personal Capabilities

Because of the analysis of factors influencing purchasing choices and energy-saving behaviour also involves the consideration of various personal capabilities, a set of variables was included that could affect the frequency and amplitude of the energy-saving actions by individuals. Since many studies have found that the personal characteristics of an individual can influence an individual’s environmental consciousness and, therefore turn into an energysaving behaviours (Karp, 1996; Mostafa, 2007; Tilikidou and Delistavrou, 2008; Chen and Chai, 2010), variables measuring the age of the respondent, his/her level of education and gender were included. Additionally, since the level of income may affect the decision to adopt curtailment activities (Zimmer et al., 1994; Darnall et al., 2012), three different variables were included in the model: level of household monthly income (0-1000€; 1000€-2000€; 2000€3500€; 3500€-5000€; above 5000€) the main source of income (family assistance; loan; scholarship; salary), the role of financial resources in inducing specific behaviours (level of agreement - from strongly disagree to strongly agree - to the following sentence: I primarily pay attention to energy consumption in the household for financial reasons). Finally, the political and religious orientation of the respondent (Costa and Kahn, 2010) and his/her nationality were measured. The descriptive statistics and correlations for the study variables are summarized in Table 2.1

28

Table 2.1: Correlation matrix and descriptive statistics (*, **, and *** indicate the significance at the 10%, 5%, and 1% levels, respective) 1)

2)

3)

4)

5)

6)

7)

8)

9)

10)

11)

12)

13)

14)

15)

16)

17)

1) Purchase

1.00

2) Curtailment

0.41***

1.00

3) Personal norms

0.53***

0.52***

1.00

4) Trust institutions

0.35***

0.15**

0.27***

1.00

5) Trust NGOs

0.30***

0.18**

0.29***

0.58***

1.00

6) Trust family and

0.09

0.18**

0.03

0.06

0.19***

1.00

7) Trust private sector

0.34***

0.11

0.23***

0.41***

0.26***

0.05

1.00

8)Financial motives

0.10

0.21***

0.003

0.002

-0.08

-0.01

0.06

9) Age

-0.21***

-0.11

-0.22***

-0.17**

-0.13**

-0.03

-0.07

-0.11

1.00

10) Gender

-0.21***

-0.07

-0.18***

0.06

-0.07

-0.18***

0.05

-0.09

0.03

1.00

11) Education

-0.17**

-0.13*

-0.14**

-0.05

-0.07

0.04

-0.12*

-0.01

0.29***

0.10

1.00

12) Area of study

-0.09

0.0006

-0.17**

-0.11

-0.02

0.06

-0.03

0.07

0.11*

-0.01

0.06

1.00

13) Nationality

0.01

-0.04

0.05

-0.15**

-0.18***

-0.05

-0.03

-0.01

0.02

0.04

0.14**

-0.08

1.00

14) Source of income

-0.11

-0.06

-0.05

-0.07

-0.05

-0.03

-0.12*

-0.07

0.44***

0.08

0.36***

0.02

0.018

1.00

15) Family income

0.19***

0.24***

0.23***

0.05

0.08

-0.07

0.02

0.10

-0.14**

-0.09

-0.22***

0.07

-0.20***

-0.08

1.00

0.09

-0.01

0.03

0.05

0.09

-0.04

0.008

-0.09

0.15**

-0.06

0.01

0.14 **

-0.008

0.07

0.10

1.00

17) Religious person

0.08

0.11*

0.05

0.01

-0.02

-0.004

0.16**

0.05

0.01

0.05

0.05

-0.06

-0.13*

0.15**

0.06

0.08

1.00

Mean

6.30

0

0

0

2.17

2.09

2.44

2.13

3.12

1.41

2.29

3.04

2.93

2.89

2.76

2.34

1.60

Standard deviation

3.13

.92

.89

.88

1.06

0.96

1.10

.91

1.04

0.49

0.68

1.90

2.48

1.02

1.14

.49

Min

0

-1.30

-1.29

-1.51

1

1

1

1

1

1

1

1

1

1

1.14

1

1

Max

14

2.64

3.46

2.14

5

5

5

5

6

2

3

7

8

4

1

4

2

N

213

213

213

199

199

200

200

213

206

206

206

206

206

206

206

206

205

friends

16)

Conservative

or

1.00

liberal

29

2.4 Empirical Models

In order to test our hypotheses, two equations were constructed with green purchasing behaviours and curtailment behaviours as dependent variables. Figure 2.1 shows the relation between the dependent and independent variables tested in both equations and the related hypotheses.

Figure 2.1: Conceptual model and Hypotheses Since the nature of the dependent variable is different (green purchasing behaviours is a continuous variable whereas curtailment behaviours is categorical), two statistical techniques were applied. To evaluate the determinants of purchasing behaviours, an ordinary least squares (OLS) regression technique was used. In contrast, an ordinal logistic regression was performed due to the categorical nature of the dependent variable “curtailment behaviours” . In order to check the feasibility of applying the two statistcal techniques, it was verified that the assumptions underlying the OLS and ordinal logistic regression were met by the equations used to test the hypotheses of this study. Regarding the equation with green purchasing behaviours as the dependent variable, the normality of residuals required for valid hypothesis testing was checked by plotting the non parametric Kernel density estimator (Fan and Gencay, 1995), which revealed the symmetry

30

of residual distribution. Secondly, the homogeneity of variance of the residuals was verified by the Breusch-Pagan test, which is one of the main assumptions for the OLS regression (Coin, 2006). The null hypothesis that the variance of the residuals is homogenous was not significant, thus so it is possible to assume that there was no heteroskedasticity. Finally, a regression specification error test was performed for omitted variables (Ramalho et al., 2011), which revealed the absence of model specification errors. Regarding the second equation with the dependent variable “curtailment behaviours”, the assumptions were positively tested that the cumulative odds ratio for any two values of the covariates was constant across response categories (Peterson and Harrel, 1990). A likelihood ratio test was applied where the null hypothesis was that there was no difference in the coefficients among models. The presence of collinearity in both equations was also checked by computing the tolerance and variance inflationary factor (VIF) for all variables. Low variance inflation factors (< 2.0) and a VIF less than 5 revealed that that multicollinearity was not present in our empirical model (O’Brien, 2007).

2.5 Results

In order to test our hypotheses, since energy-saving behaviour was measured from different perspectives (Barr et al., 2005; Suterllin et al., 2011), two separate models were constructed: a Curtailment energy- saving model (Model 1) and a Purchase-related energy-saving model (Model 2) (Table 2.2).

31

Table 2.2 Results of regression analysis MODEL 1Purchase

MODEL 2energy-

Curtailment energy saving

saving Coef.

SE

Coef.

SE

Trust institutions

.1250

.0848

.0710

.1240

Trust NGOs

.0175

.0679

.0873

.0989

Trust family and friends

.0536

.0600

.2413***

.0885

Trust private sector

.1664***

.0618

-.1175

.0911

Personal norms

.4471***

.0724

.8224***

.1178

Financial motivation

.1039*

.0620

.4132***

.0941

Age

-.0574

.0635

.1383

.0938

Gender –Female (compared to male)

-.1004

.1264

.1873

.1844

Education

-.0481

.0990

-.0976

.1464

Variable Trust of sources to provide information

Personal capabilities

Area

of

study-

Engineering

(compared

to

.2879

.1741

-.1423

.2543

study-

Humanities

(compared

to

.1350

.2415

.4179

.3559

Management

(compared

to

.2846

.1889

.1552

.2769

Area of study- Medicine (compared to economics)

.1318

.3476

.2517

.5156

Area of study- Natural science (compared to

.4606

.2280

-.1179

.3309

-.1464

.2412

.246

.3506

Nationality-Other European (compared to Italian)

-.1995

.1886

-.3155

.2971

Nationality-African (compared to Italian)

.1743

.3419

-.3907

.5000

Nationality-American (compared to Italian)

.0810

.4034

-.2883

.5850

Nationality-Asian (compared to Italian)

.1309

.1689

-.0652

.2480

Nationality-Middle Eastern (compared to Italian)

-.0150

.3061

.0727

.4438

.2303

.2013

-.1630

.2992

-.9069*

.4890

.5674

.7128

.0521

.1666

-.1493

.2458

-.0442

.1958

-.1600

.2880

economics) Area

of

economics) Area

of

study-

economics)

economics) Area of study- Other

disciplines (compared to

economics)

Nationality-Other

nationalities

(compared

to

Italian) Source of income-Loan (compared to Family assistance) Source of income- Scholarship (compared to Family assistance) Source of income-Salary (compared to Family

32

assistance) Family income

.0228

.0539

.1317*

.0792

Political orientation-Conservative (compared to

.0858

.2365

-.1844

.3475

.0793

.1347

-.0922

.1975

.2875*

.1670

-.2687

.2476

Religious person

.0518

.1217

.1908

.1801

Constant

-.8865*

.4831

N

198

198

LR chi2

--

***

F Test

***

--

Pseudo R2

--

0.1837

R-squared

0.4679

--

liberal) Political orientation-Somewhere in the middle (compared to liberal) Political orientation-None of them (compared to liberal)

First of all, Hypotheses 1 and 2 are supported, therefore, it is possible to state that consumers with strong personal norms related to energy-saving issues are more likely to purchase energy-saving products and to adopt curtailment behaviours. The results show that personal norms are positively and statistically significant (p 0.05. We conclude that mean of variables water, fruits, salads (small portion), yogurt and desserts before treatments is not significantly different from their mean after the treatment. However p-value associated with the t-test in variable salads (large portion) is small, it is less than 0.05 (salads (large portion) p-value is .002) there is evidence that the mean is different from the hypothesized value. It means that mean salads (large portion) before and after treatment changed. Number of salads (large portion) sold before the treatment decreased from 11.5 to 8.7 portions. In this case that nudge did not have positive effect and it did not increase number of salads (large portion), contrary sales decreased. Moreover we can conclude that nudge treatments did not have a significant impact on sales of water, fruits, salads, yogurts and desserts. However p-value associated with the t-test in variables soft drinks and fruit juice is small, it is

-

79

less than 0.05 (soft drinks p-value is .013, fruit juice p-value is .010) there is evidence that the mean is different from the hypothesized value. It means that mean of soft drinks and fruit juices before and after treatment changed. Number of less healthy drinks, soft drinks (e.g., Coca Cola, Sprite, Fanta) sold before the treatment increased from 542 to 675 pieces. This was an increase of 24.5 % and we can say that nudge did not have positive effect and it did not reduce number of soft drinks, contrary sales increased. Sales of healthy drinks such as fruit juice increased from 25.7 to 33.9 ( see Table 4.2). Table 4.2: T test statistics-drinks Variable Water Soft drinks Fruit juice

MeanTreatment 60.6 67.5 33.9

MeanControl 56.6 54.2 25.7

Tstatistics 0.7499 3.0597 3.2111

Pvalue 0.47 0.01 0.01

MeanTreatment 24.7 25.7 8.7 13

MeanControl 25 21.8 11.5 18.2

Tstatistics -0.1307 1.5950 -2.7097 -1.5631

Pvalue 0.89 0.14 0.02 0.15

19.3

19.3

0.0000

1

22.7

24.2

-0.4691

0.65

Table 4.3: T test statistics-food Variable Fruits Salads Small Salads Large Yogurt-without sugar Yogurt-with sugar Desserts

Our results can be seen as evidence that nudges do not always work out as planned. Our results therefore do not lend support to our hypothesis H1: Using a non-price intervention nudge (social norm) combined with an awareness-raising message and an 'easy to choose’ nudge - positively affects consumption of healthy food in the university cafeteria.

4.4.1 Why nudge do not always work out as planned?

Nudges help individuals with various decision-making flaws to eat healthier, to live longer

80

and better live. However some studies on nudges indicate that they do not always work out as planned. Rolls et al. (2007) found that altering plate sizes had no significant effect on energy intake at meals eaten in three laboratory experiments. Participants made significantly more trips to the buffet when they were given the smallest plate in one of these experiments. Adding “healthy options to “unhealthy” meals might be challenging. Psychologists also report “negative calorie illusion,” whereby adding a healthy option to weight-conscious individuals’ unhealthy meals decreases their perception of the meals’ calorie content (Marlow, 2014). Sometimes encouraging the adoption of a healthier lifestyle among overweight individuals, promoting the consumption of healthy foods might end up facilitating calorie overconsumption, leading to weight gain rather than weight loss (Chernev, 2011). Labelling requirements are introduced to help people to reduce calories and other food attributes (fat, sugar) (Marlow, 2014). However studies have found that labelling improves calorie estimates (Elbel, 2011), but evidence so far does not clearly demonstrate that required labels result in healthier eating. Elbel et al. (2009) examined the influence of menu calorie labels on fast food choices in the New York City’s labelling mandate. Elbel et al. (2009), found no change in calories purchased after the law. Similar to Elbel et al. (2009) findings, our results showed that putting labels on a healthy food such as salads did not help us to increase number of products sold during the treatment period.

We introduced a series of green footprints leading to shelves in the hope of encouraging people to take the healthier option. Green footprints on the floor had the same message as labels on the food “eat healthy”. However introducing nudge, that we called ‘easy to choose’ with green footprint did not have significant effect on sales of healthy and less healthy drinks and food.

In partnership with the local government Hansen did similar experiment, they tested two potential “social nudges” using green footprints and green arrows to try to influence choices (Hansen, 2012). In the first experiment they used green arrows pointing to stairs next to railway-station escalators, in order to encourage people to take the healthier option. Results showed that it had almost no effect. However for the second experiment they used green footprints leading to rubbish bins and this reduced littering by 46% during a controlled experiment.

81

In his work Hansen says: “There are no social norms about taking the stairs but there are about littering.” (Hansen, 2012) Bollinger et al. (2010) studied the impact of mandatory calorie posting on consumers’ purchase decisions, using detailed data from Starbucks. They found virtually no change in purchases of beverage calories. The field study replicated these findings; labelling healthy food did not lead to higher sales of healthy food. Sales of healthy and less healthy food and drinks were not impacted by manipulations. Our results showed that nudges do not always work out as planned, that sale of less healthy drinks did not decreased; contrary they increased by 25 percent. Our results similar to Bollinger et al. (2010) showed no significant change in purchase of healthy food and drinks during nudge treatment period.

From these above mentioned studies we can notice there is a number of studies that show no effect of nudges on healthy food consumption by some other means.

4.5 Conclusion In this paper we presented field experiment in a university cafeteria that examines the effect of combination of nudges social norm and ‘easy to choose’ on healthy and less healthy food and drink consumption. Going beyond existing literature, we studied the joint effect of a combination of nudges. We introduced two nudges at the same time; first we triggered a behavioural change via two different effects: awareness raising and an externally imposed norm. Additionally we used green footprints and labels encouraging people to take the healthier option. The results showed that combination of nudges used did not change consumers’ choice in a healthier direction. Moreover sale of the less healthy drinks increased by 25 percent during our treatment period (from 542 to 675). Number of salads, fruits and healthy drinks sold did not change significantly. Comparing our findings to the results from Thunström and Nordström (2013) we can say that our findings are similar since they also found no impact on sales or the market share of the healthy labelled meal from the nudge used in their study. There are several possible explanations for these findings. One of the reasons might be differences in culture. Olivier Oullier, a behavioural and brain scientist who advises the

82

French government, says: “The French have a tendency not to comply as easily with perceived social norms the way Anglo-Saxons would,” (The Economist, 2012). Moreover “Telling someone in France that their neighbour is using less electricity or saving more water is not sufficient.” (The Economist, 2012). Our results also show that informing students in France that the majority of other students at the world leading university consume healthy food was not sufficient. Another reason might be that the customer base of the field experiment consists of consumer group students that are very sensitive to money. One of the reasons why consumers choose less healthy food and drinks might be due financial reasons. But in our case prices of healthy and less healthy drinks were the same (Coca Cola 335 ml price 1.10 euros and Minute Maid Orange 335 ml price 1.10 euros. Moreover price of a healthy yogurt (without sugar and artificial aroma) was 0.45 euros compared to less healthy yogurt (with sugar and additives) 0.65 euros. Still majority of students chose less healthy yogurt (on average 13 healthy and 19.3 less healthy yogurts per day were sold during treatment period).

Some limitations should be noted about our study. One of the limitations is a relatively small number of healthy food available comparing to less healthy. Moreover fruits had a lower price comparing to desserts which may have given these products an additional benefit. The efficacy of nudge interventions could be studied over a longer time frame in order to give more realistic results.

Our results show that a nudge was not able to influence significantly consumers’ healthy food purchases. Nudges alone may not be the best solutions to encourage people to eat healthier. However nudges in combination with some other tools such as; increase of assortment, reduction of prices of healthy food, introduction of convenient lines, all together can result in winning combination. According to van Kleef et al. (2012) increase the prominence of healthy food in canteen by enlarging their availability, while permitting access to unhealthy food, might me a promising strategy to promote sales of healthy food.

The examples from the literature review section illustrate how small changes in the environment can lead to major positive effects on health and economics and they can be used in public health prevention strategies. In our opinion nudge brings additional policy tools into play that in combination with some other traditional tools (awareness campaigns, education about healthy food) are required to change consumer behaviour. However time will show can

83

‘nudges’ convince policy makers and administrations to consider them to improve the wellbeing of individuals.

An interesting topic for further research would be identification of other important factors that nudge consumers towards healthier food choice in various environments such as restaurants and grocery stores. More research is needed to analyse long-term effects of nudges on healthy food purchase in various environments.

84

References Ajzen, I. (1985). From intentions to actions: A theory of planned behavior (pp. 11-39). Springer Berlin Heidelberg. Asch, S. E. (1951). Effects of group pressure upon the modification and distortion of judgments. Groups, leadership, and men, 222-236. Ball, K., Jeffery, R. W., Abbott, G., McNaughton, S. A., & Crawford, D. (2010). Is healthy behavior contagious: associations of social norms with physical activity and healthy eating. Int J Behav Nutr Phys Act, 7(1), 86. Bandura, A. (2001). Social cognitive theory: An agentic perspective. Annual review of psychology, 52(1), 1-26. Bollinger, B., Leslie, P., & Sorensen, A. (2010). Calorie posting in chain restaurants (No. w15648). National Bureau of Economic Research. Chernev, Alexander, “The Dieter’s Paradox,” Journal of Consumer Psychology 21, 2:178–83, 2011. Cooper Leigh, Manuel Calatrava Conesa, (2013). A fruitful nudge towards a healthy diet. Available at: http://www.inudgeyou.com/a-fruitful-nudge-towards-a-healthy-diet/ [Accessed: 20 Oct 2014]. Downs, J. S., Loewenstein, G., & Wisdom, J. (2009). Strategies for promoting healthier food choices. The American Economic Review, 159-164. Elbel, B., Kersh, R., Brescoll, V. L., & Dixon, L. B. (2009). Calorie labeling and food choices: a first look at the effects on low-income people in New York City.Health Affairs, 28(6), w1110-w1121. Elbel, B. (2011). Consumer estimation of recommended and actual calories at fast food restaurants. Obesity, 19(10), 1971-1978. Elster J. (1989) Social norms and economic theory. Journal of Economic Perspectives 3(4):99– 117. Feng, J., Glass, T. A., Curriero, F. C., Stewart, W. F., & Schwartz, B. S. (2010). The built environment and obesity: a systematic review of the epidemiologic evidence. Health & place, 16(2), 175-190. Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behavior: An introduction to

85

theory and research. Hanks, A. S., Just, D. R., & Wansink, B. (2013). Smarter Lunchrooms can address new school lunchroom guidelines and childhood obesity. The Journal of pediatrics, 162(4), 867-869. Hanks, A. S., Just, D. R., Smith, L. E., & Wansink, B. (2012). Healthy convenience: nudging students toward healthier choices in the lunchroom. Journal of Public Health, 34(3), 370-376. Hansen, P.G. & Jespersen, A.M. (2012). Nudge, adfærdsøkonomi, og ‘økonomisk psykologi’ – fra eksperiment til skraldespand. Psykologisk set, vol. 87-88., p. 15-23. Available at: http://www.inudgeyou.com/green-nudge-nudging-litter-into-the-bin/ Harris, J.L., Pomeranz, J.L., Lobstein, T. & Brownell, K.D. (2009) A crisis in the marketplace: how food marketing contributes to childhood obesity and what can be done. Annual Review of Public Health, 30, 211–225. Halpern D. (2005) Social capital. Cambridge/Malden, MA: Polity Press; 2005. Just, D. R., & Wansink, B. (2011). The flat-rate pricing paradox: conflicting effects of “all-youcan-eat” buffet pricing. The Review of Economics and Statistics, 93(1), 193-200. Just, D. R., Mancino, L., & Wansink, B. (2007). Could behavioral economics help improve diet quality for nutrition assistance program participants?. USDA-ERS Economic Research Report, (43). Loewenstein, G., Brennan, T., & Volpp, K. G. (2007). Asymmetric paternalism to improve health Long, J,( 2015) Get More Done by Following 'The Productivity Diet' (Infographic) Available at: http://www.entrepreneur.com/article/247711 [Accessed: 26 June 2015]. Marlow, M. L. (2014). Weight Loss Nudges: Market Test or Government Guess?. Working Paper, Mercatus Center at George Mason University, Arlington, VA. Moseley, A., & Stoker, G. (2013). Nudging citizens? Prospects and pitfalls confronting a new heuristic. Resources, Conservation and Recycling, 79, 4-10. ObÉpi-Roche (2009), Enquête épidémiologique nationale sur le surpoids et l’obésité. OCDE - Organisation de coopération et de développement économiques (2009), Panorama de la santé. OECD report. Obesity update 2012 Available at: http://www.oecd.org/health/49716427.pdf

86

[Accessed: 20 Oct 2014]. Okun, M. A., Ruehlman, L., Karoly, P., Lutz, R., Fairholme, C., & Schaub, R. (2003). Social support and social norms: Do both contribute to predicting leisure-time exercise?. American Journal of Health Behavior, 27(5), 493-507. Oullier, O., Cialdini, R., Thaler, R. H., & Mullainathan, S. (2010). Improving public health prevention with a nudge. Economic Perspectives, 6(2), 117-36. Price J, Just D. Getting kids to eat their veggies. In: International Association of Agricultural Economists 27th Triennial Conference, Beijing, 2010. Putnam R. (2000) Bowling alone: the collapse and renewal of American community. New York: Simon and Schuster Robinson, E., Harris, E., Thomas, J., Aveyard, P., & Higgs, S. (2013). Reducing high calorie snack food in young adults: a role for social norms and health based messages. International Journal of Behavioral Nutrition and Physical Activity, 10(1), 73. Rolls, B. J., Roe, L. S., Halverson, K. H., & Meengs, J. S. (2007). Using a smaller plate did not reduce energy intake at meals. Appetite, 49(3), 652-660. Seymour, J. D., Lazarus Yaroch, A., Serdula, M., Blanck, H. M., & Khan, L. K. (2004). Impact of nutrition environmental interventions on point-of-purchase behavior in adults: a review. Preventive Medicine, 39, 108-136. Sherif, C. W. (1963). Social categorization as a function of latitude of acceptance and series range. The Journal of Abnormal and Social Psychology, 67(2), 148. Thaler RH, Sunstein C. Nudge: improving decisions about health, wealth, and happiness. Yale University

Press, 2008.

The Economist (2012) Behavioural economics: Herding the masses Available at: http://www.economist.com/blogs/freeexchange/2012/03/behavioural-economics

[Accessed:

20 Oct 2014]. Thunström, L., & Nordström, J. (2013). The Impact of Meal Attributes and Nudging on Healthy Meal Consumption—Evidence from a Lunch Restaurant Field Experiment. Modern Economy, 4(10a), 1-8. van Kleef, E., Otten, K., & van Trijp, H. C. (2012). Healthy snacks at the checkout counter: A lab and field study on the impact of shelf arrangement and assortment structure on consumer

87

choices. BMC public health, 12(1), 1072. Wansink, B., & Cheney, M. M. (2005). Super bowls: serving bowl size and food consumption. Jama, 293(14), 1723-1728.

88

5. Conclusions This work was based on primary data collected through a survey and two field experiments at the University of Pisa and University of Strasbourg. The three chapters of this thesis were focused on the understanding consumer behaviour in relation to energy consumption, recycling and healthy food consumption by using two different approaches (‘nudge’ as a behavioural economics approach and “Attitude Behaviour Context” (ABC) theory). How should policy makers and managers take all our findings into account? In chapter 2 our results show that more innovative means should be used to engage citizens or consumers from unsustainable practices to more environmental friendly actions. For instance, institutions and environmental associations could consider partnering with energy-saving companies to promote their innovative products on the market. Private companies could also play a pivotal role in developing the market demand for energy-saving products. This means that managers should work on building the level of trust of consumers in their communication and marketing strategies by providing credible and scientifically-based information on environmental performance. In chapter 3 our result suggests that schools, universities and companies should use awareness raising and externally imposed norms in order to nudge their students and employees to recycle more. Moreover they should use more convenient and accessible bins for recycling, making it easier to recycle. For policy makers nudge should be seen as low cost solution for that can be applied to a wide array of recycling and green behaviour issues. Our results in chapter 4 show that a nudge was not able to influence significantly consumers’ healthy food purchases. Nudges alone may not be the best solutions to encourage people to eat healthier. However nudges in combination with some other tools such as; increase of assortment, reduction of prices of healthy food, introduction of convenient lines, all together could result in winning combination and promote sales of healthy food.