UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Towards a new economic paradigm : exploring mental models and message framing effects about ecological… Tomaselli, Maria Fernanda 2017

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2017_november_tomaselli_maria.pdf [ 2.7MB ]
Metadata
JSON: 24-1.0357253.json
JSON-LD: 24-1.0357253-ld.json
RDF/XML (Pretty): 24-1.0357253-rdf.xml
RDF/JSON: 24-1.0357253-rdf.json
Turtle: 24-1.0357253-turtle.txt
N-Triples: 24-1.0357253-rdf-ntriples.txt
Original Record: 24-1.0357253-source.json
Full Text
24-1.0357253-fulltext.txt
Citation
24-1.0357253.ris

Full Text

TOWARDS A NEW ECONOMIC PARADIGM:                     EXPLORING MENTAL MODELS AND MESSAGE FRAMING EFFECTS ABOUT ECOLOGICAL ECONOMICS by  Maria Fernanda Tomaselli   MSc., The University of British Columbia, 2011  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in The Faculty of Graduate and Postdoctoral Studies (Forestry) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) October 2017  © Maria Fernanda Tomaselli, 2017     ii  Abstract The transition to a sustainable economic paradigm may be one of the most important issues of our times. This study contributes to the effective communication of ecological economics, by: 1) identifying mental models on people’s perceptions about economic growth and the environment, 2) exploring the prevalence of expansionist and ecological attitudes, and segmenting the audience based on these attitudes, and 3) exploring the effects of different messages (about the transition to economies not centered on growth) on people’s thoughts, emotions and attitudes. Sixty interviews and 1,250 online surveys were carried out in British Columbia and Canada, respectively. Data were analysed with NVivo 10, IBM SPSS Statistics 23 and Latent Gold 5.1.  Based on the interviews, five mental models were described. These sat in a spectrum of views anchored to an expansionist or to an ecological worldview. The most expansionist views (Cluster A) expressed great faith in indefinite economic growth and human ingenuity. The most ecological perspectives (Cluster E) acknowledged limits to economic growth, recognized the ecological crisis and expressed techno-skepticism. The other perspectives were in the middle of the spectrum. Based on the surveys, three audience segments were identified. Participants in Cluster 1 (41.1%) were the most optimistic towards technology and indefinite economic growth. Members of Cluster 2 (36.3%) did not express strong opinions. Participants in Cluster 3 (22.6%) acknowledged human unsustainability, expressed higher environmental concern and did not believe in indefinite growth. Sociodemographic factors (e.g. gender, political identification) correlated with the mental models and segments. Regarding the framing experiments, the messages influenced participant’s thoughts and emotions. Environmental messages invoked more references to resources and sustainability, while well-being messages generated more comments about overconsumption and happiness. Loss-framed messages caused greater negative emotions than gain-framed messages and the environmental message focused on losses generated the least hope and the greatest fear and anger among frames. There was no evidence that attitudinal responses were influenced by the frames. Most participants agreed with    iii  moving into an economic model with reduced consumption levels. This study provides data on topics that have been little explored and offers insights about the impacts of different post-growth messages.         iv  Lay Summary This study described five mental models on how BC residents perceive the relationship between economic growth, society and the environment. The mental models sat in a spectrum of views anchored to an expansionist or an ecological worldview. Participants with an expansionist worldview were optimistic towards technology, human ingenuity and indefinite economic growth, while those with an ecological worldview were concerned about the ecological crisis, acknowledged limits to growth and were skeptical about technology. The other mental models were in the middle of the spectrum. Based on an attitudinal survey, this study identified three audience segments among the Canadian public. Those in Cluster 1 (41.1%) aligned with expansionist attitudes, participants in Cluster 2 (36.3%) held no strong opinions, while those in Cluster 3 (22.6%) aligned with ecological attitudes. Finally, results show that different messages (about transitioning to economies not centered on growth) influenced people’s thoughts and emotions, but not their attitudes.        v  Preface The research in this dissertation was done under the supervision of Drs. Robert Kozak, Stephen Sheppard, Alan Jacobs and Robert Gifford. Under their guidance, I carried out the research design, data collection, data analyses and writing of this dissertation. Qualitative data analysis for Chapter 4 was aided by UBC undergraduate student, Catherine Moreau.    This research was approved by the University of British Columbia Behavioural Research Ethics Board. The certificate of approval for data collection methods carried out in Chapter 2 is UBC BREB # H14 - 00547 and for Chapter 3 and 4 is UBC BREB H15 – 01728.       vi  Table of Contents  Abstract ............................................................................................................................... ii Lay Summary ..................................................................................................................... iv Preface................................................................................................................................. v Table of Contents ............................................................................................................... vi List of Tables ...................................................................................................................... x List of Figures .................................................................................................................. xiii Acknowledgements .......................................................................................................... xvi Dedication ....................................................................................................................... xvii 1. Introduction ................................................................................................................. 1 1.1 Background .......................................................................................................... 2 1.1.1 The current economic paradigm: Growth as a natural imperative ..................... 4 1.1.2 Ecological economics: Designed for a finite planet .......................................... 6 1.1.3 Sustainable degrowth: A necessary transition for a global equitable and sustainable steady state economy.............................................................................. 17 1.2 Research Objectives ................................................................................................ 21 1.3 Theoretical Framework ........................................................................................... 22 1.3.1 Message framing .............................................................................................. 22 1.3.2 Mental models .................................................................................................. 28 1.4 General Methodology ............................................................................................. 31 1.4.1 Geographic scope ............................................................................................. 31 1.4.2 Methodological approach................................................................................. 33 1.5 Structure of the Thesis ............................................................................................ 34    vii  2. Exploring Mental Models about Economic Growth, Society and the Environment in the BC Population ............................................................................................................. 36 2.1 Introduction ............................................................................................................. 36 2.2 Methodology ........................................................................................................... 42 2.2.1 Study area and recruitment of participants ...................................................... 42 2.2.2 Data collection methods ................................................................................... 44 2.2.3 Data analysis .................................................................................................... 47 2.3 Results ..................................................................................................................... 52 2.3.1 Perceptions of economic growth ...................................................................... 56 2.3.2 Views on limits to growth ................................................................................ 58 2.3.3 Perceptions of people and society .................................................................... 60 2.3.4 Views on the importance of the economy, society and the environment ........ 62 2.3.5 Views on sustainability .................................................................................... 63 2.3.6 Reactions to ecological economics .................................................................. 66 2.3.7 Clusters and cultural cognition of risk ............................................................. 70 2.4 Discussion ............................................................................................................... 73 2.5 Conclusions ............................................................................................................. 82 3. Identifying Like-Minded Audiences for the Communication of Ecological Economics in the Canadian Population ............................................................................ 83 3.1 Introduction ............................................................................................................. 83 3.2 Methodology ........................................................................................................... 88 3.2.1 Data collection methods ................................................................................... 88 3.2.2 Data analysis .................................................................................................... 91 3.3 Results ..................................................................................................................... 95 3.3.1 Respondent segmentation results ..................................................................... 96    viii  3.3.2 Concern with environmental and economic issues ........................................ 102 3.3.3 Sociodemographic characteristics of segments ............................................. 103 3.4 Discussion ............................................................................................................. 105 3.5 Conclusions ........................................................................................................... 113 4. Framing of Ecological Economics .......................................................................... 115 4.1 Introduction ........................................................................................................... 115 4.1.1 Framing theory ............................................................................................... 116 4.2 Methodology ......................................................................................................... 121 4.2.1 Data collection methods ................................................................................. 121 4.2.2 Data analysis .................................................................................................. 127 4.3 Results ................................................................................................................... 130 4.3.1 Confounding variables ................................................................................... 130 4.3.2 Framing effects on cognitive responses ......................................................... 130 4.3.3 Framing effects on emotional reaction .......................................................... 132 4.3.4 Framing effects on attitudinal responses ........................................................ 144 4.4 Discussion ............................................................................................................. 147 4.5 Conclusions ........................................................................................................... 164 5. Conclusions ............................................................................................................. 166 5.1 Summary of Empirical Findings ........................................................................... 167 5.1.1 Objective 1: Mental models regarding the economy, society and the environment ............................................................................................................ 167 5.1.2 Objective 2: Prevalence of expansionist and ecological attitudes in the Canadian population and audience segmentation ................................................... 169 5.1.3 Objective 3: Message framing effects about transitioning to economies not centered on growth .................................................................................................. 171    ix  5.2 Synthesis of Empirical Findings, Theoretical Significance and Practical Implications................................................................................................................. 173 5.3 Strengths and Limitations of this Research .......................................................... 176 5.4 Future Research Directions ................................................................................... 177 5.5 Concluding Remarks ............................................................................................. 178 Bibliography ................................................................................................................... 179 Appendices ...................................................................................................................... 192 Appendix A: Interview protocol ................................................................................. 192 Appendix B: Visual aids used during the interviews .................................................. 194 Appendix C: Sociodemographics survey .................................................................... 196 Appendix D: Cultural cognition of risk survey........................................................... 197 Appendix E: Output examples of NVivo 10 cluster analysis ..................................... 198 Appendix F: Survey example...................................................................................... 201 Appendix G: Parameter estimates for ordinal regressions (main effects) .................. 212 Appendix H: Parameter estimates for ordinal regressions (significant interaction effects)......................................................................................................................... 215        x  List of Tables Table 1.1 Examples of some policy proposals for addressing each major goal of ecological economics. ....................................................................................................... 17 Table 2.1 Sociodemographic characteristics of the sample, classified by participants from MV and BCI. ..................................................................................................................... 45 Table 2.2 Values assigned to each scale item in the cultural cognition scale .................. 51 Table 2.3 Label given to each cluster and brief description of its intended meaning. ..... 53 Table 2.4 Summary of the most prominent perspectives and sociodemographics in each cluster. ............................................................................................................................... 54 Table 2.5 Perceived benefits and costs of economic growth across all clusters. ............. 57 Table 2.6 Participants’ views on people and society categorized by cluster. .................. 61 Table 2.7 Relative importance of the economy, society and the environment in relation to each other. ......................................................................................................................... 62 Table 2.8 Definitions of sustainability provided in each cluster. ..................................... 64 Table 2.9 Main criticisms and limitations identified by participants in relation to the concepts of ecological economics. .................................................................................... 68 Table 2.10 Broad representation of general patterns of thought in each cluster, classified by the main themes of analysis. ........................................................................................ 75 Table 3.1 Sociodemographic characteristics of survey respondents. ............................... 89 Table 3.2 Goodness-of-fit statistics for 2- to 6-cluster solutions using 12 variables (initial model). .............................................................................................................................. 92 Table 3.3 R2 values for each variable in the initial model. .............................................. 93 Table 3.4 Goodness-of-fit statistics for 2- to 6-cluster solutions using 8 variables (improved model).............................................................................................................. 93 Table 3.5 Relative frequencies, means and standard deviations for 12 items for all participants. ....................................................................................................................... 97 Table 3.6 Conditional probabilities for each cluster. ..................................................... 101 Table 3.7 Overall R2 and means for all participants, means for each cluster profile and Kruskal-Wallis results for comparisons between cluster mean ranks for each survey statement included in the model. .................................................................................... 102    xi  Table 3.8 Concern for the state of the natural environment and the Canadian economy classified by cluster. ........................................................................................................ 103 Table 3.9 Sociodemographic characteristics for each cluster in terms of gender, political affiliation, age, income and education. ........................................................................... 104 Table 3.10 Chi-square test for gender and cluster. ......................................................... 104 Table 3.11 Chi-square test for political identification and cluster. ................................ 105 Table 4.1 Sociodemographic characteristics of survey respondents. ............................. 122 Table 4.2 Description of message frames used in this study. ........................................ 124 Table 4.3 Wording for each treatment condition. .......................................................... 125 Table 4.4 Numerical coding for different sections of the survey, categorized by the type of variable. ...................................................................................................................... 129 Table 4.5 Most frequent cognitive responses that emerged in each condition. ............. 131 Table 4.6 Frequency proportions, means and standard deviations for reported feelings of hope between treatments. ................................................................................................ 133 Table 4.7 Frequency proportions, means and standard deviations for reported feelings of fear between treatments. ................................................................................................. 134 Table 4.8 Frequency proportions, means and standard deviations for reported feelings of anger between treatments. ............................................................................................... 134 Table 4.9 Test of model effects for the dependent variables hope, fear and anger. ....... 135 Table 4.10 Test of interaction effects for dependent variables hope, fear and anger ..... 137 Table 4.11 Number of responses pertaining to the content of the message and the general context. ............................................................................................................................ 139 Table 4.12 Most commonly mentioned themes that emerged for hope for each frame, in relation to the message and context. ............................................................................... 140 Table 4.13 Most commonly mentioned themes that emerged for fear for each frame, in relation to the message and context. ............................................................................... 141 Table 4.14 Most commonly mentioned themes that emerged for anger for each frame, in relation to the message and context. ............................................................................... 143 Table 4.15 Means and standard deviations for attitudinal questions categorized by five experimental conditions. ................................................................................................. 144    xii  Table 4.16 Relative frequencies, means and standard deviations for attitudinal statements for all participants. .......................................................................................................... 146      xiii  List of Figures Figure 1.1 Simplified models illustrating the fundamental visions of ecological economics (left) and neoclassical economics (right). ......................................................... 8 Figure 1.2 Research approach used for addressing each objective posed in this study. .. 34 Figure 2.1 Examples of how some participants depicted the relationship between the economy, society and the environment. ............................................................................ 46 Figure 2.2 Characterization of views about economic growth. ....................................... 56 Figure 2.3 Word clouds representing the initial associations brought forth by participants when thinking about economic growth. ............................................................................ 57 Figure 2.4 Differences in views among clusters regarding the existence of limits to growth. .............................................................................................................................. 59 Figure 2.5 Characterization of opinions about people and society. ................................. 61 Figure 2.6 Participants’ opinions regarding the current state of sustainability, classified by cluster. .......................................................................................................................... 65 Figure 2.7 Cultural cognition of risk results for all respondents ..................................... 71 Figure 2.8 Cultural cognition of risk classified by gender ............................................... 71 Figure 2.9 Cultural cognition of risk classified by place of residence (i.e. metro Vancouver or the BC Interior) .......................................................................................... 72 Figure 2.10 Cultural cognition of risk classified by clusters ........................................... 73 Figure 2.11 Clusters classified in the expansionist versus ecological spectrum (based on Rees 1995). ....................................................................................................................... 76 Figure 3.1 Plot of log-likelihood values (CAIC, BIC and AIC) for 2- to 6-cluster solutions for the improved model. .................................................................................... 94 Figure 3.2 Reported concern about the state of the Canadian economy. ......................... 95 Figure 3.3 Reported concern about the state of the natural environment. ....................... 95 Figure 3.4 Frequency with which participants think about how the economy and the environment affect each other. .......................................................................................... 96 Figure 3.5 Proportion and numbers of participants classified into each cluster. ............. 97 Figure 3.6 Profile plot of means for 3-cluster model ..................................................... 100    xiv  Figure 3.7 Kruskal-Wallis pairwise comparisons for income (left figure) and education (right figure). ................................................................................................................... 105 Figure 3.8 Some perceptions that emerged in each cluster, represented in the expansionist-ecological spectrum ................................................................................... 106 Figure 4.1 Main and interaction effects explored in this chapter................................... 121 Figure 4.2 Kruskal-Wallis pairwise comparisons for a) hope, b) fear and c) anger. ..... 133 Figure 4.3 Percentage responses for all participants to attitudinal item #1: Continued economic growth is essential for improving people’s quality of life. ............................. 147 Figure 4.4 Percentage responses for all participants to attitudinal item #2: Economic growth is the best measure of social progress. ................................................................ 148 Figure 4.5 Percentage responses for all participants to attitudinal item #3: Politicians should give less priority to economic growth as a major public policy goal. ................. 148 Figure 4.6 Percentage responses for all participants to attitudinal item #4: A 'good life' is possible without continuous economic growth. .............................................................. 149 Figure 4.7 Percentage responses for all participants to attitudinal item #5: In view of limited natural resources, people should figure out ways to increase quality of life while reducing overall material consumption. .......................................................................... 149 Figure 4.8 Percentage responses for all participants to attitudinal item #6: Economic growth will not be limited by the availability of natural resources. ............................... 150 Figure 4.9 Percentage responses for all participants to attitudinal item #7: The benefits of economic growth outweigh its negative consequences. ................................................. 150 Figure 4.10 Percentage responses for all participants to attitudinal item #8: We should continue growing our economy despite any large negative consequences. .................... 151 Figure 4.11 Percentage responses for all participants to attitudinal item #9: We should eventually transition into an economic model based on reduced levels of consumption 151 Figure 4.12 Percentage responses for all participants to attitudinal item #10: A sustainable economic model will only be possible if we stabilize the size of our population. ...................................................................................................................... 152 Figure 4.13 Percentage responses for all participants to item #11: How likely or unlikely are you to support a Canadian politician that does NOT pursue economic growth as a major policy goal?........................................................................................................... 152    xv  Figure 4.14 Percentage responses for all participants to item #12: What level of economic growth do you think the government should aim for in the next 10 years? ... 153 Figure 4.15 Point assigned to economic growth, environmental issues and social well-being in order of importance by all participants. ............................................................ 153      xvi  Acknowledgements I offer my profound gratitude to my supervisor, Dr. Robert Kozak for his continuous support and for his great insights and feedback throughout this process. I extend my sincere thanks to my PhD committee members, Drs. Stephen Sheppard, Alan Jacobs and Robert Gifford for their invaluable guidance and assistance. It was a pleasure and a privilege to work, collaborate and learn from each one of you. Additionally, I am deeply grateful to Professor William Rees for his course on Ecological Economics, which shifted my worldview about the economy and gave me hope that a feasible alternative to our unsustainable paradigm exists.  This investigation would not have been possible without the contribution of many people that were willing to participate in this study. Special thanks to Amanda Mjolsness for all her assistance during my stay in Prince George. Many thanks to Catherine Moreau for her hard work and enthusiasm while analyzing the surveys. I am very grateful to Jorma Neuvonen and Brian Bosworth who helped me in the data collection stage of this thesis by connecting me with potential research participants. I am very thankful to the University of British Columbia, and particularly the Faculty of Forestry, for granting me scholarships and awards throughout my studies. I also extend my appreciation to the Social Sciences and Humanities Research Council (SSHRC) of Canada for funding the fourth year of my doctoral studies. I am also indebted with Dr. Robert Kozak who funded the fieldwork (interviews and surveys) for this research. Without his financial and moral support, this dissertation would not have been possible.  A special thanks to all my fellow colleagues and friends who gave me their continuous support and shared their positive energy with me in the good and bad times. Special thanks to Ana Elia Ramon, Reem Hajjar, Andrea Vasquez, Rene Reyes, Jennifer Romero, Kahlil Baker, Maria Jose Ruiz, Laura Morillas, Trudie Van Der Burg, Kathy Scalzo and Colin Leech-Porter, among others. Finally, I am deeply grateful to my parents who have unconditionally supported and encouraged me along these years, and to my husband Hernan, who has been my continuous source of inspiration, creativity and strength.     xvii  Dedication  To my husband and parents   1  1. Introduction The unprecedented economic growth achieved during the current and past century has been attained through the unsustainable use of resources and ecosystem services. For decades, ecological economists and others have called for a change in paradigm; highlighting the importance of recognizing that the human economy is part of a bigger, but non-growing, system: the ecosphere. Although the ideas of ecological economics are not new, the influence of this transdiscipline is still marginal in academic, political and public circles. Research on ecological economics has been largely focused on building the theory and economic tools suitable for an ecological macroeconomy, but empirical research is still lacking, especially on areas related to public opinion and communication.  The main purpose of this investigation is to explore ways in which some principles and concepts of ecological economics – such as transitioning into non-growing steady state economies – can be better communicated to, and accepted by, the public. This research borrows from the fields of psychology, behavioural economics and political communication, and specifically uses framing theory to assess whether attitudes relevant to ecological economics are influenced by the way the information is presented. In addition, it explores people’s attitudes and mental models about the interaction between the economy, society and the environment, and segments the audience based on these attitudes. Ultimately, this study aims to determine how each audience segment moderates framing effects. Interviews and online survey experiments are the main data collection methods. This introductory chapter is organized into five main sections: 1) the background presents the problem statement and introduces ecological economics; 2) the research objectives present the broader and specific focus of this dissertation; 3) the theoretical framework summarizes concepts and definitions around framing theory and mental models, and introduces relevant research gaps; 4) the methodology describes the methodological approach, data collection, and analysis employed for tackling each objective; and 5) the structure of the thesis describes how the thesis is organized.       2  1.1 Background Humanity will face unprecedented ecological and social challenges in the 21st century (Greer, 2011; Heinberg, 2011; Steffen et al., 2011). The planet’s ability to provide goods and services is dwindling, while our human impact keeps growing (Rees, 2003a; Steffen et al., 2011). The ecological footprint analysis shows that, for more than three decades, the world has been in a state of overshoot, as we are consuming more resources than can be regenerated (Wackernagel & Rees, 1996). It has been contended that humanity has already crossed four out of nine ‘planetary boundaries’:1 climate change, biosphere integrity (e.g. loss of biodiversity), land system change and modifications of biochemical flows (e.g. nitrogen and phosphorous cycles) (Steffen et al., 2015). Other challenges are related to acidification of oceans, resource scarcity and a decline in ecosystem services, among others (Steffen et al., 2011). Humans – one species among the millions that live on the planet – have appropriated 25% of the planet’s net primary production and now influence about 75% of vegetated lands (Krausmann et al., 2013). The unprecedented exponential growth in the human enterprise since mid-20th century and its effects on natural systems has been termed the Great Acceleration, with increased consumption identified as its main driver (Steffen et al., 2011). Our impact is such that scientists have warned about the increasing risks of disrupting the Earth’s Holocene-like state (Steffen et al., 2015). The climatic stability of the Holocene has allowed human civilization to thrive and is the only geologic state that certainly can support current society (Steffen et al., 2011). Moreover, a new geological era has been proposed – the Anthropocene – which is dominated  by human influence (Crutzen, 2002) and where “humanity itself has become a global geophysical force” (Steffen et al. 2011, p. 741).  This dire ecological situation is only exacerbated by the increasing levels of wealth and income disparity evidenced across the globe (Dietz & O’Neill, 2013). Current levels of global economic inequality are such that, in 2015, the wealth of the richest 1% surpassed that of the remaining 99%. What is more, in relation to the previous five years, the wealth                                                  1 A boundary is a limit under which it is safe for humans to function; crossing one increases the risk of destabilizing the Holocene-type state of the Earth system. The Holocene is the current geological epoch that began about 11,000 years ago, after the last ice age (Steffen et al., 2015).     3  owned by the bottom 50% declined by 38%, while that of the top 1% increased by 45% (OXFAM, 2016). This situation has not only economic, but also political ramifications, as economic power is used to exert political influence in favour of the very wealthy, reinforcing inequality even more (OXFAM, 2016). This predicament force us to question and identify the underlying drivers of these alarming ecological and social problems. Some authors hypothesise that humans have an inherent tendency to be unsustainable, and, if possible, will use all available resources and occupy all accessible niches (Hardin, 1968; Rees, 2010). This inclination helped us to survive and expand in an ‘empty’ human world, but is becoming problematic and potentially devastating for our species once we approach planetary limits. “By depleting natural capital and eroding life support systems, the continued material growth undermines the future of global civilization” (Rees, 2010, p.6). In 2005, the Millennium Ecosystem Assessment Board released this warning: “Human activity is putting such strain on the natural functions of Earth that the ability of the planet’s ecosystems to sustain future generations can no longer be taken for granted” (Millenium Ecosystem Assessment, 2005, p. 5).  Many individuals and institutions have been calling for a paradigmatic change. “[…] we need to fundamentally alter our relationship with the planet we inhabit” (Steffen et al., 2011, p. 739). These authors make the plea to move from resource exploitation towards planetary stewardship, while other authors have explicitly appealed for a new economic paradigm (Boulding, 1966; Daly & Farley, 2011; Jackson, 2011; Victor, 2008). For decades, ecological economists have pointed out that the economy and, specifically, the drive for continuous material growth, is the root cause for these multiple ecological issues. The economy depends on the use and transformation of natural resources (Rees, 2003a) and the unprecedented growth seen during the last decades has been possible due to the extraction of cheap fossil fuels (Heinberg, 2011; Nikiforuk, 2012) and the unsustainable use of resources (UNEP, 2011). If societies neglect these multiple ecological warnings and continue with the current approach, they will likely face, at one    4  point or another, economic, social and/or ecological collapse (Greer, 2011; Jackson, 2011; Donella Meadows, Meadows, & Randers, 1992).  1.1.1 The current economic paradigm: Growth as a natural imperative  The notion of progress was initially a broad idea that encompassed advances in various areas, such as science, health, arts, wealth and social organization. However, it was not until the past century that this notion was conflated by many with the concept of economic growth (Victor, 2008). Economic growth is defined as the increase in the amounts of goods and services produced in an economy determined by their market value, and is measured by the increase in real Gross Domestic Product (GDP). GDP is often used as a proxy for rising living standards and, therefore, well-being (Heinberg, 2011). “A large GDP does not ensure that all of a nation’s citizens are happy, but it is surely the best recipe for happiness that macroeconomists have to offer” (Mankiw & Scarth, 2004, p. 44). The importance granted to this indicator by economists and politicians cannot be overstated; most policies are assessed according to their impact on GDP and much of the study of macroeconomics is focused on understanding and explaining the dynamics of this indicator (Dietz & O’Neill, 2013). Unless otherwise stated, the use of the term ‘economic growth’ in this dissertation refers to GDP growth. Economic growth is thought to increase employment and living standards, as more goods and services are produced and consumed in an economy (Mankiw & Scarth, 2004). Between 1910 and 2005, the global economy grew by more than 1,600%, a rate much greater than the growth in population, which was of 274% during the same time frame (Krausmann et al., 2013). Despite this extraordinary growth, factors such as technological improvements, mechanisation and efficiency gains generally result in the use of less labour for producing the same amount of goods. This means that the economy needs to keep growing in order to keep employing a similar number of people (Victor, 2008). "The shortage of employment has become more important than the shortage of products. Whereas in the past we needed to have more people at work because we needed the goods and services they produce, now we have to keep increasing production simply to    5  keep people employed" (Victor, 2008, p. 12). In this way, keeping up aggregate demand is central for the economic model. The belief and devotion to unlimited economic growth (as measured in GDP) has filtered down to most governments, institutions and economic policies around the world (Gowdy & Erickson, 2005; Ockwell, 2008). It has become a major goal for the majority of nations, thus perpetuating the underlying assumption that indefinite exponential economic growth is possible and desirable. Some critics have highlighted that the obsession with economic growth has become a dogma or doctrine that few dare to question or challenge. Nonetheless, as a result of ecological deterioration, social inequity, global economic crises and lower rates of global growth, the dominant economic paradigm is being questioned in academic, social and political circles2 (Hopkins, 2008; Jackson, 2011; New Economics Foundation, 2009; Schneider, Kallis, & Martinez-Alier, 2010; Victor, 2008). For instance, a few influential mainstream economists and heterodox media are now contesting the desirability and possibilities of continuous growth (Drews & van den Bergh, 2016; Norgard, Peet, & Ragnarsdottir, 2010). Although criticisms to the current growth paradigm still tend to be on the margins, they present a renewed opportunity to re-examine the dominant worldviews guiding society and may open new possibilities of co-existence between people and nature. This could be a prime opportunity to put forward some of the guiding principles and proposals of ecological economics, such as the need to transition to non-growing steady state economies (as described by Daly and Farley, 2011).                                                  2 This is partly revealed in the number of organizations working on promoting different economic models (e.g. New Economics Foundation, New Economy Coalition, Postgrowth Institute) and in a few media articles. For example:  Simms, A. (May 1, 2013) Endless growth will not deliver a healthy economy: The UK must seize this opportunity to create an economy that can flourish without addiction to relentless expansion, The Guardian. Retrieved on June 22, 2017 from: https://www.theguardian.com/environment/blog/2013/may/01/endless-growth-not-deliver-healthy-economy Neril, Y. (November 9, 2015) Questioning the growth mentality, The Huffington Post blog. Retrieved on June 22, 2017 from: http://www.huffingtonpost.com/yonatan-neril/the-sabbatical-year-and-t_b_8110104.html     6  1.1.2 Ecological economics: Designed for a finite planet In the past decades, the transdiscipline3 of ecological economics has confronted some important gaps in the theory and practice of mainstream economics by integrating and addressing the relationship between economic and ecological systems (Costanza, Daly, & Bartholomew, 1991). Ecological economics “[…] tries to manage the whole system and acknowledges the interconnections between humans and the rest of nature (Costanza et al., 1991, p. 6).” This differentiates it from conventional economics and ecology in that economics places emphasis on humans only, while ecology focuses mainly on non-humans. It is the only heterodox form of economics that centres on the economy as a social system and as one restricted by the biophysical environment (Gowdy & Erickson, 2005). A central tenet of ecological economics is that the human economy is embedded in nature and, by default, is an extension of natural and biological processes (Røpke, 2004). Moreover, it attempts to ground economics in physical reality and takes into account the laws of thermodynamics (Ockwell, 2008). Although ecological economics is an heterodox field with multiple perspectives (Spash, 2011), the main view of ecological economics adopted in this thesis is the one proposed by Herman Daly4 and others who have been influential in the field, such as Peter Victor, William Rees and Robert Costanza, who highlight the need to move away from economic growth as a main policy goal. Moreover, many of the ideas and proposals presented in this dissertation emerge from the seminal textbook “Ecological Economics: Principles and Applications” by Herman Daly and Joshua Farley (2011). According to Daly and Farley (2011), ecological economics espouses three main goals, in order of importance: 1) achieving a sustainable scale for the human economy; 2) attaining a fair distribution of resources; and 3) generating an efficient allocation of these scarce resources. Each of these is discussed below.                                                   3 Ecological economics is considered a transdiscipline because it tries to integrate the perspectives of multiple disciplines like economics, ecology, psychology and political science, among others (Costanza et al., 1991; Gowdy & Erickson, 2005). 4 Herman Daly has often been called the “founding father of ecological economics.”    7  1.1.2.1 Achieving an optimal sustainable scale Ecological economics views the human economy as an open system that occurs within a closed system – the ecosphere (see Figure 1.1). The economy is seen as part of nature and dependant on it for its development and growth. Biophysical constraints and the finite nature of global carrying capacity are acknowledged. Essential to this view is that there are limits to indefinite increases in material growth and energy use (Daly, 1991; Daly & Farley, 2011; Rees, 2011). “Ecological Economics emphasizes the limits to material and energy throughput and the problems then posed by the modern economic obsession with increasing consumption” (Spash, 2011, p. 359). Moreover, there is a recognition that there are limits to GDP growth because, ultimately, GDP growth reflects total resource and physical throughput (Daly, 2013). As explained by Agnolucci, Flachenecker and Söderberg (2017, p. 3): “Although the size of the impact of changes in income on the use of material depends on the material requirements of economic goods being consumed, economic activity ultimately leads to an increase in the use of materials as long as economic activity affects the consumption of economic goods.” Trade has allowed people to overcome local limitations, but has contributed to the illusion of unlimited resources (Daly & Farley, 2011). According to Daly and Farley (2011) this worldview greatly diverges from the implicit vision of neoclassical economics, illustrated in the circular flow of exchange value (see Figure 1.1), where the economy is represented as an isolated system independent from the environment. In the neoclassical view “nothing enters from the environment nor exits to it. The physical environment is completely abstracted from” (Daly, 1994, p. 22).     8                       Figure 1.1 Simplified models illustrating the fundamental visions of ecological economics (left) and neoclassical economics (right) (adapted from Daly and Farley, 2011 and Mankiw & Scarth, 2004). Rees (1995) differentiates the worldview associated with ecological economics (labeled as ‘ecological’), with the one associated with neoclassical economics (labeled as ‘expansionist’). In a nutshell, by Rees (1995) explains that expansionist worldview sees humans as masters of nature, recognizes indefinite economic expansion as natural and unrestrained from limits, believes that economic growth and free markets are paramount for social progress and ecological sustainability, and is heavily techno-optimistic. Contrasting this perspective, an ecological worldview sees the economy as fully dependent on the environment, recognizes ecological limits to economic growth and acknowledges that GDP is an inadequate measure of social well-being and ecological health, and is more skeptical about technology’s capacity to replace complex ecological systems (Rees, 1995).  Another point of divergence between conventional and ecological economics lies in the production function. In neoclassical economics, the production function often includes capital and labour (Mankiw & Scarth, 2004), while ecological economists expand it to include natural capital, resources, energy and waste (Daly & Farley, 2011). In this sense, in traditional economic accounts, the contribution of nature to the economic process is often not included (Costanza et al., 1991). Moreover, “for ecological economists, energy is a fundamental factor enabling economic production” (Ockwell, 2008, p. 4601). Ecosphere Income  Economy Ecological Economics Labor  Households Firms     Neoclassical economics Expenditure  Goods     9  Another point of contention between worldviews lies around the substitutability between capitals. On the one hand, the conventional view often espouses that natural capital can be substituted with human-made capital (i.e. weak sustainability) (Costanza et al., 1991; Pezzey & Toman, 2002). With a weak sustainability rule, an economy is deemed sustainable if total savings are greater than the combined depreciation of natural and manufactured capital (Pearce & Atkinson, 1993). On the other hand, some ecological economists are more skeptical of the capacity of human-made capital to substitute for nature (i.e. strong sustainability) (Gowdy, 2000; Ockwell, 2008). With a strong sustainability basis, an economy is deemed sustainable when the stock of natural capital remains constant (Ott, Muraca, & Baatz, 2011). In other words, to be sustainable, humans should live off of the natural income (i.e. flow of services and resources) generated by natural capital. Relatedly, there has been a call to measure wealth in biophysical terms (e.g. physical stock of resources), rather than solely in monetary terms (Daly & Farley, 2011). Daly and Farley (2011) contend that there is an optimal scale for the macroeconomy; a point where marginal costs equal marginal benefits, after which further economic expansion generates more losses than advantages. Fundamental to this view is the idea that there are opportunity costs to economic growth. In fact, globally, we may well be in a period of uneconomic growth, where, due to ecological degradation and other costs, further economic growth is actually making us poorer rather than richer (Daly, 2013). The pattern of scarcity has changed, as natural capital is becoming increasingly scarce, whereas human-made capital is becoming more abundant (Costanza et al., 1991). This is supported by research that shows that, when ecological degradation and other costs are taken into account, economic welfare has actually decreased since the late 1970s, while GDP has continued rising exponentially (Kubiszewski et al., 2013).  The idea of an optimal scale for the macroeconomy underscores diametrically opposed views between conventional and some ecological economists (Daly & Farley, 2011). “Where conventional economics espouses growth forever, ecological economics envisions a steady state economy at optimal scale” (Daly & Farley, 2011, p. 23). A steady    10  state economy is an optimally sized economy with relatively constant throughput of materials and energy, and stable fluctuations of people (Daly & Farley, 2011; Rees, 2011). In order to be sustainable, renewable resources in a steady state economy should not be used faster than their regeneration rates, waste should not be emitted more rapidly than the ecosystem’s absorption capacity, and non-renewables should not be depleted quicker than the rate at which renewable alternatives can be developed (Daly & Farley, 2011). A steady state economy need not be stagnant and can, in fact, be dynamic. Many aspects can continually change, evolve and improve, such as technologies, mixes of products and businesses, information and institutions, among others (Dietz & O’Neill, 2013). “After all, John Stuart Mill, one of the founding fathers of economics, recognized both the necessity and the desirability of moving eventually towards a ‘stationary state of capital and wealth’, suggesting that it ‘implies no stationary state of human improvement’” (Jackson, 2011, p. 122). 1.1.2.1.1 Criticisms, counterviews and author positionality   “The decades old question ‘Is economic growth environmentally sustainable?’ remains contested despite its apparent simplicity” (Ward et al., 2016, p. 1). Some of the arguments used to dispute the validity of the ideas presented above and dismiss the idea of limits to material and GDP growth are mostly related to technological progress, efficiency gains, material and economic decoupling, substitution between natural and human-made capital, and monetary signs of scarcity (Simon, 1981). In relation to resource scarcity, decreasing prices of some natural resource commodities has been used as evidence to suggest that these are not becoming scarcer (Krautkraemer, 2005) and moreover, that they may be becoming more abundant (Simon, 1981). The main argument is that once prices start to rise, this will motivate people to find substitutes and develop alternatives that will overcome any limitations, indefinitely (Simon, 1981).5 Nonetheless, Krautkraemer (2005) warns that even if natural resource commodities are not becoming more scarce (using prices as the main indicator), the provision and potential scarcity of environmental                                                  5 The subsection on ‘Efficient allocation’ states some of the limitations of using prices as an indicator of scarcity from an ecological economics point of view.     11  services may be threatened, as these are not included in market accounts and are not subject to price signals. In addition, technological optimists have faith that technology and innovation will overcome resource scarcity and the depletion of nature.  Another argument used to dismiss limits to GDP growth relates to efficiency gains achieved in past decades. For instance, the global energy intensity per unit of GDP decreased by 26% between 1990 and 2005 (IEA, 2008). Similarly, today we use less materials and resources per unit of GDP (i.e. material intensity) than what we did at the beginning of the past century (Krausmann et al., 2009). This trend has also been evidenced in some countries. For example, the Chinese economy (as measured by GDP) grew by a factor of 20 (between 1990 and 2012), while material use increased only by a factor of five and energy use by a factor of four during the same time. In Germany, while GDP has continued growing, energy use has dropped by 10% and material use by 40% (Ward et al., 2016). These and related data are often used to endorse the idea of economic decoupling. Decoupling occurs when economic growth (as measured in GDP) is disassociated from environmental impacts (Ward et al., 2016). Many cases of relative decoupling exist (as illustrated above). However, there is little evidence that the economy will de-materialize in absolute terms (Jackson, 2011; Ward et al., 2016). “The decoupling debate itself is polarized with a preponderance of neo-classical economists on one side (decoupling is viable) and ecological economists on the other (decoupling is not viable)” (Ward et al., 2016, p. 2). Efficiency gains are often subject to the Jevons Paradox, which occurs when efficiency improvements tend to increase overall consumption (Daly, 2013). For example, despite the remarkable gains in efficiency and reductions in material intensity made since the 1900s, global aggregate material and energy use have increased significantly, to a point where overall material use per capita doubled between 1900 and 2005 (Krausmann et al., 2009). In this sense, an ecological economics worldview is more skeptical of the possibilities of decoupling GDP growth from energy use (Ockwell, 2008). Using another example, Jackson (2011) points out that, if the CO2 target of 450 ppm is to be reached and global incomes rise at 2% per year, carbon intensity would need to be reduced to 14gCO2 per dollar by 2050 from 768    12  gCO2 per dollar in 2007. If the same income trends continue, by 2100 “nothing less than a complete decarbonization of every single dollar will do to achieve carbon targets” (Jackson, 2011, p. 81). Along similar lines, Rees (1995) pointed out two decades ago that, if the economy kept growing at 3% per year (which is often considered a modest rate of growth), the environmental footprint per unit of consumption would need to be reduced by 90% within the next 30 years to meet demands within ecological limits; beyond that, a ‘complete dematerialization’ would be needed. Many put forward the idea that a service-based economy can be decoupled from environmental impact, but, as Ockwell (2008, p. 4602) points out, “[…] this notion ignores the large amounts of energy involved in producing services.” Moreover, many countries have been able to transition to service-based economies by partly shifting the manufacturing of goods to other countries (Ockwell, 2008). In summary, there is little evidence that any country has achieved absolute decoupling in the past decades, nor environmental impact has been reduced globally (Ward et al., 2016). Based on historical data and modelled projections, Ward et al. (2016, p.10) conclude: “Our model demonstrates that growth in GDP ultimately cannot plausibly be decoupled from growth in material and energy use, demonstrating categorically that GDP growth cannot be sustained indefinitely. It is therefore misleading to develop growth-oriented policy around the expectation that decoupling is possible.” This thesis takes the position (alongside Daly, 2013 and other ecological economists) that there are limits to material growth, as well as GDP expansion. This position is taken due to two main reasons: 1) as illustrated above, there is not a strong body of evidence supporting absolute decoupling; that is, we do not know whether globally, GDP can continue growing indefinitely while reducing material use and environmental impacts in absolute terms; and 2) using a precautionary principle approach and following Rees’s (1995) logic, and given the high stakes at hand, it is safer and more prudent to assume that indefinite exponential GDP growth is unsustainable. This position could reduce the risk of ecological and social collapse and may allow societies to plan for a smoother    13  transition to economic systems not centered on GDP growth. In other words, it may be much riskier for global society to assume that GDP growth can be sustainable, because if this assumption is proven wrong, the costs can be catastrophic and possibly irreversible (e.g. runaway climate change, continued loss of biodiversity). On a related point, this thesis also aligns with the position that further GDP growth in already large economies does not generate significant gains in well-being (Jackson, 2011; Victor, 2017). Thus, GDP growth is an inadequate indictor of well-being and progress (Van den Bergh, 2017). This last point will be discussed more thoroughly in the next subsection.  1.1.2.2 Equitable distribution of resources Economic growth and the trickle-down effects are often used as justification for not dealing with the equitable distribution of resources (Dietz & O’Neill, 2013). However, in a non-growing steady state economy, a just distribution is central to achieving sustainability (Daly & Farley, 2011). As overall income and wealth remain relatively stable, society should set up institutions and mechanisms to enable an equitable distribution of resources. This does not mean that every person will have the same resources as every other, but that the differences between the poor and the rich should be much reduced (Dietz & O’Neill, 2013). This is important considering that well-being is generally assessed in relative terms, meaning that, after a certain level of income, how we compare ourselves with others matters more than what we have in absolute terms (Ariely, 2009; Jackson, 2011). Equitable distribution is also important, because inequality can have pervasive effects on societies. “Life expectancy, child well-being, literacy, social mobility and trust are all better in more equal societies. Infant mortality, obesity, teenage pregnancy, homicide rates and incidence of mental illness are all worse in less equal ones” (Jackson, 2011, p. 154). Ecological economists care about intra-generational equity as much as intergenerational equity (Daly & Farley, 2011).    In this sense, there is a need to use more comprehensive measures of progress that move beyond consumption-based and market exchange indicators. Indicators influence the behaviour of systems (Donella Meadows, 2008) in that they “[...] arise from values (we measure what we care about), and they create values (we care about what we measure).    14  When indicators are poorly chosen, they can cause serious malfunctions in systems (Donella Meadows, 1998, p. viii).” Thus, the focus should be placed on improving overall quality of life and achieving sustainable well-being, rather than on increasing GDP (Costanza et al., 2014; Daly & Farley, 2011; Dietz & O’Neill, 2013). “[…] it is becoming clear to ecological economists that increasing GDP should be seen for what it is: a measure of means, not ends. These ends include the promotion of well-being, employment, social justice, environmental quality, and biodiversity […] (Victor, 2017, p. 19). Furthermore, GDP has a number of limitations as an indicator of well-being. For instance, benefits and costs are conflated, such that harmful events like pollution and crime boost GDP (Costanza et al., 1991). Also, critical aspects that matter for our well-being are not measured, such as natural capital, volunteer work, social connections and housework, as these have no assigned monetary value (De Graaf & Batker, 2011). In addition, GDP does not expose how wealth is distributed within a society. “As a society, we need to realize that GDP growth (growth fetishism) is a constraint in our search for human progress, and without it we will arrive at better welfare outcomes” (Van den Bergh, 2017, p. 23). To address these issues, different proposals have been put forward. For instance, new indicators have been proposed and developed. For instance, the Genuine Progress Indicator (GPI) (previously the Index for Sustainable Economic Welfare (ISEW)) uses similar information as GDP, but adjusts it to account for ecological degradation and environmental costs, income distribution, negative activities (like crime) and other aspects not valued in the market (Kubiszewski et al., 2013). In the United States, while GDP has grown exponentially, GPI peaked in the 1980s (Dietz & O’Neill, 2013) and similar trends have been evidenced in many countries globally (Kubiszewski et al., 2013). Another indicator, the Happy Planet Index (HPI) also aims to include ecological degradation into national accounts. HPI comprises life expectancy and life satisfaction divided by a nation’s ecological footprint. Costa Rica, together with several other Latin American nations, are leading the way with this ranking (Jeffrey, Wheatley, & Abdallah,    15  2016). Furthermore, it has been proposed that, due to the significant shortcomings of GDP as an indicator of progress, we should ignore GDP per capita altogether (i.e. be indifferent about GDP growth) and rather focus on direct indicators of well-being, employment, environmental quality, etc. This position has been labeled “agrowth” (van den Bergh, 2011; Van den Bergh, 2017).  1.1.2.3 Efficient allocation The efficient allocation of resources via the market mechanism is often debated and discussed by ecological economists (Gowdy & Erickson, 2005). Markets do not always reflect accurate scarcity of resources through pricing (e.g. collapse of the North Atlantic Cod fisheries); but even when they do, a high price can be the best incentive to deplete a species (e.g. one blue fin Tuna was sold at the record price of US$ 1.8 million in Japan in early 2013)6 (Daly & Farley, 2011). In addition, life-essential ecosystem functions do not even have markets. Even if markets would be developed for these vital ecosystem services, internalizing prices is a highly complicated task, as there is incomplete information on ecosystem functions and we have little experience valuing non-market goods. For instance, many of the costs generated by economic activity will only appear in the future and are highly uncertain in the present. As essential resources become scarcer (e.g. clean water, the atmosphere’s absorptive capacity), their marginal value will increase rapidly, likely muddling the calculation of an adequate price. Overall, the market mechanism has been effective for increasing the availability of various market goods, while decreasing the existence of non-market goods and services (Daly & Farley, 2011). Despite the limitations mentioned, it has been argued that valuing nature and ecosystem services can bring attention to policymakers and society in general, about the almost ‘infinite’ value of ecosystems to human societies and our economies (Costanza et al., 1997).                                                  6 Narula, S.K. (January 5, 2014) Sushinomics: How Bluefin Tuna became a million-dollar fish, The Atlantic. Retrieved on June 22, 2017 from: https://www.theatlantic.com/international/archive/2014/01/sushinomics-how-bluefin-tuna-became-a-million-dollar-fish/282826/     16  1.1.2.4 Policies proposed for advancing to a steady state economy Daly and Farley (2011, p. 414-417) propose six main principles that should be considered when designing policies aimed at achieving a steady state economy. Some of these include: policies should strive to attain the necessary degree of macro-control with the minimum sacrifice of micro-level freedom and variability; they should leave a margin of error when dealing with the biophysical environment (i.e. precautionary principle); they must be able to adapt to changed conditions (i.e. adaptive management); and they must recognize that we always start from historically given initial conditions. With these principles in mind, various policies have been proposed to address each major objective of ecological economics. Some of these are listed in Table 1.1. Some of these proposals are already occurring in various places, with varying degrees of success. As one example of a bottom-up approach, the transition movement, which embraces many of the principles of ecological sustainability and social equity, has expanded in the past decade from a few cases in the UK to more than 1,000 transition groups in about 50 countries. Some of the activities embraced by transition groups include: community currencies, local food, community energy and local businesses, among others. The transition movement is just one example of solutions that are emerging from various community groups, activists, businesses and individuals, interested in transitioning to a different paradigm.7 Nonetheless, there is still a long way to go before these policies are implemented and embraced at national scales.                                                    7 More information about the transition movement can be found at: https://transitionnetwork.org/ (Retrieved on April 30, 2017).    17  Table 1.1 Examples of some policy proposals for addressing each major goal of ecological economics.  Major goal Policy examples Sustainable scale - Caps on resource use and waste generation (i.e. maximum quotas). Quotas could be traded in the marketplace.  - Ecological taxes and/or subsidies (e.g. carbon taxes).  - Stabilization of population and labour force (via education and family planning).  Just distribution - Maximum wage differentials (i.e. smaller differences between high and low wage earners). - Caps on income and wealth.  - Establish a citizen’s or liveable income. - Progressive taxation and more investment in social programs.  - Promote more employee-owned business structures. Efficient allocation - Price market goods closer to their true costs. - Address information asymmetry between market and non-market goods, possibly by imposing more regulations on the advertising industry.  - Recognize the value of non-market goods by offering subsidies for their provision.  Other policies - Promote more local trade of goods and services.  - Reform monetary and financial systems by eliminating debt-based money.  - Promote the development of local currencies. - Encourage investment and business models that generate social and environmental returns (not only financial). - Ensure employment opportunities for most by work-time reductions and/or guaranteed jobs. - Set new indicators for measuring progress. Note: These proposals were obtained mainly from: Daly and Farley (2011) and Dietz and O’Neill (2013), and to a lesser extent from: Jackson (2011), Kallis (2011), van den Bergh (2011), and Victor (2008). 1.1.3 Sustainable degrowth: A necessary transition for a global equitable and sustainable steady state economy? If a global equitable and sustainable steady state economy is to be achieved, many industrialized nations will need to reduce the size of their economies in terms of throughput (and possibly GDP) (Assadourian, 2012; Trainer, 2010), thus allowing less industrialized nations to use the remaining global ecological capacity (Kerschner, 2010; Martínez-Alier, Pascual, Vivien, & Zaccai, 2010; Schneider et al., 2010). Proponents of ‘sustainable degrowth’ argue that a voluntarily decrease in the material throughput of an economy – a reduction in the production and consumption of resources – is possible and    18  that it can be done in a socially equitable, democratic and sustainable manner (Martínez-Alier et al., 2010; Schneider et al., 2010). Although sustainable degrowth is focused on reducing material consumption, this will likely entail reductions in GDP, as it is assumed that significant declines in throughput cannot be achieved with a growing GDP (Kallis, 2011). However, sustainable degrowth is different from an economic recession in that a recession is an involuntary and unplanned reduction of GDP in a growth-oriented economy that generally has harmful social consequences such as unemployment, insecurity and poverty. On the other hand, degrowth represents a deliberate and intentional transition to live more simply, while still improving people’s quality of life (Kallis, 2011). It is not an end in and of itself, but the means to an optimal sized or steady state economy (Kerschner, 2010). Many authors argue that, due to internal conditions of the economic model and biophysical constraints such as resource scarcity and environmental services decline, degrowth is an inevitable future for economies (Greer, 2011; Heinberg, 2011; Kallis, 2011; Klitgaard & Krall, 2012). The question, then, is not how to avoid it, but how to make “[...] a prosperous and stable, rather than a catastrophic, descent” (Kallis, 2011, p. 873). Like a steady state economy, sustainable degrowth implies a radical change in the way society is structured; transformations in market exchange, banking, investment and employment are required, and certainly, so too is a fundamental transformation of societal values and ideals (Kallis, 2011; Klitgaard & Krall, 2012; Latouche, 2007; Trainer, 2010). Selective sectors of the economy will likely be reduced, especially those that are detrimental to the ecosphere (Kallis, 2011). “Without growth many industries would be seriously depleted or eliminated, because producing more is what they do” (Trainer, 2010, p. 2). Nonetheless, other activities would certainly be promoted within this model, such as those related to increasing organic agriculture, recovering the top soil and investing in natural capital, among others. Trainer (2010) also reflects on the impacts of degrowth for financial institutions as money creation and interest rates will likely be non-existent. Difficult questions are raised about job creation, ownership of assets, the role of markets and the political structures that will be required for deliberately degrowing the economy.    19  Another difficult area of discussion, for sustainable degrowth as well as for a steady state economy, revolves around the real possibilities of making a democratic transition. This is especially complicated considering that most politicians live with the promise of continuous economic growth as a means of solving our complex problems, raising, at the same time (together with the media and advertisement), people’s material aspirations (Matthey, 2010). In addition, consumerism is advertised as the mechanism of finding personal identity and happiness (Hamilton, 2010). Undoubtedly, in this context, obtaining popular support for sustainable degrowth, the steady state economy and even ‘agrowth’, will be very challenging. Van den Bergh (2017, p. 24) explains that discussions around GDP and growth are often “dogmatic in nature” and that “many economists agree that GDP per capita is not a good measure of social welfare, but are then still unwilling to set it aside.” Nonetheless, there is increasing recognition among politicians, economists and others about the shortcomings of GDP growth (Van den Bergh, 2017). An additional topic for consideration revolves around the possible connotations that the public may confer to different post-growth expressions (e.g. sustainable degrowth, steady state economy). “Finding a new and captivating name for the steady state economy could help attract a critical mass of people committed to taking the concept forward. What the name should be, though, remains an open question” (Dietz & O’Neill, 2013, p. 171). Words and language are important. “They can subtly influence the way we think about social issues” (Perloff, 2010, p. 216). For example, some authors argue that ‘sustainable degrowth’ could be a strong motto and a powerful label to counter the prevailing growth-based model (Martínez-Alier et al., 2010); it is a ‘missile’ word and is one of the few terms that indicates some sort of limits (Demaria, Schneider, Sekulova, & Martinez-Alier, 2013). Kallis (2011, p. 873) describes it as a “rallying slogan for a social coalition built around the aspiration to construct a society that lives better with less.” Moreover, Demaria et al. (2013) present degrowth as a common representative frame and point of convergence for various social and environmental movements. Nonetheless, other researchers claim that the term could be construed as abstract and negative (Dale, Herbert, Newell, & Foon, 2012) and that it may not generate the public support required (Drews & Antal, 2016; van den Bergh, 2011). In a similar manner, ‘steady state    20  economy’ may not be an adequate label either, as it may bring forth negative associations and could be coupled with stagnation (Dietz & O’Neill, 2013). The arguments presented here, of whether these terms are appropriate or not, have not been tested experimentally.  In addition to the ‘name’ given to the post-growth economy, attention has also been granted to the way in which the ‘story’ of this new proposition is told. Dietz and O’Neill (2013) highlight the importance of using a compelling narrative that appeals to emotions, although they also warn against ‘protective cognition’ as this new information may clash with pre-existing beliefs. “The key to bypassing protective cognition is to frame information about economic degrowth in a way that prevents people from feeling threatened. One possibility is to focus the conversation on the needs that all people share [...] and how the economy can help meet these needs without growth” (Dietz and O’Neill, 2013, p. 172). In a similar way, other individuals and institutions promoting similar goals have also pointed to the value of narratives and frames for successfully transmitting the ideals and goals of the new economy.8 However, which frames are successful and under what conditions they work is yet to be known, as there has been little empirical testing about these issues. The conceptual framework for a new ecological macroeconomy (whether steady state or degrowth) is a work in progress and there are still many related ambiguities and open questions (Trainer, 2010; van den Bergh, 2011). Scholars have primarily focused on discussing and designing the theory and concepts that would support the macroeconomics, institutions and politics for this new model. Empirical studies have received less attention (Berg & Hukkinen, 2011; Weiss & Cattaneo, 2017), especially studies on public opinion (Drews & van den Bergh, 2016), communication and behavioural dimensions. Thus, the level of public support for such an ambitious endeavor is still very unclear. Furthermore, moving into a new economy will involve shifting the fundamental vision upheld by the current economic paradigm (Daly & Farley, 2011;                                                  8 See: Korten, D. (April 4, 2013) Opening keynote at New Economy Summit, University of British Columbia. Retrieved on August 22, 2013 from: http://neweconomyatubc.ca/ New Economy Working Group. (n.d.) The framing meta-story: Equitable living economies. Retrieved on August 22, 2013 from: http://neweconomy-wg.thenextsystem.org/visions/new-economy-story     21  Dietz & O’Neill, 2013) and challenging the status quo (of endless economic growth) requires research to identify effective ways in which these concepts can be framed, communicated and accepted by the public.  1.2 Research Objectives The general aim of this doctoral dissertation is to explore and uncover more effective ways in which some of the tenets and propositions of ecological economics (as conceptualized by Daly and Farley, 2011), such as the need to transition into steady state economies, can be better communicated to the public. Specifically, this study focuses on mental models and message framing effects. This investigation represents an initial attempt to present concepts of ecological economics to the broader public. It does not aim to explore how to change behaviours; rather, it is directed at earlier stages of behaviour change such as seeing, hearing and knowing (Sheppard, 2012).  In this context, the specific objectives of this dissertation are to: 1. Identify, describe and bring to light existing mental models on how people perceive the relationship between economic growth, society and the environment. A secondary objective is to explore reactions to the alternative paradigm of ecological economics and investigate other variables (e.g. cultural cognition of risk, sociodemographics) that could relate to the identified mental models. 2. Explore the prevalence of expansionist and ecological attitudes among the general population in Canada and segment the audience based on these attitudes. A secondary objective is to determine how sociodemographic factors and other variables correlate with each identified segment.  3. Determine the effects of different message frames (related to the transition to economies not centered on growth) on people’s cognitive responses, emotional reactions and attitudes. A secondary objective is to explore how mental models and other variables (e.g. audience segments, sociodemographics) moderate framing effects.     22  These three objectives are highly inter-related. Data obtained for Objective 1 will inform data collection methods for Objective 2, while results for Objective 2 will allow determining if the mental models identified for Objective 1 replicate in a larger sample. Also, the audience segments identified in Objective 2 will be employed as a moderator variable for addressing Objective 3. 1.3 Theoretical Framework The questions posed by this research, require the integration of multiple disciplinary fields. Thus, this investigation is informed by the disciplines of behavioural economics, psychology9 and political communication. This section first describes message framing effects, synthesizes the relevant literature and points to some important research gaps. Then, it describes the pertinent literature regarding mental models, their application to natural resource and environmental topics, and indicates relevant research gaps.   1.3.1 Message framing Framing has been conceptualized and operationalized in multiple ways (Chong & Druckman, 2007; Weaver, 2007). This investigation uses the concept of frames in communication, which are embedded in media and political communication (Chong & Druckman, 2007; Scheufele, 1999). Framing effects occur when changes in the presentation of an issue or problem influence people’s responses, attitudes, evaluations or choices (Chong & Druckman, 2007; Nabi, 2003). Frames in communication should not be confused with frames in thought, which are meant to represent internal mental structures or schemas (Chong & Druckman, 2007; Scheufele, 1999).10 Research on frames in communication has been categorized into equivalence or issue framing (Chong & Druckman, 2007). In equivalence framing, logically equivalent options are framed differently. For example, a surgery’s survival rate could be framed as                                                  9 This study is mainly informed by research on environmental psychology, although it also uses theories from social and cognitive psychology, such as the Elaboration Likelihood Model (ELM) and schemata theory. 10 Frames in communication have also been called ‘media frames’, while frames in thought have been called ‘individual frames’ (Scheufele, 1999).    23  a 90% survival rate or as a 10% mortality rate (note that both options are equivalent). Experiments show that preference for the surgery drops from 80% to 50% when framed in terms of its mortality probability (Levin, Schneider, & Gaeth, 1998; Tversky & Kahneman, 1981). Similarly, the percentage of fat in food could be framed in a positive (e.g. 75% lean) or a negative way (e.g. 25% fat). Research shows that a positive frame influences people to make more optimistic assessments of the specific food or item (Levin et al., 1998). On the other hand, in issue framing, qualitatively different aspects or dimensions of the same issue are highlighted and made more salient (Chong & Druckman, 2007; Entman, 1993; Weaver, 2007). By emphasizing certain aspects in a communication while omitting others, people’s considerations and thought patterns are influenced (Chong & Druckman, 2007; Nisbet, 2009). For example, consenting to a public rally could be framed as an issue of defending free speech or as one threatening the public order (Nelson, Clawson, & Oxley, 1997). Likewise, a resource extraction issue could be framed in relation to its economic benefits or in terms of its environmental impacts (Shen 2004). In issue framing, the different frames employed are not logically equivalent (Druckman, 2004).  Much of the framing research has been operationalized as some form of equivalence framing. However, there is still little consensus on which type of message frames are the most effective (Levin et al., 1998). Part of the inconsistencies could be due to differences in the operationalization of research and the use of alternative terminology. For example, some terms may not be perceptually equal for respondents and even their tone can be inherently positive or negative (e.g. life, debt) (Levin et al., 1998; Maheswaran & Meyers-Levy, 1990). In their typology of framing effects, Levin et al. (1998) classified equivalence framing as: 1) attribute framing; 2) goal framing; and 3) risky-choice framing. Attribute framing focuses on different characteristics of an item or issue, such as success versus failure rates, or goals achieved versus goals failed. Research shows that positive messages are consistently more effective than negative ones. These findings are primarily explained by the associative model, as positive phrases tend to evoke more optimistic thoughts and positive associations than negative ones (Cheng, Woon, & Lynes, 2011; Maheswaran & Meyers-Levy, 1990; Spence & Pidgeon, 2010).     24  Goal framing focuses on the positive effects of conducting a behaviour or in the negative outcomes of not. Goal framing has been traditionally operationalized as gain and loss framing (Levin et al., 1998). Negative messages tend to have stronger effects than positive ones, although evidence is mixed. Some authors have explained these findings using prospect theory (Cesario, Corker, & Jelinek, 2013), as loss frames are expected to be more effective for behaviours considered risky, while gain frames are expected to be more effective in low risk situations (Rothman & Salovey, 1997).11 Loss aversion theory can also be used to explain these results, as negative messages could be more effective due to our negativity bias and tendency to avoid losses (as losses are perceived more strongly than gains) (Levin et al., 1998).  Lastly, in risky-choice framing, consequences of varying riskiness are presented in alternative ways (i.e. positive or negative). The riskiness of an option is related to the probability of its occurrence and not to the characteristic of the issue itself. For example, a 70% probability that 150 people will be saved would be considered a risky option because it involves some probability of occurrence, whereas the riskless option is presented with certainty (e.g. 100 people will be saved with certainty) (Levin et al. 1998). In the positive frame, outcomes are expressed as gains or as positive (e.g. lives saved) while outcomes are expressed in negative ways or as losses (e.g. lives lost) in the negative frame. Experimental results with risky-choice framing are heterogeneous, but, in general, prospect theory is useful for explaining these results (Levin et al., 1998) as negatively framed messages are more effective for outcomes that involve some level of risk, whereas positive frames are more effective for outcomes that will happen for certain. In risky choice framing, positive frames tend to enhance risk-averse responses in relation to negative frames. Framing research has been implemented in the fields of marketing, health, political communications, general decision-making, environmental issues and, most recently, climate change (Cheng et al., 2011; Davis, 1995; Loroz, 2007; Spence & Pidgeon, 2010).                                                  11 Levin et al. (1998) discuss the high degree of subjectivity involved in categorizing and perceiving risky behaviour.    25  Equivalence framing research often uses similar frames across topics. This is in contrast with issue framing studies, where different frames are often used and developed according to the topic of interest, thus making it difficult to draw comparisons across studies. For example, research on climate change has compared the effectiveness of frames that focus on local impacts versus the global or more distant impacts (Spence & Pidgeon, 2010; Wiest, Raymond, & Clawson, 2012). Research on arctic drilling has focused on the environmental costs versus the economic benefits of developing such projects (Shen, 2004). Several other studies use message frames that are often employed in the media, newspapers and political communication (Durfee, 2006; Nelson, Clawson, et al., 1997).  Issue framing effects have also been explained theoretically. The availability heuristic theory argues that framing works by making some issues more mentally accessible or available than others. “Much of the work on framing seems to regard it as an extension of the priming literature, with accessibility as the main theoretical explanation for framing effects” (Gross & D’Ambrosio, 2004, p. 3). Nelson, Oxley, and Clawson (1997) argue that framing not only makes any specific issue more accessible, but that it also affects the weight that is given to some dimension of our attitudes. For example, an economic frame may influence the weight that some people give to economic considerations in contrast to environmental ones. Another explanation for framing effects is that people often do not hold consistent or stable attitudes; this is why attitudes can be easily changed depending on how questions are framed (Chong & Druckman, 2007; Shen, 2004).  1.3.1.1 Moderating variables of framing effects12 Studies show that framing effects are moderated by a multitude of other variables. The level of involvement with a specific issue (i.e. issue involvement) could be an important moderating variable, as it influences how people process messages. According to the Elaboration Likelihood Model (ELM), message processing happens by two alternative routes: the central processing route, which is more thoughtful, cognitive, slow and                                                  12 Moderating variables are those variables that influence the strength of the relationship between the independent variable and the dependent variable.    26  systematic; and the peripheral processing route, which is simpler, quicker and involves less elaboration. The processing route taken varies according to the individual’s motivation (e.g. involvement in the topic) and ability (e.g. time, knowledge) (Perloff, 2010). In conditions of high involvement, the central processing route is often used, whereas in conditions of low involvement, the peripheral route is often employed. Some evidence suggests that loss-framed messages are more effective under high involvement (i.e. central processing), and gain-framed messages are more persuasive under low involvement (i.e. heuristic processing) (Maheswaran & Meyers-Levy, 1990; Meyers-Levy & Maheswaran, 2004). This is explained by our negativity bias, as negative information is outweighed when making judgements under high involvement and by the associative model, as we use heuristic cues under low involvement (such as the positive feelings associated with a message) (Maheswaran & Meyers-Levy, 1990). Prior knowledge and level of sophistication also seem to interact with framing effects (Price, Lilach, & Cappella, 2005). Framing effects seem to be stronger when individuals already hold previous information about an issue. This may occur because framing makes some considerations more available, so knowledgeable individuals are likely to have those considerations in their mind (regardless of whether they agree with them or not) (Druckman & Nelson, 2003; Nelson, Oxley, et al., 1997). In this sense, message effectiveness could be improved by targeting messages according to the stage of awareness of the audience (Cheng et al., 2011; Pelletier & Sharp, 2008). Nonetheless, framing effects can be weakened when people hold strong attitudes about an issue (Levin et al., 1998; Rothman & Salovey, 1997) or when the frame is inconsistent with previous beliefs (Lakoff, 2004; Myers, Nisbet, Maibach, & Leiserowitz, 2012; Shen, 2004). That being the case, determining the mental models of the audience should be typically the first step in a framing exercise (Cheng et al., 2011; McDonald, 2009; Shome & Marx, 2009) and various target messages should be developed to reach diverse audiences (Entman, 1993; Gifford & Comeau, 2011; Myers et al., 2012).  Framing effects also vary in group settings. Experiments show that effects are consistent in groups where all individuals receive the same frame, but, these effects are minimal (or    27  even disappear) in groups where individuals receive different frames (Druckman, 2004; Druckman & Nelson, 2003). In addition, socio-demographic characteristics, such as gender, age and education have influenced results in some studies (Gifford & Comeau, 2011; Lockwood, 2011; Van de Velde, Verbeke, Popp, & Van Huylenbroeck, 2010), as well as political affiliation (Gross & D’Ambrosio, 2004; Hardisty, Johnson, & Weber, 2010; Iyengar, 1991; Lockwood, 2011; Price et al., 2005; Shen, 2004). Finally, individual differences may also interact (Cesario et al., 2013; Myers et al., 2012; Tversky & Kahneman, 1981). For example, gain-framed messages may work better with promotion-oriented people, whereas loss-framed messages may be more effective with prevention-oriented individuals (Cesario et al., 2013). 1.3.1.2 Research gaps regarding framing effects Although framing studies have been applied extensively, notable research gaps persist. Much research has been focused on university students, so framing studies should be continually extended to other target populations (Davis, 1995; Loroz, 2007). In addition, new frames should be studied, developed and tested for their effectiveness according to specific characteristics of the audience (Cesario et al., 2013;13 Gifford & Corneau, 2011; Myers et al., 2012), especially in the topics of sustainability and climate change (Gifford & Comeau, 2011; Nisbet, 2009). Few studies have looked into the moderating relationship between framing effects and different audience segments (Hine et al., 2016; Shen, 2004). Furthermore, emotional reactions to framing have been largely overlooked and there is little empirical research in this regard (Gross, 2008; Gross & D’Ambrosio, 2004; Lecheler, Schuck, & de Vreese, 2013; Myers et al., 2012). Finally, little studies that have researched message framing effects on topics related to the communication of ecological economics have been uncovered.                                                  13 For example, Cesario et al. (2013) propose that regulatory focus theory is adequate for predicting frame effects based on whether people are prevention- or promotion-focused. They highlight the need for further testing in fields other than health.     28  1.3.2 Mental models As mentioned above, framing effects are moderated by individuals’ knowledge and mental structures (Shen, 2004). These mental structures are labeled differently depending on the field or the author, but the most common terms used are: mental models, mental schemas,14 cognitive maps, cultural models,15 frames in thought, interpretative frames and audience frames (Chong & Druckman, 2007; Kearney & Kaplan, 1997; Lakoff, 2010; McDonald, 2009; Morgan, Fischhoof, Bostrom, & Atman, 2002; Scheufele, 1999; Scheufele & Nisbet, 2007; Shen, 2004). This investigation will employ the term ‘mental model’ due to its frequent use in the literature.  Mental models are conceptual structures or mental organizing frameworks that guide individuals’ information processing and understanding of the world (Cox, 2013; Scheufele & Nisbet, 2007; Shen, 2004). They are simplified and incomplete models of reality that include the underlying beliefs, associations and assumptions attached to concepts or issues (Morgan et al., 2002). For instance, using mental model analysis, Bostrom, Morgan, Fischhoff and Read (1994) found that people often confuse ozone depletion with the greenhouse effect. Mental models are used to “[…] categorize new information quickly and efficiently, based on how that information is defined or described by the media” (Scheufele & Nisbet, 2007, p. 255). Mental models are acquired by personal experience, by learning from others and by theory (i.e. how things are supposed or should work) (Rosner, 1995). Researchers differ in regards to where mental models exist in memory. Some argue that they are located in working memory and, thus, are transitory (i.e. they change depending on the task at hand); however, others believe that these are stable structures stored in long-term memory (Salter, 2015).  In order to influence the interpretation of an issue, frames in communication must activate some aspects or dimensions of people’s mental models (Chong & Druckman,                                                  14 The differences between mental models and mental schemas is not fully clear in the literature (Jones et al. 2011 based on Johnson-Laird 1983). Some authors have used them interchangeably (McDonald, 2009), while others have characterized schemas as static and generic (i.e. broad) and mental models as more dynamic and specific (Jones et al., 2011).  15 Cultural models are schemas shared by a cultural group.    29  2007; McDonald, 2009). However, the comprehension around a topic will become problematic if the appropriate schema is not in the audience memory or is not provided in the message (Lakoff, 2010; Scheufele & Nisbet, 2007). Moreover, if the framed information does not fit individuals’ pre-existing beliefs, it will likely be ignored or rejected (Price, Tewksbury, & Powers, 1997). In this way, mental models influence our selective attention and may reinforce confirmation bias (Jones, Ross, Lynam, Perez, & Leitch, 2011). Thus, they may be difficult to change (Rosner, 1995), and, therefore, communications should be carefully crafted so that they trigger meaningful associations with the listener. For example, Shen (2004) showed that frames that match the audience’s pre-existing schemas tend to be more effective than the ones that do not.  Research on mental models has been applied to different topics and fields. In the context of environmental and sustainability topics, mental models have been used to identify general understandings and common misconceptions about climate change (Bostrom et al., 1994), to explore conceptions and representations of natural systems (Bang, Medin, & Atran, 2007; Jones, Ross, Lynam, & Perez, 2014) and general human-environment interactions (Lynam & Brown, 2012), to delve into water use and management strategies (Pahl-Wostl & Hare, 2004; Stone-Jovicich, Lynam, Leitch, & Jones, 2011), to explore sustainability conceptions among farmers (Hoffman, Lubell, & Hillis, 2014) and even to encourage social learning (Pahl-Wostl & Hare, 2004).  Elicitation techniques for uncovering mental models vary greatly (Morgan et al., 2002). Some of the methods that have been used include: card sorting, where participants are asked to organize and arrange different concepts, and in some cases explain the reasoning behind their choices (Salter, 2015); content analysis, which is focused in analyzing the repetition of concepts; procedural mapping that is centered on identifying and delineating the processes behind the implementation of tasks (Carley & Palmquist, 1992); and interviews and/or questionnaires, which are used to delve directly into the content of the mental models (Morgan et al., 2002). Some studies use a combination and/or a modification of these methods. For example, for risk communication, Morgan’s et al. (2002) methodology first creates an expert model to map the most updated scientific    30  knowledge on a specific topic. Subsequently, lay people’s mental models are explored to allow for comparisons and identify misunderstandings and knowledge gaps in the perceptions of lay audiences. Finally, communications are drafted in a way to address people’s knowledge gaps and misconceptions. This methodology often uses open-ended and structured interviews to delve into the content of the mental models (Morgan et al., 2002). Multiple biases may arise at the moment of eliciting mental models. Some of these include: illusory expertise, when people do not express beliefs that could make them look like non-experts; illusory discrimination, the suppression of beliefs that are perceived as inconsistent by the respondent; and illusory correctness, when people try to conform to what they believe is correct rather than expressing what they genuinely think (Gentner & Whitley, 1997). Researchers studying mental models should be cognizant of the potential effects that different methodologies may cause on respondents. 1.3.2.1 Research gaps regarding studies on mental models Gladwin et al. (1997) claim that the mental models held in western cultures are a main source of unsustainability, as these are often biased towards anthropocentrism, disconnection, individualism and efficiency. ‘Unlearning’ some of the old problematic models and replace them with new ways of thinking is very important (Rosner, 1995). In this sense, Lakoff (2010) highlights the need to build an adequate system of mental frames within society, so that the dimensions of the environmental crisis can be more easily appreciated and grasped with. Although studies on mental models have increasingly been applied to natural resource management (Jones et al., 2011), few have been devoted to understanding people’s mental models in relation to ecological economics and related concepts (e.g. economic growth, limited resources, limits to growth, uneconomic growth). Moreover, “[...] effective communication and education require an understanding of people’s existing cognitive maps so that information may be framed in a way that encourages people to notice and integrate the new information rather than ignore or reinterpret it” (Kearney & Kaplan, 1997, p. 581). However, more research is needed to better understand the connection between frames in communication and    31  mental models, especially the role of mental models in moderating framing effects (Hine et al., 2016; Shen, 2004). This research gap is especially relevant for this dissertation, as no study to my knowledge, has studied the degree to which mental models can moderate how people perceive different messages related to post-growth economics.  1.4 General Methodology 1.4.1 Geographic scope This study was carried out in Canada, although the first objective (i.e. mental model analysis) took place in the Province of British Columbia (BC) only. Canada is a high-income nation that ranks very highly in human development – tenth in the world in 2015 (UNDP, 2016). Regarding ecological impact, Canada has one of the largest ecological footprints per capita in the planet at 7 hectares per person. Due to its large territory, abundant natural resources and small population, Canada’s biocapacity (and likely BC’s as well) still exceeds its footprint;16 however, if everyone in the world consumed as Canadians do, more than three planets would be needed (Global Footprint Network, 2012). Canada, and certainly BC, could be in a privileged position to plan for a smoother transition towards creating a sustainable economy, as it has achieved a high standard of living, while still having large amounts of natural capital. This is a special situation, as most industrialized countries place an ecological demand that far exceeds their biocapacity (Global Footprint Network, 2012). While the province of BC has taken important steps in implementing climate change policy, such as aiming to reduce GHG emissions by 80% by 2050 (from 2007 levels) and being the first jurisdiction in North America to put in place a revenue-neutral carbon tax (Government of British Columbia, 2016), it is yet to be seen whether these plans can be achieved or if they can be taken even further. The provincial government has also promoted a ‘Green Economy’ model which aims to achieve environmental sustainability, while still pursuing economic growth                                                  16 Biocapacity refers to the availability of ecologically productive land. Canada’s biocapacity has diminished from 24 hectares per capita in 1960 to 15 hectares per capita in 2012 (Global Footprint Network, 2012).    32  as the main objective; a plan that is strongly focused on technology and efficiency gains (Government of British Columbia, 2014). Similar objectives are being pursued by the Canadian government.17 Vancouver, the largest city in BC, has ambitious sustainability goals, aiming to become the greenest city in the world by 2020. Most of Vancouver’s targets in this regard are related to improving transportation and reducing waste, carbon emissions and the per capita ecological footprint (City of Vancouver, 2012).  The Green Party of BC, which had slightly more than 15% of popular support in the 2017 election,18 runs on a platform that is closely related to the principles of ecological economics. Moreover, BC’s residents elected the first green caucus in North America, as three green candidates won seats in the legislature. Furthermore, the Green Party’s influence has leaped from having a marginal influence in BC’s politics, to holding the balance of power in the province.  British Columbia and Canada offer a very interesting case study for this research as there is some tension between those that promote environmental sustainability and those that push strongly for exploiting and exporting the natural resources of the country. This emerging struggle is somewhat evidenced in the different perspectives around pipeline expansion. For instance, multiple rallies have taken place in BC (some supported by local municipalities and First Nation groups) for advocating climate action and coastal protection, but at the same time, the federal government has approved the construction of two new pipelines.19 This tension between different values and objectives could offer important data and fascinating insights, thus making the residents of Canada, and British Columbia, a reasonable target population for this investigation.                                                  17 See Canadian government budget report (March 22, 2016) A clean growth economy, Chapter 4 (149-167). Retrieved on May 4, 2017 from: http://www.budget.gc.ca/2016/docs/plan/ch4-en.html 18 Data as of May 2017 after BC’s provincial election. Retrieved on June 22, 2017 from: https://en.wikipedia.org/wiki/Green_Party_of_British_Columbia 19 On rallies against pipeline expansion see: Smith, C. (November 19, 2016) Major Vancouver rally and march scheduled today in opposition to Kinder Morgan pipeline, The Georgia Straight. Retrieved from on June 22, 2017: http://www.straight.com/news/831651/major-vancouver-rally-and-march-scheduled-today-opposition-kinder-morgan-pipeline On pipeline approval by the federal government see: Tasker, J.P. (November 29, 2016) Trudeau cabinet approves Trans Mountain, Line 3 pipelines, rejects Northern Gateway, CBC news. Retrieved on June 22, 2017 from: http://www.cbc.ca/news/politics/federal-cabinet-trudeau-pipeline-decisions-1.3872828     33  1.4.2 Methodological approach20 This mixed methods study incorporated both qualitative and quantitative approaches (Creswell, 2014). The overall methodology employed for tackling each objective is shown in Figure 1.2. Qualitative methods were used for addressing the first objective of identifying mental models of economy and environment, as they allowed for an in-depth exploration of people’s thought processes and the meaning granted to concepts (Creswell, 2014). Semi-structured interviews were ideal for delving into and uncovering mental models. For tackling the second objective on exploring participants’ attitudes, quantitative methods were employed as they allowed for surveying larger numbers of people and obtaining numeric trends about their attitudes. An online survey was the preferred data collection method. The methodologies for answering these two first phases of research constituted an exploratory sequential mixed methods design, as the qualitative research phase was used to inform the data collection methods (i.e. survey questions and variables) of the quantitative phase (Creswell, 2014).  For addressing the third objective on message framing effects, a mix of quantitative and qualitative methods were used. Quantitative methods were employed for testing theories and exploring relationships between variables. Quantitative data was collected with experimental close-ended items included in an online survey experiment. Qualitative methods allowed for an exploration of the reasoning behind people’s choices and reactions. Data were collected in the online survey with a series of open-ended items. By integrating both methodological approaches, it was possible to acquire a more complete picture of framing effects.                                                     20 A comprehensive description of the methodologies employed for addressing each research question is described in Chapters 2, 3 and 4, respectively.     34   Figure 1.2 Research approach used for addressing each objective posed in this study.  1.5 Structure of the Thesis This doctoral thesis contains five main chapters. This introduction (Chapter 1), provides the context for this investigation and introduces the most important research gaps identified in the relevant literature. In addition, it describes the main objectives that will be addressed by this dissertation, and briefly covers an overview of the methodological approaches.  Chapter 2 presents the findings for Objective 1 regarding the mental models held by the BC population about economic growth, society and the environment. It also delves into the factors that relate most strongly with the identified mental models (e.g. cultural values, demographic factors) and explores reactions to the alternative paradigm of ecological economics. The information presented in this chapter is employed in later Objective #3            Framing effects General approach Quantitative & Qualitative Method                  Online survey with Canadian sample (survey company) Analysis                     Kruskal-Wallis test and ordinal regression (SPSS) and thematic coding (NVivo) Research design            Experimental (between-subjects) Objective #1     Public’s mental models General approach Qualitative Research design       Cross-sectional              Method                     Mental model interviews with non-probability sample of BC population Analysis                 Cluster analysis (NVivo) Objective #2  Audience segmentation  General approach Quantitative Research design            Cross-sectional  Method                     Online survey with sample of Canadian population (survey company panel) Analysis                 Latent class analysis (Latent Gold) and correlational (SPSS)    35  stages of the investigation, specifically for informing data collection methods of Chapter 3 and exploring if the mental models identified replicate in a larger population. Chapter 3 addresses Objective 2 by examining the prevalence of expansionist and ecological attitudes among the Canadian population and presenting the results of the audience segmentation based on these attitudes. It also explores the relationship between each identified segment and other variables (e.g. sociodemographics, participants’ levels of involvement with economic and environmental issues). The identified segments are used in Chapter 4 for testing interaction effects in message framing. Chapter 4 addresses Objective 3 by testing the effects of different messages (focused on transitioning to an economic system not centered on economic growth) on people’s cognitive responses, emotional reactions and attitudes. In addition, it explores the influence of other variables in moderating framing effects (e.g. audience segments, participants’ level of involvement, sociodemographics).  Finally, Chapter 5 synthesizes the main findings of the overall research, locates it within the relevant literature, summarizes the main implications and outlines pathways for future research.      36  2. Exploring Mental Models about Economic Growth, Society and the Environment in the BC Population  This research chapter explores the mental models held by the BC public on issues surrounding economic growth, society and the environment, and probes reactions towards the main premises of ecological economics. It also delves into the factors that correlate most strongly with the identified mental models (e.g. cultural values, sociodemographic factors). The chapter provides exploratory data and does not aim to generalize its results to the population at large. The information obtained here is employed in later stages of this investigation (see Chapter 3 in this dissertation).  This chapter is structured as follows: 1) the introduction describes the need to transcend current unsustainable worldviews and states the main objectives of this research; 2) the methodology describes the methodological approach, data collection, and analysis; 3) the results describe the main mental models identified in the sampled population; 4) the discussion examines the most important implications of the results and relates them to relevant literature; and, 5) the conclusions summarize the main findings of this empirical chapter.  2.1 Introduction  The world is facing an unprecedented ecological crisis. The planet’s capacity to provide goods and services is shrinking, while our human load – in terms of population size and consumption – keeps growing (Rees, 2003a). The world has been in a state of overshoot for more than three decades (Wackernagel & Rees, 1996), which means that resources are being used at a faster rate than they can be regenerated. Four out of nine ‘planetary boundaries’ have already been crossed: climate change, biosphere integrity (e.g. loss of biodiversity), land system change and alterations to biochemical flows (e.g. nitrogen and phosphorous cycles) (Rockstrom et al., 2009; Steffen et al., 2011).21 This situation is only worsened by the increasing levels of wealth and income inequality that exist within and                                                  21 Crossing a boundary increases the risk of permanent and sudden environmental change (Rockstrom et al., 2009).    37  between countries (Dietz & O’Neill, 2013), which confirms the need for greater equity in the use of limited natural resources (Raworth, 2012). In this context, various scientists argue that the global community will face remarkable challenges during the present and upcoming decades (Heinberg, 2011; Millenium Ecosystem Assessment, 2005; Rees, 1995; Steffen et al., 2011; Union of Concerned Scientists, 1992).  It is increasingly being recognized that the current economic model is at the centre of this crisis. The economy depends on the use and transformation of natural resources (Rees, 2003a) and the unparalleled levels of economic growth seen during the last decades have arguably been the result of the ongoing extraction of cheap fossil fuels (Heinberg, 2011; Nikiforuk, 2012) and the unsustainable use of resources (UNEP, 2011). Large international organizations (e.g. UNEP, OECD) and multiple governments have started to recognize the need to change the current economic paradigm (OECD, 2011; UNECE & FAO, 2014; UNEP, 2011) and some have even admitted that “getting the economy right” is key for attaining sustainability (UNEP, 2011, p. 17). However, few have gone so far as to acknowledge that the source of unsustainability may lie in the economic system’s need for continuous growth.  Since the last century, the concept of economic growth has often been conflated with the idea of progress (Victor, 2008) and GDP is often used as a proxy for rising living standards and well-being (Heinberg, 2011). An example of this follows: “A large GDP does not ensure that all of a nation’s citizens are happy, but it is surely the best recipe for happiness that macroeconomists have to offer” (Mankiw & Scarth, 2004, p. 44). The importance granted to this indicator by economists and politicians cannot be overstated; most policies are assessed according to their impacts on GDP, and macroeconomics is largely focused on understanding and explaining the dynamics of this indicator (Dietz & O’Neill, 2013). The preoccupation with unlimited economic growth has filtered down to most governments and institutions around the world: “The overarching priority of economic growth was easily the most important idea of the twentieth century” (McNeill, 2000, p. 336). As mentioned in the introduction, unless otherwise stated, the use of the term ‘economic growth’ in this dissertation refers to GDP growth.    38  For more than four decades, some ecological economists have stressed the need to move into an optimally-sized sustainable economy, more aptly called a steady state economy (Daly, 1991; Rees, 2011). A steady state economy is described as one with relatively constant stocks of materials and energy and stable numbers of people, that uses resources at or below their regeneration rate (Daly & Farley, 2011). In ecological economics, the economy and society are seen as part of nature and fundamentally dependant on it for their growth and development. In this view, economic growth is not seen as the ultimate goal of economic processes, but rather as a limited and temporary mechanism to achieve specific goals. More importantly, it is recognized that there are trade-offs between economic expansion and maintaining ecosystems, so indefinite material growth is seen as very costly, harmful and unattainable in the long term (Daly & Farley, 2011). Fundamental to the ecological economics perspective is that achieving a high quality of life and well-being for society can be accomplished with lower rates of material and energy throughput (Dietz & O’Neill, 2013).  Nonetheless, moving to a different economic paradigm will involve shifting the fundamental vision upheld by the current economic model. Among other aspects, the western worldview has been characterized as anthropocentric, individualistic, reductionist, mechanistic and rationalistic (in opposition to ecocentric, communitarian, holistic, organic and intuitive) (Gladwin et al., 1997). These traits are reflected in the Dominant Social Paradigm (DSP), which is described as a worldview that evokes optimism and confidence towards science and technology, free markets, efficiency, unlimited economic growth and limited government involvement (Dunlap & Van Liere, 1984; Kilbourne, Beckmann, & Thelen, 2002; Shafer, 2006). The DSP is so deeply ingrained that it is rarely ever questioned by political actors of any conviction (Shafer, 2006) and while it is not endorsed by all the population, it certainly guides to a great extent individual and societal behaviour (Dunlap & Van Liere, 1984). Despite the prevalence of the DSP, it has been argued that a new worldview – the New Ecological Paradigm (NEP) – may be emerging and starting to replace the basic tenets of the DSP (Dunlap, Van Liere, Mertig, & Jones, 2000). A high NEP worldview recognizes the delicate balance of nature, sees humankind as part of nature and subject to its rules,    39  admits nature’s intrinsic value (i.e. anti-anthropocentric), acknowledges limits to growth, and/or recognizes more seriously the possibility of an ecological crisis (Dunlap et al., 2000). High support for the DSP has been found to be inversely correlated to environmental attitudes and concern (Kilbourne et al., 2002) and with high NEP scores (Shafer, 2006). Along similar lines, Rees (1995) deconstructed the western dominant economic worldview and distinguished two opposing paradigms in the sustainable development debate – the expansionist and ecological worldviews. The expansionist perspective (which is more aligned with neoliberal economics) is characterized by considering: the economy and economic growth as unrestrained by the environment; humans as masters of nature; technology as able to compensate for depleting natural capital; free markets as instruments for achieving sustainability; and growth as the solution for social and environmental problems (Rees, 1995). On the other hand, the ecological worldview (which is more aligned with ecological economics) is characterized by recognizing that: the economy is embedded and dependent on the ecosphere; material and income growth are constrained by the flow of goods and services from the environment (i.e. natural income); nature is seen as the real producer, whereas the economy transforms and consumes what is generated by nature; the economy is governed by physical laws (e.g. thermodynamics) with a unidirectional flow of useful energy into more dissipated and disorganized forms; prices are unreliable indicators of scarcity; and technology cannot fully substitute for complex ecological services such as biodiversity, climate regulation, soil fertility, among others (Rees, 1995). Rees (1995) argues that the expansionist worldview is a fundamental driver of multiple ecological problems, cautioning that, if this view remains prevalent and unchanged, it is likely to generate greater ecological deterioration and thus, increase the likelihood of political instability and economic decline. “[…] global ecological decline is the inevitable consequence of fundamental incompatibilities between the dominant, growth-oriented cultural paradigm and biophysical reality” (Rees, 2003b, p. 30). Along similar lines, many authors have highlighted the importance of ‘unlearning’ and replacing the problematic and unsustainable ways in which we see the world (Gladwin et al., 1997; Donella Meadows,    40  2008; Rees, 2010; Rosner, 1995). “What we face in the West today is not an ecological crisis, nor a crisis of economics, nor a crisis of structure. It is a crisis of the mind. A crisis of the stories we tell ourselves, of the position we wish to give ourselves in the creation, and of the purpose that we give to our existence" (Palmer, 1992, p. 178 quoted in Gladwin et al., 1997, p. 245).  Despite the importance of this topic for achieving sustainability, little research has been devoted to explore and deconstruct economic worldviews among the general population, in order to better inform approaches for shifting these worldviews towards sustainability. For instance, people’s mental models about economic relationships have not been well studied (Dixon, Griffiths, & Lim, 2014) and little current research, to my knowledge, has determined the actual prevalence of expansionist thinking among the general population. If there is any possibility that ideas for a new economy may take hold, it is imperative to understand people’s current attitudes and reasoning about these issues.  Given this context, the main objective of this research is to identify, describe and bring to light existing mental models about how people perceive the relationship between economic growth, society and the environment. A secondary objective is to explore reactions to some ideas of the alternative paradigm of ecological economics. Mental models22 have been described as mental organizing frameworks that shape people’s understanding of the world and guide information processing (Cox, 2013; Scheufele & Nisbet, 2007; Shen, 2004). They are simplified and incomplete cognitive models of external reality that include the underlying beliefs, associations and assumptions attached to concepts or issues (Jones et al., 2014; Lynam & Brown, 2012; Donella Meadows, 1998; Morgan et al., 2002). Mental models are acquired by personal experience, by learning from others and by theory (i.e. how things are supposed to work) (Rosner, 1995) and are influenced by our culture and language (Donella Meadows, 1998). Besides their importance in shaping our worldviews and in influencing the decisions that we make,                                                  22 The term ‘mental models’ is used in this investigation due to its frequent use in the literature. The terminology of these mental structures varies depending on the field or the author, with most commonly used terms being: mental models, mental schemas, cognitive maps, cultural models, frames in thought, interpretative frames and audience frames (Chong & Druckman, 2007; Kearney & Kaplan, 1997; Lakoff, 2010; McDonald, 2009; Morgan et al., 2002; Scheufele, 1999; Scheufele & Nisbet, 2007; Shen, 2004).     41  mental models are crucial in the uptake of new information, as they can be used to organize new data quickly and efficiently (Scheufele & Nisbet 2007, p. 255) and tend to filter out conflicting information (Price & Tewksbury, 1997; Shome & Marx, 2009). In this way, mental models influence people’s selective attention and confirmation bias (Jones et al., 2011) and, thus, can be very persistent and difficult to change (Rosner, 1995).  In the context of environmental and sustainability topics, mental models have been used to identify general understandings and common misconceptions about climate change (Bostrom et al., 1994), natural systems (Bang et al., 2007; Jones et al., 2014) and general human-environment interactions (Hoffman et al., 2014; Lynam et al., 2012; Pahl-Wostl & Hare, 2004; Stone-Jovicich et al., 2011). However, little research on mental models has been done in relation to people’s views about the interactions between economic growth, society and the environment.  Thus, this investigation aims to offer an initial empirical exploration of these understudied topics. In addition, it aims to generate theory and hypotheses on the structures of mental models and their relationships to other factors like sociodemographics and cultural cognition of risk. Sociodemographic factors, like political views, have been linked to environmental attitudes, as conservatives tend to show less environmental concern than liberals (Shafer, 2006). Regarding gender, women tend to have stronger pro-environmental attitudes and behaviours than men (Zelezny, Chua, & Aldrich, 2000). Also, the notion of cultural cognition of risk has been increasingly favoured as a good predictor of policy support for a number of topics, including environmental issues (Kahan & Braman, 2006), as egalitarians tend to be more supportive of climate policies and environmental regulation, than individualistic and hierarchical individuals (Kahan & Braman, 2006; Leiserowitz, 2006).  In summary, the research questions that this chapter aims to address are: 1) What are the mental models that the public hold in relation to economic growth, society, and the environment? 2) What factors correlate most strongly with the identified mental models? Are they related to cultural cognition of risk and sociodemographic factors (e.g. gender,    42  place of residence, political orientation)? 3) What are some of the initial reactions that emerge when participants are presented with ideas related to ecological economics? Do these reactions differ in relation to a particular mental model?  2.2 Methodology A qualitative approach was employed to address the research questions posed above. Qualitative methods are often used to delve into the inner experiences of people, to explore how individuals grant meaning and significance to things, to study new areas of research, and to develop hypotheses that can be later tested with quantitative methods (Corbin & Strauss, 2015). In addition, they allow the researcher to explore in greater depth research participants’ thinking and understanding of issues. This can be crucial in the interpretation of mental models and patterns of thought – as is done here – especially since this topic has not been previously explored. Despite the advantages of qualitative methods, an important drawback is that generalizations to a broader population cannot be made due to the non-probabilistic sampling strategies used.  2.2.1 Study area and recruitment of participants This study was carried out in the Province of British Columbia (BC) in Canada. As mentioned in the introductory chapter, BC offers a very interesting case for this research as it faces conflict between those that promote environmental sustainability and those that push strongly for exploiting and exporting the natural resources of the province. On one hand, BC is a global producer of products derived from natural resources. On the other hand, it is considered to be one of the ‘greenest’ provinces in Canada. Sixty British Columbians participated in this qualitative study. The purpose of the overall sampling strategy was to recruit a broad diversity of participants. Data collection was carried out from April to October, 2014 in Metro Vancouver (MV) and during June, 2014 in the BC Interior (BCI), specifically Prince George and Williams Lake. These municipalities were selected in order to explore if there were noticeable differences in opinions between participants from large, medium and small urban centres. A total of 36 participants were sampled in MV and 24 in the BCI. Multiple mechanisms were used to    43  recruit a diversity of participants. General internet searches of professional associations, private businesses and other institutions were used to identify potential participants that work in multiple sectors (e.g. education, transportation, forestry, health, government). Approximately 35 personal emails were sent to potential participants, nine of whom agreed to participate.23 Additionally, approximately 80 short letters with information about the research were delivered to people frequenting public areas (e.g. malls, bus and metro stations, fast food restaurants).24 Twenty-eight participants were successfully recruited in this manner. Finally, 23 participants were recruited using snowball sampling through initial participants.25 An incentive of $15 CAD was provided to each person, which was delivered in cash or donated to charity depending on each person’s preference. The sample sociodemographics are summarized in Table 2.1. These are similar to the sociodemographics of the BC population in terms of gender, age and ethnicity, except that age groups 30 to 39 and 50 to 59 were over sampled (by about 10% each), while those more than 70 years old were under sampled. Also, respondents identified as First Nations were slightly overrepresented. In terms of education, people with college or university education, especially those with graduate degrees were overrepresented, while people with less than high school education were underrepresented.26 As shown in Table 2.1, the sample characteristics between MV and the BCI were quite similar, except that most respondents with graduate education and East Asians were recruited in MV, while First Nations were recruited only in the BCI.                                                   23 This sampling strategy was more successful in MV. Only one participant in the BCI was recruited in this manner.  24 The researcher delivered these information letters by approaching random people in public places and by posting them in public billboards. 25 Some respondents recommended family, friends, neighbours, colleagues or acquaintances who they thought could be interested in the interview and/or could provide interesting perspectives. This sampling strategy aimed at finding a diversity of perspectives. 26 Data for establishing these comparisons was obtained from multiple sources. For age, data was collected from BC Stats, 2015 data, retrieved on February 2, 2017 from: http://www.bcstats.gov.bc.ca/StatisticsBySubject/Demography/PopulationEstimates.aspx  For ethnicity, data was obtained from Wikipedia, 2006 data, retrieved on February 2, 2017 from: https://en.wikipedia.org/wiki/Demographics_of_British_Columbia#Ethnicity  For education, data was gathered from Statistics Canada, 2009 data, retrieved on February 2, 2017 from: http://www.statcan.gc.ca/pub/81-599-x/81-599-x2012008-eng.htm    44  Regarding occupation, 65% of all participants were still active in the workforce, 17% were retired, 8% were students (of which half worked part-time), 3% were homemakers and 7% were unemployed. Occupations varied greatly. Some of the sectors represented included (from the most to the least common): education; communications; forestry; government; health; transportation; retail; technology; accounting; mining; real estate; and tourism. 2.2.2 Data collection methods Elicitation techniques to measure mental models vary greatly and there is no single preferred methodology27 (Morgan et al., 2002; Salter, 2015), with some studies using multiple methods (Jones et al., 2014; Pahl-Wostl & Hare, 2004). A few elicitation methods include: sorting techniques (organization and arrangement of different concepts) (Bang et al., 2007); mapping (visual representation of concepts and their connections) (Kearney & Kaplan, 1997); procedural mapping (focused in processes behind tasks) (Carley & Palmquist, 1992); and interviews or questionnaires (Bostrom et al., 1994; Morgan et al., 2002). Although each method may be appropriate for a specific context, an advantage of interviews is that they allow respondents greater flexibility than other approaches and a deeper exploration of participants’ thinking (Kearney & Kaplan, 1997). Therefore, they were the preferred method for exploring participants’ mental models. One-on-one semi-structured interviews took place in coffee shops, private offices, libraries and parks. On average, each interview lasted approximately 60 minutes and addressed the following topics: initial reactions towards economic growth; benefits and costs of economic growth; opinions on development and progress; views on limits to growth; connections and associations between the economy, society and the environment; views on sustainability; and initial reactions towards some of the ideas of ecological economics (see Appendix A for the complete interview protocol). Generally, for each topic covered, the discussions started with an open question which had little guidance and direction. This was followed by more structured probes and questions. As a way of                                                  27 For a more detailed review of mental models elicitation techniques, see Salter (2015).     45  reducing potential biases (see chapter’s introduction), the interviewer emphasised to participants that the interview was about opinions and that there were no correct answers.  Table 2.1 Sociodemographic characteristics of the sample, classified by participants from MV and BCI. Sociodemographics Place of residence MV BCI All sample N 36 24 60 Gender (% male) 50.00% 50.00% 50.00% Political  Conservatives 14.00% 8.00% 12.00% Moderates 31.00% 17.00% 25.00% Liberals 25.00% 21.00% 23.00% Other 19.00% 37.00% 27.00%a None 11.00% 17.00% 13.00% Age  <20 11.00% 0.00% 7.00% 20-29 11.00% 17.00% 13.00% 30-39 17.00% 33.00% 23.00% 40-49 11.00% 8.00% 10.00% 50-59 31.00% 17.00% 25.00% 60-69 14.00% 17.00% 15.00% >70 5.00% 8.00% 7.00% Education Less than High School 0.00% 8.00% 3.25% Some High School 6.00% 0.00% 3.25% Finished High School 17.00% 21.00% 18.25% Technical Training 6.00% 13.00% 8.25% Some College University 19.00% 21.00% 20.00% Bachelor Degree 19.00% 21.00% 20.00% Graduate Degree 33.00% 17.00% 27.00% Ethnicity White 72.00% 67.00% 68.00% South Asian 3.00% 8.00% 5.00% East Asian 17.00% 4.00% 11.00% Indigenous  0.00% 29.00% 11.00% Other 8.00% 0.00% 5.00% a Of these, 7% were NDP, 8% Green, 5% Socialist and 7% other.    46  Visual aids assisted in the elicitation of the mental models. They helped to uncover participants’ beliefs about issues and assisted the interviewer in explaining concepts related to ecological economics (see Appendix B for a complete illustration of these). Specifically, in Section 2 of the interview, a chart of global economic growth since the 1960s was presented (see Appendix B.1) to help participants visualize the rate of growth and the current size of the world economy. In Section 5 of the interview, circular figures of multiple sizes were employed (see Appendix B.2) to assist participants in describing how they perceive the relationship between the economy, society and the environment (Figure 2.1 depicts some examples). Each participant freely moved these templates around and explained his or her conception of the relationship between these systems. Interestingly, some participants were very keen and interested in using these visual aids. Finally, other graphics (see Appendix B.3) were used to assist the interviewer to describe the main proposals of the steady state economy.   Figure 2.1 Examples of how some participants depicted the relationship between the economy, society and the environment. Hand-written notes were taken during all conversations. With the exception of one interview, all were taped with an audio recorder.28 Of the remaining 59 interviews, four were not well recorded due to technical issues. After the interview, each participant completed a short self-administered survey, with sociodemographic questions (see Appendix C). The survey also included a short version of the cultural cognition of risk scale (obtained from Kahan, Jenkins-Smith, & Braman, 2011) (see Appendix D). Cultural cognition asserts that people tend to adjust their beliefs and perceptions of risk according to values that support their ‘group’ identity (Kahan et al., 2011). There are two main                                                  28 One participant did not give her consent to be recorded.    47  attitudinal dimensions; the first is related to individualistic or collective orientations regarding social organization (e.g. individualistic or communitarian individuals), while the second is about the perceived need for more or less socially stratified rules to control behaviour (e.g. hierarchical or egalitarian individuals) (Kahan et al., 2011; Leiserowitz, 2006). The first section of the cultural cognition survey measures participants’ predispositions towards individualistic versus communitarian values, whereas the second section evaluates egalitarian versus hierarchical values (see Table 2.2 or Appendix D).   All participants completed the sociodemographic survey, while three did not fill out the cultural cognition survey in its entirety, mainly due to a lack of time. Overall, many participants expressed that they found the interview process interesting, as they had not thought with such depth about these issues before.29 Weeks after the interviews took place, I received three emails from participants who were grateful for the experience, with one mentioning changes in his business practices as a result of this process. 2.2.3 Data analysis Interviews were transcribed verbatim into Microsoft Word and transferred to NVivo 10 Qualitative Software. Sociodemographic data was first coded into Excel and then inserted into NVivo as a Casebook. Data on cultural cognition was coded and analyzed in Microsoft Excel and in IBM SPSS statistics 23 software. In order to maintain confidentiality, each participant was assigned a unique code, which referred to his or her place of residence (i.e. R for rural BCI, U for urban MV), gender (i.e. M for male, F for female) and the order of the interview. For example, RM5 was the fifth man to be interviewed in the BCI, while UF7 was the seventh woman to be interviewed in MV. Once the interviews were input into NVivo, a thematic analysis was carried out. All data were classified into general descriptive codes, most of which were based on the interview                                                  29 For instance, UF2 shared her enthusiasm about the new knowledge she gained in the interview: “Well, I’m glad that I got to hear about ecological economy, I’m going to have to look this up, I’m very, very, very interested.”       48  protocol (i.e. theory-driven deductive process), while some emerged from an inductive analyses of the data. Themes derived from the interview protocol included: opinions and initial impressions on economic growth, views about limits to growth, thoughts and understanding related to sustainability and views on ecological economics, among others. Themes that were derived in an inductive manner from the data included: views about current society and human nature, respondent’s perceptions on the environment and opinions about overpopulation, among others. Additional rounds of more detailed coding were carried out to classify the data into more specific categories and subcategories. This process yielded about 20 general ‘umbrella’ codes and more than 400 subcodes. In order to ensure consistency while coding, a codebook was developed in Microsoft Excel with a description of the content of each code and subcode. In addition, while coding, multiple matrices were generated in Microsoft Excel for organizing, condensing and visualizing the data more easily. In total, nine matrices were created, depicting and classifying information for each participant in relation to the multiple themes identified by the analysis.30  After multiple rounds of coding and categorizing data, NVivo 10 Cluster Analysis was used to identify homogenous groups of participants. Cluster analysis is an exploratory and descriptive method that aims to identify groups whose members are relatively similar to each other and different from the rest (Everitt, Landau, Leese, & Stahl, 2011; Norusis, 2009). The success of the methodology is determined by finding ‘meaningful clusters’ (Bartholomew, Steele, Moustaki, & Galbraith, 2008), in the sense that the groups identified should be interpretable and explainable.31 NVivo 10 groups respondents based on coding similarity. Specifically, NVivo calculates a similarity index between each pair of sources using a correlation metric, such as Pearson, Jaccard or Sorensen coefficients. For calculating the similarity index, NVivo builds a matrix with sources (i.e. participants) as rows and nodes (i.e. codes) as columns. If a source is coded in a node, then it is                                                  30 Binary coding was often used in these matrices in order to make quick summations and identify recurring patterns.  31 Everitt et al. (2011, p. 4) explain the nature and purpose of cluster analysis: “So it should be remembered that in general a classification of a set of objects is not like a scientific theory and should perhaps be judged largely on its usefulness, rather than in terms of whether it is ‘true’ or ‘false’”.    49  granted a value of 1 and if it is not coded, then it is valued at 0. Based on the similarity index, sources are clustered into groups with a complete linkage (farthest neighbor) hierarchical clustering algorithm, so that sources coded similarly are closer together and the ones coded differently are further apart. The default number of clusters is 10, although the researcher can choose any desired number of groups.32 Cluster analysis was run iteratively with different similarity metrics (i.e. Jaccard, Pearson and Sorensen) in NVivo 10, in order to determine which clusters were consistent and which were less steady. In addition, some codes that were deemed as less relevant for the research (e.g. views on the past, opinions on overpopulation) were removed from the analysis in order to explore their effects on the clustering. Appendix E provides examples of NVivo output using different coding and different similarity metrics. After this iterative process, a five cluster solution clearly emerged from this analysis, with each cluster comprising between 5 and 10 participants. In order to interpret each cluster, a conceptually clustered matrix was developed33. This type of matrix brings together the major topics, variables or themes for facilitating the analysis (Miles, Huberman, & Saldana, 2014). The most relevant themes for this research obtained during the coding process were placed in rows and each of the clusters were placed in separate columns. Responses for each participant within each cluster were deconstructed and quantified, which allowed identifying important similarities within participants in a group and differences between those in other groups.34 Upon completion of this analysis, it was possible to start characterizing and describing each of the five clusters in terms of a shared mental model. In addition, participants that were not clearly assigned to any group during NVivo’s analysis were allocated to the cluster with which they shared the greatest similarity. Lastly, a new sixth cluster was created, comprised of                                                  32 Results are, by default, displayed as a dendrogram (as shown in Appendix E), although other graphical representations are available. The similarity index (correlation coefficient in coding) between sources is also provided in NVivo’s output.    33 In a simplified way, Table 2.4 illustrates the structure of this matrix. 34 For instance, views on limits to growth and the perceptions on human nature were clearly different between clusters.    50  participants that shared some similar characteristics between them and that did not clearly belong to any of the other groups.  Data on cultural cognition of risk were tabulated and analyzed in Microsoft Excel. Each point in the Likert-type scale was given a value from 1 to 6 (Table 2.2 illustrates how the values were assigned), the middle point being 3.5. For group predisposition (i.e. individualistic or communitarian), a value closer to 6 indicated a highly communitarian worldview (i.e. high group), whereas values closer to 1 indicated higher individualistic tendencies (i.e. low group). For grid predisposition (i.e. hierarchical or egalitarian), a value closer to 6 showed a higher hierarchical worldview (i.e. high grid predisposition), whereas a value closer to 1 indicated higher egalitarian perspectives (i.e. low grid predisposition) (Kahan & Braman, 2006; Kahan, Braman, Gastil, Slovic, & Mertz, 2007). For each participant, computed values were placed on scatter plots, which enabled quick comparisons between individuals and between groups. IBM SPSS statistics 23 software was used to determine if differences between groups were statistically significant.      51  Table 2.2 Values assigned to each scale item in the cultural cognition Likert-type scale.   Cultural Cognition Scale Group Predisposition  Measures People in our society often disagree about how far to let individuals go in making decisions for themselves. How strongly you agree or disagree with each of these statements? Strongly Disagree Moderately Disagree Slightly Disagree Slightly Agree Moderately Agree Strongly Agree Individualism  1. The government interferes far too much in our everyday lives. 6 5 4 3 2 1 Community 2. The government should do more to advance society's goals, even if that means limiting the freedom and choices of individuals.  1 2 3 4 5 6 Individualism 3. It's not the government's business to try to protect people from themselves. 6 5 4 3 2 1 Community 4. Sometimes government needs to make laws that keep people from hurting themselves. 1 2 3 4 5 6 Individualism 5. The government should stop telling people how to live their lives. 6 5 4 3 2 1 Community 6. Government should put limits on individual choices so they don't get in the way of what's good for society.  1 2 3 4 5 6 Grid Predisposition  Measures People in our society often disagree about issues of equality and discrimination.  How strongly you agree or disagree with each of these statements?  Strongly Disagree Moderately Disagree Slightly Disagree Slightly Agree Moderately Agree Strongly Agree Egalitarian 1. Discrimination against minorities is still a very serious problem in our society. 6 5 4 3 2 1 Egalitarian 2. We need to dramatically reduce inequalities between the rich and the poor, whites and people of color and men and women. 6 5 4 3 2 1 Hierarchy 3. Society as a whole has become too soft and feminine. 1 2 3 4 5 6 Hierarchy 4. We have gone too far in pushing equal rights in this country. 1 2 3 4 5 6 Hierarchy 5. It seems like minorities don't want equal rights, they want special rights just for them. 1 2 3 4 5 6 Egalitarian 6. Our society would be better off if the distribution of wealth was more equal. 6 5 4 3 2 1 Note: Scale was obtained from Kahan et al. (2011).    52  2.3 Results  Six clusters, labelled Clusters A, B, C, D, E and X, were identified in this analysis, each representing a different mental model.35 Cluster X (n=8) has been removed from the upcoming description and analysis in this chapter, as there was not enough data from participant’s responses to provide a clear understanding of how this group interprets the spectrum of ecological and economic worldviews, which is the main focus of this research chapter.36 This section first summarizes the perceptions within each cluster, while providing sociodemographic information for each group. It also provides a more detailed account of the main points of difference between clusters, classified by the main themes that emerged during the analysis. This is followed by a description of the main reactions within each cluster towards the ideas of ecological economics. Finally, results on cultural cognition of risk are presented. For easier identification, each cluster was given a nominal label which embodied the most prominent views found in each group. Table 2.3 describes the intended meaning of each label. However, for maintaining an adequate reading flow, the letter nomenclature (i.e. A, B, C, D and E) will be commonly used when referring to the clusters. The prevailing opinions found in Clusters A through E, arranged by the emergent themes of the analysis, are summarized in Table 2.4. The main themes included: initial perceptions towards economic growth; views about limits to growth; perceptions towards people and                                                  35 The stability of membership in each cluster varied. As reflected in Appendix E, Cluster A was the most stable group, with 10 participants consistently assigned here. Similarly, Cluster X was quite stable with five participants. Clusters D and E were relatively consistent, although less than the previous groups. On average, five participants were consistently grouped in Cluster D, while nine were consistently grouped in Cluster E. Cluster B was the least consistent of all, fluctuating between 10 and 14 members, of which six participants were consistently clustered here and seven others were often grouped here, but not always. 36 The data seems to indicate that participants in this group (which has been labeled ‘The Uninformed’) were generally unaware of and unfamiliar with many of the topics and terminology covered during the interviews, like the terms economy, economic growth and sustainability, among others. Most in this cluster seemed to be more concerned about affording the basics of life (e.g. income, employment), than about big macro-concepts (like the ones touched by this research). Nonetheless, this does not mean that they were not aware about multiple environmental issues and ecological problems. Demographically, Cluster X comprised 3 men and 5 women. This group had the largest proportion of First Nations (50% were Indigenous and 50% were white) and of participants with lower levels of formal education (i.e. many with high school or less than high school education).     53  society; views on the relative importance of different systems; views on sustainability; and reactions towards ecological economics.   Table 2.3 Label given to each cluster and brief description of its intended meaning. Label Description A The Expansionist Rejects the idea that there are limits to economic expansion, because human ingenuity and technology will come up with solutions. B The Hesitant Undecided about the existence of limits to economic expansion, because natural resources are ultimately finite, but human ingenuity and technology may come up with appropriate solutions. C The Indifferent Admits constraints to economic expansion, due to various reasoning (e.g. unlimited growth seems unnatural, financial debt). D The Green Recognizes limits to economic expansion due to natural resource availability. Limits are often seen as physically and/or temporally distant. E The Ecological Acknowledges limits to economic expansion due to ecological constraints. Limits are often seen as proximate and impending.  Slightly less than half of all participants were classified into Clusters A and B, while 40% were categorized into Clusters C, D and E.37 Cluster B was the largest group containing 28% of participants, followed by Cluster A that contained 18% of the sample. The smallest group was Cluster C with 12% of participants. As a general trend, many participants in Clusters A and E seemed to have thought beforehand about some of the issues discussed and presented relatively consistent and stable opinions. This tendency was not as evident for participants in Clusters B, C and D. A few individuals even expressed surprise by their own answers, reflecting that they have not thought about these issues before with much depth. That said, these observations should be interpreted with caution, as the flexibility or rigidity of the mental models identified in this research was not formally measured.                                                   37 The remaining participants were classified into Cluster X.    54  Table 2.4 Summary of the most prominent perspectives and sociodemographics in each cluster.  Themes Cluster A  n=11 (18%) The Expansionist Cluster B  n=17 (28%) The Hesitant Cluster C  n=7 (12%) The Indifferent Cluster D  n=8 (13%) The Green  Cluster E  n=9 (15%) The Ecological Perceptions of economic growth Very positive. Benefits outweigh costs. Growth = progress and advancement.  Positive. Benefits usually outweigh costs. Growth is inevitable and necessary. Ambivalent, although slightly positive. Unsure about the balance of benefits and costs. Neutral and slightly negative. Sometimes negatives of growth outweigh the benefits. Largely negative. Costs outweigh benefits, mainly due to environmental reasons. Views on limits to growth No limits to growth, because human ingenuity is infinite. Unsure about limits. Resources are finite, but human ingenuity is infinite. Yes, there are limits. Unlimited growth is unnatural, debt is large and resources are finite. Yes, there are limits. Natural resources are finite.  Yes, there are limits, mainly due to ecological reasons (going beyond natural resource scarcity). Perceptions of people and society Very optimistic. Humans are very smart, creative and innovative. Fairly optimistic. People are very smart, but we are also greedy and selfish. Somewhat negative. People are greedy and materialistic, but also very smart. Less optimistic. People are greedy, consumeristic and hard to change. Not optimistic. People are greedy, don’t think in advance, nor change quickly. Relative importance of economy, society and environment Mixed views. For some, the economy is the most important. For a few it is society, and for others, it is the environment. Mixed views. For some, all are important. For others, it is the environment, the economy or society. Mixed views. For some, all are important. For others, it is society and the environment, or the economy and society. For most, the environment is the most important system. For most, the environment is the most important system. Views on sustainability Many (industrialized) countries are sustainable or heading there. Growth and sustainability are compatible through efficiency and technology. Growth is a pre-requisite for environmental protection. Mixed views. For some, we are sustainable or making efforts; for others we are not. Growth and sustainability could be compatible, through innovation and using more renewable resources. Mixed views. For some, we are sustainable or making efforts; for others we are not. Growth and sustainability could be compatible for some, with increased regulation and lower rates of growth. The world is not sustainable. Limits may be reached sometime this century. Growth and sustainability may be compatible for a few through lower rates of growth and true costing. We are not sustainable. Collapse is likely and possibly will happen in the near future. Most believe growth and sustainability are not compatible.       55   Themes Cluster A  n=11 (18%) The Expansionist Cluster B  n=17 (28%) The Hesitant Cluster C  n=7 (12%) The Indifferent Cluster D  n=8 (13%) The Green  Cluster E  n=9 (15%) The Ecological Reactions to ecological economics Closed to very closed about ideas. Pros: Make sense for some. Cons: Ideas do not account for technology and ingenuity; decision-making issues; human nature (e.g. greed). Open or ambivalent about ideas. Pros: Make sense for most. Cons: Human nature (e.g. selfishness); political aspect; tough sell. Open to very open about ideas. Pros: Make sense and less risky for some. Cons: Human nature (e.g. greed); tough sell. Open to very open about ideas. Pros: Make sense for most. Cons: Human nature (e.g. materialism, status quo bias); tough sell; job generation. Very open towards ideas. Pros: Make sense to all. Cons: Human nature (e.g. status quo bias, greed); elite control; tough sell & political aspect. Sociodemographics Gender Age  Ethnicity  Political  10 men, 1 woman. Older participants: 7 were older than 50. 8 white, 2 Asians, 1 Indigenous. 5 Conservatives, 3 Liberals, 1 Moderate, 2 None.  9 men, 8 women. Mixed and balanced.  13 white, 3 Asians, 1 Persian. 2 Conservatives, 6 Liberals, 5 Moderates, 1 NDP, 1 Socialist, 2 None.  1 man, 6 women. Half were 30 to 39 and the other half were 60 to 69.  5 white, 2 Asians.  3 Liberals, 1 Moderate, 1 NDP, 2 None.   4 men, 4 women. Younger participants: 5 younger than 30.  6 white, 2 Asians.  3 Moderates, 2 Liberals, 2 Greens, 1 None.   3 men, 6 women. Mixed between 30 and 70. 8 white, 1 Indigenous.  1 Moderate, 1 NDP, 4 Greens, 2 Socialists, 1 None.        56  The following subsections describe the main points of difference between clusters, classified by the emergent themes seen in Table 2.4. Letters A, B, C, D or E have been added to each participant’s code, indicating the cluster to which each individual belongs. For example, UM4A represents a participant that belongs to Cluster A, whereas RM12B represents a participant that belongs to Cluster B. 2.3.1 Perceptions of economic growth Perceptions ranged from being very positive towards economic growth in Cluster A to being less positive about it in Cluster E (see Figure 2.2). As shown in Figure 2.3, words like prosperity, development and modernity emerged in Cluster A in relation to the economy, while there was a greater connection to corporations, resource extraction and destruction in Cluster E. In a similar way, when asked to assess the balance of benefits and costs of global growth, respondents in Cluster A generally believed that the benefits outweighed the costs, while the opposite emerged in Cluster E, mainly due to ecological degradation.  Figure 2.2 Characterization of views about economic growth. The following quotes reveal the differences in views: “100 to 1, absolutely! […]. If anybody tells you that economic growth hasn’t been a good thing, I would just close the books as they don’t know what they are talking about.” (UM12A)  “The damage done to Mother Earth is not recorded on there [...] It’s damage beyond repair.” (RM7E)   MORE POSITIVE PERCEPTIONS OF ECONOMIC GROWTH LESS POSITIVE A   B   C   D   E    57   Figure 2.3 Word clouds representing the initial associations brought forth by participants when thinking about economic growth. Participants in Cluster B leaned towards believing that the benefits of economic growth outweigh the costs, most in Cluster C were unsure of this balance, while increasing costs, especially related to the state of the environment, were recognized in Cluster D. Of all participants, about 40% were initially positive towards economic growth, 28% were ambivalent, and the remaining 32% were more negative about it. Notably, the different benefits and costs mentioned were quite similar across groups, as shown in Table 2.5. Table 2.5 Perceived benefits and costs of economic growth across all clusters.  Benefits of economic growth Costs of economic growth  Increased standard of living  Greater ease of life (e.g. more access to basic and non-basic goods and services)  Better health services  More and better education  More innovation and technology  Job generation   Environmental issues (e.g. resource depletion, pollution and climate change)   Increased inequality  Lower quality of life (e.g. quick pace of life)  Greater materialism  Health related issues (e.g. obesity)  Loss of social connections Note: This list only includes the six most mentioned aspects, organized from the most to the least mentioned. Cluster C Cluster D Cluster E Cluster B Cluster A   58  In addition, for about half of the participants in Cluster A and a few in Clusters B and C, economic growth was associated with ‘moving forward’; while for some in Cluster B, it was seen as inevitable and necessary. A few respondents (mainly in Cluster A and, to a lesser extent, B and C) recognized that growth can have high social and environmental costs, but it can also be clean, modern and equitable. In contrast, growth was often associated with mistrust in Cluster E, as some considered it a smoke screen used by politicians for benefitting corporations, while others linked it to falsehoods and unnecessary destruction. 2.3.2 Views on limits to growth Three very distinct opinions emerged on the topic of limits to growth (see Figure 2.4). All respondents in Cluster A (representing 18% of the sampled population) did not seriously believe in any constraints to growth, mainly because they believed that human ingenuity is infinite. This was explained by UM13A: “I don’t think there’s a limit on the resourcefulness of mankind.” Although it was acknowledged that natural resources are finite and possibly scarce, it was believed by some participants that these can be substituted and replaced with other resources (e.g. synthetic products) and that we will become more efficient and innovative in using these. For instance, two participants believed that humans’ dependence on nature could be reduced by increasing our reliance on artificial goods. Along similar lines, RM2A expressed his faith in our ability to find solutions: “I mean, there are materials yet to be discovered. I’m freakin’ sure that our kids are going to discover that. They are pretty smart.” This respondent also explained that, with enough research and technology, nature and even some extinct species, may be able to be recover. In this sense, about half of participants in Cluster A considered environmental impacts as reversible. Moreover, UM4A saw current challenges as the opportunities for tomorrow: “A lot of the things that we’re looking at and mourning, ‘Oh my goodness look at what they are doing’, may become raw materials just some 20 years from now […]. So, we can’t look at it too negatively.” Some participants also used historical data as evidence to point out the multiple innovations that have occurred in the   59  past century. For example, UM1A explained that due to technological progress, fossil fuels have become more abundant rather than scarcer.                Figure 2.4 Differences in views among clusters regarding the existence of limits to growth. Most participants in Cluster B (13 out of 17 representing 22% of the sampled population) were ambiguous about the existence of limits to growth. On the one hand, resource scarcity was considered an issue, especially considering increasing consumption from emerging economies. On the other hand, human ingenuity was also seen as a key factor for overcoming limitations. RF2B clearly expressed her ambivalence: “Yeah, yes and no. Yes, because some of them will ran out [of natural resources]. No, because I think, you know, things like... I don’t know, they’ll find something, they’ll find a way […].” Four participants talked about resource substitution and the potential for technology to find resources beyond the planet. UM10B explained: “Earth is a limited finite space in the universe and there’s supposedly, you know, infinite other spaces in the universe [...]. If we have access to that, we could support our species for millions of years theoretically [...].” Nonetheless, some individuals were more skeptical about this, especially when increasing global consumption is taken into account. Interestingly, although RM5 was generally optimistic, he also described some of the risks involved: “Well, that’s the hope, that we’ll be able to come up with substitutes […]. That’s not necessarily going to work. It may actually be, uh... you know, a real hard limit, but it has worked so far, so you kind of cross your fingers and keep going. That’s all we can do.” Historical evidence was also provided by few participants, but unlike Cluster A, it was used to argue that all great civilizations have collapsed in the past. Only three participants in this group believed that the economy can continue expanding indefinitely and one (UM15B) believed in limits, mainly because exponential growth seemed unnatural to him.  NO LIMITS VIEWS ON LIMITS TO GROWTH YES THERE ARE LIMITS UNSURE A    B   C           D                  E    60  All participants in Clusters C, D and E believed in limits to economic growth (representing 40% of the sampled population), but the reasoning provided in each group was different. On the one hand, continuous growth seemed unnatural for some in Cluster C. For example, RM4C explained that things work in cycles rather than in a continuous upward trend, while UF18C was not able to articulate her explanation: “I feel this in an intuitive level and a gut level and I haven’t intellectualized it [...].” Other reasoning provided in this cluster included the planet’s finite carrying capacity and the growing financial debt of global economies. Also, it was acknowledged that resource scarcity may impose a limit, but only after participants were probed with a specific question about this. In contrast, resource scarcity was the main explanation provided by participants in Cluster D.38 Distinctively, participants in Cluster E gave ecological reasons more broadly as major arguments for the existence of limits (e.g. climate change, ocean acidification), and not only related to natural resource scarcity. Six participants expressed concern about crossing important tipping points and the possibility of heading into collapse. For example, UF17E revealed her fear: “I try not to think about it because I wouldn’t sleep at night […]. We’re getting way to a point of no return.”  2.3.3 Perceptions of people and society39 The perceptions about people and society that emerged in this study point to important differences between groups, as shown in Figure 2.5. Participants in Cluster A, and to a lesser extent, in Cluster B, showed more optimism and confidence towards humankind, while participants in Clusters D and E had greater recognition of our limitations. Table 2.6 summarizes the views that emerged.                                                    38 This reasoning emerged naturally from all respondents except UF11D, who was explicitly asked: Do you think that natural resources will impose a limit? 39 No questions were directly aimed at exploring this topic. These opinions emerged naturally during the interview process in the context of what was being discussed.    61                     Figure 2.5 Characterization of opinions about people and society. Human ingenuity was the most cited trait in Clusters A and B. All members of Cluster A talked about ingenuity in some way or another. Among other things, people were seen as smart, motivated, creative, innovative, knowledgeable and fully cognizant of the impacts of our actions. UF16B revealed his faith in human resourcefulness: “We can always find... I think we are very motivated because as human beings we are kind of a species that we try to maybe even finding natural resources outside of Earth [...].” Despite this, about half of the participants in Cluster B recognized that people and society are greedy and too money-driven.  Table 2.6 Participants’ views on people and society categorized by cluster.  Note: The number of participants that mentioned a trait is presented (most participants mentioned more than one trait). a Idea that people tend to react only when they are faced with crisis, not beforehand.  Cluster Perceptions about people and society Ingenious Adaptable Oriented to compete Hard to change  Reactiona Loss Averse Desire for more Limited cognition Selfish Greedy & money driven Materialistic Entitled Short-term oriented Unequal wealth distribution A (n=11) 11 1 1 -- -- 1 2 -- 1 1 2 1 1 1 B (n=17) 12 5 4 1 6 -- 6 2 4 8 5 4 3 4 C (n=7) 3 1 -- 1 1 1 -- -- 1 5 5 4 -- 2 D (n=8) 1 1 -- 5 2 2 1 3 1 2 3 2 1 2 E (n=9) 1 -- 2 4 2 -- 1 5 2 7 3 1 -- 3 Total 28 8 7 11 11 4 10 10 9 23 18 12 5 12 MORE OPTIMISTIC PERCEPTIONS OF PEOPLE AND SOCIETY LESS OPTIMISTIC A    B   C           D                  E    62  Participants in Cluster C were slightly more negative and they often referred to society’s materialism, greed and sense of entitlement (although some also recognized our ingenuity). The most prevalent perspectives in Cluster D related to difficulties in changing habits and behaviours, but also to an incomplete understanding of the world (i.e. limited cognition). Finally, greed was the most prominent trait that emerged in Cluster E, with many believing that it will impede any positive change. Some respondents in this group talked about our limited cognition in relation to biases in our perceptions (e.g. denial) and our incomplete knowledge of ecological functions.  2.3.4 Views on the importance of the economy, society and the environment Various perspectives emerged about the importance of the environment, society and the economy in relation to one another. As shown in Table 2.7, 11 participants (largely from Clusters A and B and few from C) thought that the economy was the most important system (of these 11, three and two put it at the same level as society and as the environment, respectively). It was argued that a strong economy provides the funds to oversee and protect society and the environment. UM4A explained that nature should be protected only if there are the financial means to do so. Interestingly, no participants from Clusters D and E thought that the economy was the most important system. Table 2.7 Relative importance of the economy, society and the environment in relation to each other.          A (n=11) 4 2 2 1 -- 1 1 -- B (n=17) 2 1 3 1 1 -- 5 4 C (n=7) -- -- -- 1 1 2 3 -- D (n=8) -- -- 5 -- -- -- 1 2 E (n=9) -- -- 6 -- -- 1 2 -- Total 6 3 16 3 2 4 12 6 Note: The economy is symbolised by the contraction EC, society by SOC and environment by ENV. When two or three systems are portrayed together, it represents that both or all were seen as equally important. The question mark represents ambivalent answers or no data. Eleven participants (mostly from Clusters A and C) believed that society was the most important system (of these 11, three and four put it at the same level as the economy and   63  the environment, respectively). Some thought that everything should be done for the benefit of people. For instance, UM13A explained his view: “I think that’s key, so how do we support society. […] We are here because mankind is, we are above […].” RM6A thought that the environment should be protected, but contingent on society’s needs.  Twenty-two participants (half from Clusters E and D and the rest distributed among all the others) placed the environment as the most important system (of these 22, four placed it on top with society and two with the economy). Many in Clusters E and D visualized society and the economy as subsystems of the environment. RM10E explained his worldview: “If you don’t have a healthy environment, this [economy] eventually is going to collapse; it will collapse.” Moreover, many participants from Cluster E acknowledged that humans are dependent on other organisms for survival and that we are part of the natural cycle; thus, human life would not be possible without the environment. RM7E explained his perspective: “[…] Earth is called Mother Earth. Mother means somebody that provides, gives you nutrition, keeps you alive.” RM10E also talked about the importance of other species for human survival, while UF7E described the place of humankind in nature: “Humans are, you know, we are nature. We can’t start thinking we are superior.”  Few participants from Clusters A and B put the environment above the other two systems. Interestingly, UF9B was surprised by her own answer: “Yeah, interesting for me too because I would never thought that I would put [the environment above], really, I mean, I’m environmentally conscious, but I’m not, I’m not like over the top [...].” All systems were equally important for 13 participants (mainly from Clusters B and C) and it was often highlighted that achieving a balance between these is fundamental.  2.3.5 Views on sustainability Participants defined sustainability in multiple ways (see Table 2.8). Most definitions were shared among most clusters, except one that emerged only in Cluster E; togetherness, which was related to living in harmony between people and with other species. UF4E explained that sustainability means thinking about the survival of everyone (including   64  non-humans) and treating natural resources with authentic care. Similarly, RF5E explained her view: “Finding ways to live at peace with everything... the Earth, other people.”  Table 2.8 Definitions of sustainability provided in each cluster.  Definitions of sustainability Cluster  Resource use rate Balance between systems  Reducing impacts on environment Continuity Togetherness Economic definition Conserving  the environment Brundtland report definition Othera A (n=11) 1 3 4 -- -- 1 1 -- 1 B (n=17) 6 1 4 2 -- 3 1 -- 2 C (n=7) 1 2 1 -- -- 1 1 1 2 D (n=8) 1 3 -- 3 -- -- -- 1 2 E (n=9) 2 2 -- 2 5 -- 1 1 1 Total 11 11 9 7 5 5 4 3 9 Note: Each cell indicates the number of participants that provided that definition (some mentioned more than one definition). a This category included concepts like: self-sufficiency, renewable resources, etc.  Cluster A had the largest proportion of participants who believed that society is currently sustainable or very close to being so, with not one participant acknowledging that society is not sustainable. Figure 2.6 summarizes views regarding sustainability across groups. The belief that industrialized countries are reducing their impacts on the environment and even have lower footprints than less industrialized nations was very prevalent in this group.40 Some participants argued that richer countries can afford to protect their environment, whereas countries with lower incomes cannot. UM12A explained that wealthier countries have a lower impact due to lower rates of population growth: “So, the wealthier we get, the fewer children we have. So over time, you know, if everybody gets wealthy, we solve these inequality problems. […]. Maybe in 150, years we’ll have it solved.”                                                   40 This idea was not evidenced in any of the other clusters, except for two participants in Cluster C.   65                                                                                 Figure 2.6 Participants’ opinions regarding the current state of sustainability, classified by cluster.   66  The proportion of participants that believed that society is sustainable decreased in Clusters B and C and was reduced to none in Clusters D and E. The opposite trend emerged for the belief that society is unsustainable, which was high in Cluster E and decreased towards Cluster A. Issues like the slow development of renewable energy, high resource consumption and increased financial debt were mentioned in Clusters B and C as examples of unsustainability, while in Cluster E, all participants talked about large ecological impacts such as climate change, ocean acidification, biodiversity loss, resource depletion and pollution in general. With concern, RF5E questioned: “How much more can the Earth take?” Many in Cluster E believed that humanity is way off target and possibly heading to collapse. RM10E described his fears: “When is it going to happen? You know, in a way, it’s good we don’t have a crystal ball because I don’t really want to know.”  2.3.6 Reactions to ecological economics Of all participants, 60% mentioned that the ideas of ecological economics were sound and logical, in the sense that society and the economy will not be able to persevere without a healthy environment. In addition, a few participants in Clusters B and C stated that a steady state economy seems less risky and with a longer term view than the current economic model based primarily on exponential growth. A few participants in Cluster D thought of the steady state economy as an inevitable path for economies, due to natural resource constraints. The following quotes reveal some positive initial reactions: “I think the idea definitely makes sense. I think a lot of people are perhaps maybe recognizing that this is sort of the way to move forward” (UF16B). “First thing that came to my mind was, Oh yeah! Somebody, somebody is thinking the right way. I absolutely agree. Like, you cannot, your society and your economy cannot exceed your, the natural resources of the world” (UF2C).  “Ahh... well this definitely makes sense. As I was saying, from a thermodynamic point of view [...], you can’t just keep growing at this sort of rapid rate” (UM5D).   67  “I don’t understand really how people can think other than like this […]. We need to see the world as a whole and balance” (RF5E). Despite the positive reactions regarding the basic premises of ecological economics, many doubts and concerns also emerged. These are summarized in Table 2.9. Participants in Cluster A were the most unenthusiastic, as some believed that the ideas of ecological economics were contrary to human nature. UM12A explained his view: “People have an innate desire to produce and to achieve. You can’t turn that off.” Some participants mentioned that the ideas underestimate human ingenuity and technological improvements. UM4A explained: “I would look at that [ecological economics], saying that the people who think that don’t appreciate human intelligence, because I think that, given challenges, people will rise to those challenges.” He emphasized his faith in human ingenuity to continue modifying and transforming the environment for society’s benefit. Some participants in Cluster A were concerned about decision-making in a steady state economy, especially with respect to resource distribution and accountability. A few were particularly concerned about the loss of individual freedoms and other unintended consequences. Others were concerned about the availability of funding for social services and environmental protection with no economic growth. Finally, some people associated the idea of zero growth with moving backwards, going back to the middle ages or with ceasing all activities. This is partly revealed in statements by UM13A: “Well, the economists will say that we need growth to sustain; if you are not growing, you are going backwards”; and UM12A: “We have to acknowledge and support, in fact, the human desire for progress. You know, during the middle ages when people lived in sod huts and lived to the age of 26, I don’t think they were happy. There was zero growth, zero growth, but they weren’t very happy, I don’t think that was a good model.”      68  Table 2.9 Main criticisms and limitations identified by participants in relation to the concepts of ecological economics.   Criticisms about the ideas Limitations related to human nature Limitations related to implementation Cluster Ingenuity not accounted for Sounds socialistic Possible loss of freedoms Unplanned consequences Moving backwards Greed Materialism Status quo bias Reaction Fear Selfishness Loss aversion Tough sell Politically unfeasible Control by elites Population growth Employment generation Lack of financial resources Decision-making A (n=11) 4 -- 2 1 3 2 1 1 -- -- 1 1 1 1 1 1 -- 2 3 B (n=17) 1 4 2 2 1 3 5 2 4 1 4 2 7 8 1 4 1 1 -- C (n=7) -- -- -- -- -- 4 1 1 1 1 -- -- 2 -- 1 -- 1 -- -- D (n=8) -- -- -- -- 1 1 3 3 -- 1 -- 2 3 2 1 1 3 -- -- E (n=9) -- 1 -- -- -- 2 1 4 1 2 -- -- 2 2 3 1 -- 1 -- Total 5 5 4 3 5 12 11 11 6 5 5 5 15 13 7 7 5 3 3 Note: Each cell indicates the number of participants that mentioned an issue. Some participants mentioned multiple issues. A larger set of negative issues were mentioned by participants in Cluster B. Materialism, greed, selfishness and our incapacity to react prior to crises, were seen as the main obstacles. Political aspects were also seen as a key impediment, especially short-term political cycles and government’s financial resources dependent on growth. For instance, RM5B thought that any politician promoting less consumption “is going to be toasted”. In this sense, the ideas were seen as difficult to communicate and persuade. UM2B explained his view: “If you tell people, ‘we want to stop economic growth’, that’s like not a very popular idea upon hearing it, because you are like ‘economic growth is a good thing’.” Interestingly, UF8B reflected on her own reaction: “I would be scared because, like I said, I want to be successful, I want a good career, […] and if the economy is, you know, like at a standstill, if it is not demanding as much as it used to, then I might have to look at other countries for a job. […].” Two participants pointed out the difficulties of moving to a steady state economy unilaterally by any one country, as it will likely require global cooperation. For four participants, the ideas of ecological economics brought up   69  notions related to communism or socialism, with one thinking that it is a step backwards. He explained: “You can’t go backwards, you know, you can only go ahead; otherwise, if you go backwards, you are collapsing. […] We have to grow” (RM12B). In Cluster C, greed was identified as the major barrier to implementing ecological economics’ proposals, while in Cluster D, materialism, loss aversion (i.e. living with less) and status quo bias (e.g. habits, fear of change) were seen as the most important impediments. UF13D explained her concern: “[…] if you tell us that we have to go with what we had a few decades ago, we are going to think that we are going back and we are not moving forward. It’s a bit scary.” A few participants identified challenges in the communication of these ideas. UM5D saw it as especially challenging, mainly because growth is often equated with progress. In addition, three individuals were concerned about job creation in a steady state economy. Similar challenges were identified by participants in Cluster E, such as status quo bias, greed and control by elites. As in previous clusters, the difficulties in transmitting the ideas of ecological economics were also discussed, especially due to the potential impacts in reducing consumption and employment.  An important point mentioned by six participants related to re-defining ‘growth’ so that its meaning goes beyond material aspects to also include spirituality, creativity, health, knowledge, social connections, time for community and family, cultural advancement, well-being and happiness. In this sense, a few participants suggested that ecological economics’ messages should focus on positive aspects of transitioning to a steady state economy (e.g. increasing ecological health, greater social well-being) rather than focusing on aspects that will decrease (e.g. economic growth, consumption). For example, in reference to Appendix B.3, RM8D explained that a non-increasing steady line visually does not motivate people as it does not reflect advances in standard of living. “[...] people don’t want to see a graph that goes like this [steady]. People are so visual, they are going to say, this doesn’t increase? How is my satisfaction going to increase in this graph? So there’s a challenge there right away.” Interestingly, UF18C had a very   70  positive reaction towards the term ‘sustainable degrowth’; she associated it with having more leisure time and better relationships.  2.3.7 Clusters and cultural cognition of risk  The results for the cultural cognition survey for all participants are shown in Figure 2.7. except for three individuals who did not complete the survey (the reasoning for this is explained in the methodology section in this chapter). Most participants tended to have a more egalitarian perspective than a hierarchical one. Nine participants ranked the lowest score possible in this scale (1), which means that they have a highly egalitarian view. On the other hand, the highest score obtained by a participant was 4.67, indicating a more hierarchical worldview (although no participant scored the highest possible value of 6). With respect to group worldview, more participants have a communitarian worldview, although many tend to be in the middle of the scale (between scores of 3 and 4). No participants were located in the extremes. The lowest score obtained was 1.83 indicating a more individualistic worldview, whereas the highest score obtained (by two participants) was a 5 indicating a more communitarian perspective. When results were analyzed by gender (see Figure 2.8), all women were classified as egalitarians (i.e. not one is in the hierarchical zone), while the only individuals classified as hierarchical, were men. Mann-Whitney U Test (non-parametric) confirm significant gender differences (p value = 0.013) as men reported to be more hierarchical than women were. When data were analyzed by place of residence (see Figure 2.9), results were not statistically significant between MV and BCI residents.        71                      Figure 2.7 Cultural cognition of risk results for all respondents. Each point represents one participant.    Figure 2.8 Cultural cognition of risk classified by gender. Each point represents one participant.     72   Figure 2.9 Cultural cognition of risk classified by place of residence (i.e. metro Vancouver or the BC Interior). Each point represents one participant. Although no statistically significant differences were found between clusters, Cluster A was the only one with participants in the individualistic hierarchical space and Clusters A and B were the only ones with participants in the hierarchical spectrum (see Figure 2.10). All participants in Clusters C, D and E were identified with an egalitarian worldview. Respondents in Cluster C were largely in the ‘communitarian egalitarian’ quadrant, while roughly half of participants in Cluster D were geared towards individualism and the other half were geared towards communitarianism. In Cluster E, most were identified with a communitarian worldview and only two with an individualistic one.     73         Figure 2.10 Cultural cognition of risk classified by clusters. Each point represents one participant. 2.4 Discussion This exploratory research aimed to identify and describe some of the mental models held by residents of British Columbia regarding the interactions between economic growth, society and the environment. In addition, it set out to explore whether these mental models are related to cultural cognition of risk, political views and sociodemographic factors. This section summarizes the main results that emerged, their theoretical and practical significance and how they complement the general literature on worldviews and mental models about the economy and the environment. In addition, study limitations and suggested avenues of future research are discussed.    74  This study identified six clusters of participants, each representing a different mental model, of which five were characterized and deconstructed. Table 2.10 depicts in a simplified way, the broad patterns that emerged in each cluster. Although there were some shared perceptions among groups – for instance, many participants thought of the environment as the most (or one of the most) important systems in relation to the economy and society – some themes pointed to important differences in how people see and understand the world. Specifically, what marked the most important qualitative differences between clusters were participant’s views on limits to economic growth and the reasoning provided for this. Members of Cluster A were the only ones that strongly believed in the possibility of indefinite growth, while members of Cluster B were the only ones that were largely ambivalent about the existence of limits to growth; no other participants (in any of the other clusters) showed this strong level of belief or ambivalence. This meant that all members of Clusters C, D, and E believed, to one degree or another, in limits to growth, but with different reasons for each group. In Cluster C, the unnatural character of indefinite growth and large financial debts were some of the most prominent reasons provided. Resource limitations did not emerge naturally as reasoning in this group, indicating that this issue was not central to their mental model. In contrast, for members of Cluster D, the main reasoning revolved around natural resources, meaning that resource scarcity is central to their way of thinking about this topic. Participants in Cluster E went beyond resource scarcity to mention the general decline in ecosystem services and other ecological issues, indicating a more complex and unpredictable view of nature. Moreover, they gave clear indications of being more eco-centric than participants in any of the other groups. For instance, their conception of sustainability is based on harmonious relationships between humans and other species. Furthermore, members of this group believe that humans are part of the natural cycle, not above it. No other groups naturally and effortlessly expressed such strong eco-centric conceptions, which makes this group unique. Although this research chapter did not determine if the main variables or themes in this study represent one or multiple dimensions, some themes seem to be highly correlated. For instance, views about limits to growth and perceptions on society and human nature   75  seem to be strongly connected; that is, participants that are very optimistic about human nature especially in terms of our ingenuity are less likely to believe in limits to growth. Alternatively, it could also be stated that, participants that believe in limits to growth tend to have less optimistic views of people. Similarly, views on these themes may be also related to opinions on the current state of sustainability. That is, people that believe in limits to growth tend to recognize societal unsustainability more acutely than those who believe in indefinite growth or are ambivalent about this. That said, further study is needed to determine the genuine existence and strength of these correlations.  Table 2.10 Broad representation of general patterns of thought in each cluster, classified by the main themes of analysis.   A Expansionist B Hesitant C Indifferent D Green E Ecological Initial perceptions on growth      Views about limits to growth      Perceptions on people and society      Importance of eco, soc, and enva      Views on sustainability       Reactions to ecological economics      a Only the most mentioned option is represented in this table. By no means has it reflected the thinking of all members in the clusters.  It can be said that all mental models are located within a spectrum of views that range from an expansionist to an ecological worldview (Rees, 1995). Figure 2.11 depicts the classification of the clusters according to this spectrum. The views in Cluster A closely resemble the expansionist perspective, manifested in the optimism towards economic growth, the belief in the possibility of unlimited economic expansion and the faith in human ingenuity and technology. On the other side of the spectrum, the views expressed by participants in Cluster E align well with the ecological worldview in that the economy and society were seen as subsystems of the environment, limits to growth were clearly   More positive                           More negative   More optimistic                            Less optimistic We are closer to sustainability                        We are further from sustainability Less open to ideas   No limits Limits Unsure   Economy Economy, society and environment Environment More open to ideas   76  acknowledged and there was great skepticism about technology’s capacity to replace complex ecological systems.     Figure 2.11 Clusters classified in the expansionist versus ecological spectrum (based on Rees 1995). For more than 40 years, there has been a call for moving away from an economic worldview that sees the Earth as open and unbounded, towards a model that acknowledges limited resources and recognizes humankind as part of the ecological system (Boulding, 1966). Daly and Farley (2011) explain that the expansionist worldview emerged at a time when the world was ‘empty’ of human artifacts and ‘full’ of ecological goods and services. In that context, it made some sense to aim at limitless economic expansion. However, this context has changed as today we live in a rather ‘full’ world of manufactured capital and increasingly scarce ecological services. The results of this research indicate that the expansionist worldview described by Rees (1995) is still very much present in society today. Nonetheless, results also show that most respondents did grasp, albeit to different degrees with the idea of limits, possibly indicating that perceptions among the general population may be moving away from expansionism. However, the data obtained in this study does not provide conclusive support of this trend. Future research should explore and attempt to quantify the distribution of these worldviews in the general population.  The findings in this study reconcile well with results from the Dominant Social Paradigm (DSP). Perspectives expressed in Cluster A and, to a lesser extent, in Cluster B align with the DSP as it pertains to trust in technology, disregard or ambiguity about economic limits and highly anthropocentric views. On the other hand, opinions in Cluster E side very closely with those of the New Ecological Paradigm (NEP). Specifically, participants CLUSTER A The Expansionist CLUSTER B The  Hesitant CLUSTER C The Indifferent CLUSTER D The          Green CLUSTER E The Ecological EXPANSIONIST WORLDVIEW ECOLOGICAL WORLDVIEW   77  in Cluster E and people with high NEP scores, both reject human exemptionalism, recognize the possibility of an ecological crisis and have anti-anthropocentric views. Perspectives in Cluster D also align with the NEP, although less so than Cluster E. These results may indicate that the NEP scale may be an appropriate method for identifying the type of mental models described here, as it claims to address fundamental beliefs about human-nature relationships (Dunlap et al., 2000) and has been used in multiple studies across the world (Hawcroft & Milfont, 2010). Having said that, it should be noted that the NEP scale does not seem to directly address perceptions about the economy. For example, views on limits to growth (one of NEP’s five components), is focused on population growth and resource availability,41 rather than on actually delving directly into people’s views about limits to economic growth (as measured in GDP). Given that the NEP scale is one of the most widely used tools for exploring environmental worldviews (Hedlund-de Witt, 2012), future research could explore its correlation with the mental models identified here and its ability to partly indicate economic worldviews.  Parallels can also be drawn between the results of this study and other research, particularly on climate change. For example, the segments identified in Global Warming’s Six Americas – the Dismissive, Doubtful, Disengaged, Cautious, Concerned and Alarmed (Maibach, Leiserowitz, Roser-Renouf, & Mertz, 2011) can be compared to the clusters identified here, despite the differences in topics and methodologies. For instance, the Dismissive are similar to Cluster A in that they have strong opinions against environmental action, do not believe that human impacts are very large and are more optimistic about technology. The Doubtful and Cluster B are quite hesitant about their views and do not seem largely concerned. On the other side of the spectrum, the Alarmed and Cluster E are the most worried about the urgency of climate change and/or environmental issues and show less optimism about technology (and possibly, human ingenuity). The Concerned and Cluster D are still engaged, although less than the Alarmed and Cluster E. The Cautious and Cluster C are even less certain and do not show                                                  41 NEP items 1, 6 and 11 aim to tackle views on limits to growth and are phrased as follows: 1) We are approaching the limit of the number of people the earth can support; 6) The earth has plenty of natural resources if we just learn how to develop them; and 11) The earth is like a spaceship with very limited room and resources (Dunlap et al., 2000).    78  any sense of urgency. Similar segments to those of Maibach et al. (2011) were identified by Hine et al. (2013) in Australia, except that the Disengaged and Cautious were merged into one segment called the Uncertain. A few similarities can also be drawn from Poortinga and Darnton’s (2016) audience segmentation study in Wales, with the Pragmatists closely resembling Clusters E and D due to their prioritization of environmental sustainability over economic growth, and the Commentators and Community-focused seeming to be closer to Clusters A and B due to the larger priority granted to the economy. All of these results indicate that the mental models identified by this research, although focused more closely on economic issues, are indeed similar to segments emerging from other environmental and sustainability studies. Beyond this validation, it is difficult to establish further links with these studies due to the different focus, variables and methodologies used in these studies. Previous research on cultural cognition of risk shows that more individualistic and hierarchical people tend to be less concerned about environmental risks, while egalitarians and communitarians are generally more supportive of environmental regulation (Kahan & Braman, 2006; Kahan et al., 2007). However, this study did not find evidence of a significant correlation between the mental models and cultural cognition. Interestingly, the results here do point to some relationship between gender and cultural cognition, as only men were classified as hierarchical. This supports research that shows that white men hold more hierarchical views than women and minorities (Finucane, Slovic, Mertz, Flynn, & Satterfield, 2000). Nonetheless, it is important to consider some of the criticisms that have been made about the cultural cognition of risk theory and its operationalization of ‘culture’ and ‘cognition’ (van der Linden, 2015). Political views seem to be associated with particular mental models. All participants that self-identified as Conservatives were classified in Cluster A and, to some extent, in Cluster B, while participants that self-identified as Greens were mainly in Cluster E, and to a lesser extent, Cluster D. These results are consistent with previous research that shows that conservatives tend to have lower levels of environmental concern (Shafer, 2006) and, thus, may be more expansionist in their way of thinking about the economy.   79  However, more research is needed to determine whether there is any real statistical correlation underlying this connection. Gender also seems to be associated with particular mental models, as Cluster A comprised a disproportionate amount of men (ratio: 10 to 1), supporting research that women tend to report higher levels of environmental concern and stronger pro-environmental attitudes and behaviours than men (Hunter, Hatch, & Johnson, 2004; Zelezny et al., 2000). Again, further research is needed to determine whether there is a genuine correlation here.  Successful communications and public engagement require an understanding of the mental models held by the audience (Shome & Marx, 2009). “[...] effective communication and education require an understanding of people’s existing cognitive maps so that information may be framed in a way that encourages people to notice and integrate the new information rather than ignore or reinterpret it” (Kearney & Kaplan, 1997, p. 581). The results presented in this study can be used to improve the rhetoric around ecological economics by providing data about, and accounting for, different types audiences that exist in the population (Luks, 1998). While all groups identified similar strengths and challenges regarding some tenets of ecological economics, participants in Cluster A were the least open towards these; it was thought that they underestimate human ingenuity and technological improvement. However, most of the participants in the other clusters identified challenges not with the ideas of ecological economics per se, but rather with the difficulties of their implementation given some unfavourable characteristics of current society and/or human nature (e.g. materialism, greed, selfishness). Interestingly, the basic principles of ecological economics made sense, were logical and even became evident for many people after they were exposed to them. Thus, despite the challenges raised, these fundamental principles could be a good starting point for discussions regarding a new economic paradigm. As pointed out by Meadows (2006, p. 236) novel information can be key for systemic transformation: “That does not necessarily mean more information […]. It means relevant, compelling, select, powerful, timely, accurate information flowing in new ways to   80  new recipients, carrying new content, suggesting new rules and goals […].When its information flows are changed, any system will behave differently.” For instance, making apparent the connections and dependence of the economy on the environment could be an important step in moving the message of ecological economics forward. It is also possible that more people would be encouraged to believe that such a change in economic paradigms is feasible, by showing progress that has been made thus far, rather than by focusing on the immensity of the challenges lying ahead. In this sense, Moore Lappe (2013, p. 89) claims that we should move past the mental trap of thinking that “we must overcome human nature to save the planet” as it ends up hindering change. She calls for putting greater emphasis on positive human traits, such as our capacity for cooperation, empathy, fairness, etc. On the other hand, data from this study suggests that messages focused on environmental limits will likely only appeal to people who believe in them (i.e. members of Clusters C, D and E), while they will likely backfire with people that believe in indefinite growth (i.e. members of Cluster A). Therefore, further research is needed to explore which types of messages work best with which types of audiences. Moreover, it is likely that regardless of message, people who are closer to the ecological end of the spectrum would be more open to ecological economics proposals than people closer to the expansionist side.  This study confirms that a significant challenge for post-growth communications is that some individuals relate economic growth to prosperity, advancement and moving forward. Growth is an ontological metaphor, which means that many mental associations are generated automatically and largely unconsciously with the use of this term. It is a deep frame in many people’s mind, as it is often seen as natural and inherently positive; therefore, the idea of no growth comes as unnatural and backward (Gustafsson, 2013). “The idea of continuous growth thus seems self-evident and natural. This can be seen as a hegemonic construction of meaning - and it means that critics of the thought of growth have a linguistic challenge to meet" (Gustafsson, 2013, p. 213). These associations are often reinforced in the media, as economic growth is frequently covered as something good and positive (Dryzek, 2013). As argued by Gustafsson (2013), there could be some   81  hypocognition42 at play in relation to economic progress, as there is a lack of well-established ideas and mental frames that could help people to think in other ways about the economy. Moreover, the use of terms like degrowth or postgrowth to promote a new economic model could even be counterproductive, because by negating the frame they reject, they are likely still invoking that frame (Lakoff, 2004)43 and, furthermore, they are negating a “profoundly positive concept” (Gustafsson, 2013, p. 202). Therefore, focusing on positive frames like well-being and quality of life may be more appropriate and effective (Lakoff, 2010). Similarly, Moore Lappe (2013) pleads moving past the growth-no growth’ debate, to focus on aspects such as “flourishing” and “genuine progress”. Education and communications around post-growth economics may face significant challenges in transmitting ideas that counter the popular notion of economic growth. Hence, there is a pressing need for research that explores pathways in relation to this topic.  Lastly, it should be noted that the results of this research are purely exploratory and cannot be generalized. Future research should explore the reliability and prevalence of the mental models identified here and expand to other populations, possibly incorporating quantitative research strategies. The results generated by this investigation are a product of a specific context and methodology, thus the importance of replication and triangulation with other methodologies to corroborate the consistency of these findings.                                                   42 Hypocognition has been defined as the lack of ideas or frames that are needed (Lakoff, 2010). 43 There is a growing debate in academic and social media circles as to whether ‘degrowth’ and ‘postgrowth’ are adequate and effective frames. For examples of these, see:  Raworth, K. (December 1, 2015) Why Degrowth has outgrown its own name, Oxfam blogs. Retrieved on April 25, 2016 from: http://oxfamblogs.org/fp2p/why-degrowth-has-out-grown-its-own-name-guest-post-by-kate-raworth/ Kallis, G. (December 2, 2015) You’re wrong Kate. Degrowth is a compelling word, Oxfam blogs. Retrieved on April 25, 2016 from: https://oxfamblogs.org/fp2p/youre-wrong-kate-degrowth-is-a-compelling-word/ Dean, B. (January 15, 2015) “Degrowth” – a problematic economic frame, NewsFrames. Retrieved on April 25, 2016 from: https://newsframes.wordpress.com/2015/01/15/degrowth/  Dean, B. (August 27, 2014) The economic “growth” frame – and its opposition, NewsFrames. Retrieved on April 25, 2016 from: http://newsframes.wordpress.com/2014/08/27/economic-growth/    82  2.5 Conclusions This research identified six clusters (and described five) that possess distinct mental models pertaining the relationship between the economy, society and the environment. Views on limits to economic growth and the corresponding reasoning provided is what delineated the greatest differences between clusters. In Cluster A, the economy was seen as unrestrained from natural limits and there was great faith in human ingenuity to overcome any constraints to economic growth. In Cluster B, participants were ambivalent about the possibilities of indefinite economic growth. In Clusters C, D, and E, there was a clear recognition of limits to growth, but for different reasons. In Cluster E, large ecological imbalances were recognized, while in Cluster D responses focused on natural resource limitations. In Cluster C, the reasoning that emerged often did not relate to environmental issues. Furthermore, Cluster E was particularly different from the other groups in its clear eco-centric approach towards human-nature relationships. It could be said that these mental models are located in a spectrum of views that range from an expansionist to an ecological worldview, with Cluster A closer to the expansionist side and Cluster E closer to the ecological one. Results indicate that gender and political affiliation could be related to the mental models, as Cluster A was largely composed by males and by most self-identified Conservatives in the sample. In terms of accepting some of the ideas of ecological economics, participants in Cluster A were the least open and accepting, making more sense to participants in other clusters. By deconstructing and bringing to light some of the various mental models held by people about the relationship between economic growth, society and the environment, this research aims to contribute to increasing our knowledge about economic worldviews, their structure and their relation to other variables. Ultimately, it is hoped that this work will contribute indirectly to the transition towards a more sustainable economy and society.      83  3. Identifying Like-Minded Audiences for the Communication of Ecological Economics in the Canadian Population This chapter examines the prevalence of expansionist and ecological attitudes among the Canadian population and segments the audience based on these attitudes. It also explores the relationship between each identified segment and sociodemographic factors as well as participants’ levels of issue involvement. The findings described in Chapter 2 are employed here to inform the development of questions to measure expansionist and ecological attitudes and test whether the clusters identified in Chapter 2 are replicated in a larger sample. This chapter is structured as follows: 1) the introduction describes the current dominant worldviews and highlights the need to explore, in more depth, people’s attitudes surrounding the economy; 2) the methodology presents the data collection and analysis approach; 3) the results present the main findings related to the most prominent attitudes and describes each of the segments identified; 4) the discussion examines the main implications of the findings and their connections with other research; and, 5) the conclusions summarize the main findings and key points of this research.  3.1 Introduction   Human society is facing an unparalleled set of ecological and social challenges. These include, among others, climate change, ocean acidification, loss of biodiversity and large economic inequality (Raworth, 2012; Rockstrom et al., 2009). Although the causes for these crises are multiple and very complex, the dominant economic model has been pointed out as a fundamental driver for many of these issues (Daly & Farley, 2011; Martínez-Alier et al., 2010; Donella Meadows et al., 1992) because, among other aspects, it is based on the unsustainable consumption of resources and depends on continuous growth. However, some authors have gone even further to argue that the primary driver lies in dominant western worldviews that underlie the current economic model (Dunlap & Van Liere, 1984; Gladwin et al., 1997; Rees, 1995) which are rapidly being exported to all parts of the globe. In this sense, the importance of ‘unlearning’ and replacing the   84  problematic and unsustainable ways in which we see the world is becoming more pressing (Gladwin et al., 1997; Donella Meadows, 2008; Rees, 2010; Rosner, 1995). A worldview44 is a set of fundamental beliefs that shape and delimit people’s thinking and interpretation of world (Byrch, Kearins, Milne, & Morgan, 2007; Rees, 1995). They are “mental models about the very nature of reality” (Meadows, 1998, p. 8) and are acquired by living in a specific culture or context. Worldviews delimit what is desirable and the goals to be pursued, and give shape to institutions and technologies (Beddoe et al., 2009). People are so familiar with their own worldviews that they are often unconscious that they even have them (Byrch et al., 2007; Rees, 1995). “Therefore, most of us are generally unaware of the subtle ways in which the prevailing paradigm shapes our understanding of, and approach to, societal problems or that there may be more viable alternatives” (Rees, 1995, p. 344). Moreover, worldviews could become maladaptive if the circumstances and the environment surrounding a society change (Beddoe et al., 2009; Dunlap & Van Liere, 1984).  Various authors have tried to characterize western cultures’ predominant ways of seeing the world. For instance, Gladwin et al. (1997) have described the western worldview as biased to disconnection, reductionism, certainty, efficiency, rationalism, anthropocentrism, techno-optimism and individualism, among other traits. Along similar lines, other authors have depicted the dominant social paradigm (DSP) as a worldview that is generally committed to free enterprise, free markets, private property, technological development, indefinite economic growth and limited government involvement (Dunlap & Van Liere, 1984; Kilbourne et al., 2002; Shafer, 2006). These characteristics are so ingrained in the current paradigm, that they are rarely ever questioned (Shafer, 2006) and, while the DSP is not shared by everyone, it certainly guides to a great extent individual and societal behavior (Dunlap & Van Liere, 1984).  Similarly, Rees (1995) has described the western dominant economic worldview as one where humans are seen as the masters of the natural world, nature is perceived as                                                  44 In the literature, the term worldview is often used interchangeably with other terms like paradigms, attitudes and cognitive maps (Byrch et al., 2007; Schultz, Shriver, Tabanico, & Khazian, 2004).   85  predictable and knowable, economic expansion is seen as unrestrained from ecological limits, economic growth and free markets are identified as paramount to achieve social progress and ecological sustainability, and great faith is placed on technology (i.e. techno-optimistic). Rees (1995) has labeled this economic worldview as expansionist and put it in direct opposition to the ecological worldview.45 An ecological worldview is one where humans and their economy are seen as fully dependent on the environment, the behavior of natural systems is recognized as non-linear and unpredictable, GDP growth is seen as an inadequate indicator of social well-being and ecological health, global carrying capacity is seen as finite, biophysical limits to growth are recognized and techno-salvation is seen with great skepticism (Rees, 1995). Although segments of western society may be shifting towards more ecological ways of seeing the world (Dunlap et al., 2000; Shafer, 2006), the expansionist worldview is still very much prevalent and unchanged. Its influence in local and national policies across the world cannot be overstated (Jepson, 2004; Rees & Wackernagel, 2005) and the goal of growth is still prevalent in political discourses and global economic and even environmental agendas (see UNEP, 2011; OECD, 2011). Moreover, Rees has pointed out that, if this expansionist worldview remains unchanged, the risk of global ecological collapse and geo-political instability will likely only increase (Rees, 1995). A fervent debate between worldviews and paradigms arose in the 1970s with the publication of Donella Meadows, Meadows and Randers (1972) ‘The Limits to Growth’ (Dryzek, 2013; Norgard et al., 2010). A major premise of this book is that, if current trends of economic, industrial and population expansion continue unchanged, the world will face overshoot and collapse sometime in the 21st century, mainly due to resource and ecological constraints (Donella Meadows et al., 1992). Millions of copies were published in more than 20 languages (Norgard et al., 2010). Nonetheless, this publication faced fierce criticism from academics, economists and business people, who argued, among other aspects, that limits and collapse can be avoided through ingenuity, technological wizardry and human adaptability. Also, great faith was put in the market mechanism to                                                  45 The expansionist worldview is associated with neoliberal economics, whereas the ecological worldview is related with the heterodox transdiscipline of ecological economics (Rees, 1995).   86  indicate and solve resource scarcity issues (Dryzek, 2013; Norgard et al., 2010). Thus, over the next decades, the central message and warning of ‘The Limits to Growth’ was largely dismissed and ignored.  More than 40 years have passed since the first edition of this publication and the world is in a further state of overshoot and ecological decline (Rockstrom et al., 2009). As a result of ecological deterioration, social inequity, global economic crises and lower rates of global growth, the dominant economic paradigm is being put into question once again in academic, social and political circles (Hopkins, 2008; Jackson, 2011; New Economics Foundation, 2009; Schneider et al., 2010; Victor, 2008). Moreover, few influential mainstream economists and heterodox media are now contesting the desirability and possibilities of continuous  growth (Drews & van den Bergh, 2016; Norgard et al., 2010). Although the criticisms to the current growth paradigm are still quite marginal, they may present a renewed opportunity to re-examine the dominant worldviews guiding society and may open new possibilities of co-existence between people and nature. This may be a prime opportunity for the transdiscipline of ecological economics (as conceptualized by Daly and Farley, 2011) to put to the forefront some of its guiding principles and proposals, such as the need to transition, sooner rather than later, towards a steady state economy.  If there is any possibility of moving to a new economic paradigm that is not based on indefinite material and GDP growth, it is imperative to explore people’s attitudes and beliefs about growth, economic progress and related issues. So far, multiple research gaps still remain. For instance, there is little public opinion research regarding preferred economic end states (e.g. steady state economy versus economic growth) (Leiserowitz, Kates, & Parris, 2006) and debates about economic growth, prosperity and the environment have not yet explored the public opinion dimension (Drews & van den Bergh, 2016). Moreover, little is known about people’s views on economic growth and the predominance of expansionist and ecological thinking among the general population. Although the New Ecological Paradigm (NEP) aims at measuring pro-ecological orientation (Dunlap et al., 2000) and has been used in multiple studies across the world   87  (Hawcroft & Milfont, 2010), the revised NEP scale does not include any item to specifically explore economic attitudes. For instance, views on limits to growth (one of NEP’s five components) is focused on population growth and resource availability,46 rather than specifically delving into people’s views about limits to economic growth.  In this context, the objectives of this study are to explore the prevalence of some expansionist and ecological attitudes47 among the general population in Canada. Moreover, this research segments the audience based on these attitudes and determines how each segment correlates with sociodemographic factors (e.g. age, gender, political views), and participant’s issue involvement. In this investigation, issue involvement refers to a variable that measures the degree to which participants are interested or concerned about environmental and economic issues (Rothman & Salovey, 1997). Specifically, the research questions that this chapter addresses are: 1) What is the prevalence of expansionist and ecological attitudes among the general population in Canada? 2) Can the sampled population be classified into different segments? a) What are the characteristics of each segment? 3) How does each segment correlate with sociodemographic factors and other variables? Research has shown that messages that are tailored to specific audiences are marginally more successful than non-tailored messages (Hine et al., 2014, 2016). Thus, audience segmentation research, such as the project proposed here, can be used to inform more effective communication of some concepts of ecological economics.                                                   46 As mentioned earlier, NEP’s items 1, 6 and 11 aim to tackle views on limits to growth and are phrased as follows: 1) We are approaching the limit of the number of people the earth can support; 6) The earth has plenty of natural resources if we just learn how to develop them; and 11) The earth is like a spaceship with very limited room and resources (Dunlap et al., 2000). It is evident from these questions that they do not directly tackle views on economic limits. 47 This study does not aim to measure worldviews directly as it will not explore the foundational structures underlying beliefs and opinions, but rather it will explore more superficial attitudes. Koltko-Rivera (2004, p. 5) clarifies that: “Not all beliefs are worldview beliefs. Beliefs regarding the underlying nature of reality, ‘proper’ social relations or guidelines for living, or the existence or non-existence of entities are worldview beliefs. Other beliefs are not.”   88  3.2 Methodology 3.2.1 Data collection methods 1,250 Canadian residents participated in an online panel survey in January 2016. Among other reasons, Canada was selected as a case study for this research, as Canadians has shown to have more pro-environmental attitudes than their counterparts in some European countries and the United States (Pyman & Pammett, 2010). Similarly, in the fight against climate change, British Columbia was the first jurisdiction in North America to set a carbon price and currently, many other provinces are following suit. Although Canada still has one of the largest ecological footprints per capita in the planet (Global Footprint Network, 2012), it is one among a small group of nations that are working to institutionalize the ecological footprint as an indicator (Global Footprint Network, 2011). Canada could be in a privileged position in the world to plan for a smoother transition towards creating a sustainable economy, as it has achieved a high standard of living, while still having large amounts of natural capital. This is a special situation, as most industrialized countries place an ecological demand that far exceeds each nation’s biocapacity (Global Footprint Network, 2012).  This survey was implemented by ResearchNow market research agency. ResearchNow randomly contacted a number of potential participants from their Canada-wide pre-enlisted online panel48 and provided them with basic information about the survey (e.g. subject, estimated length of exercise). Interested participants clicked on the survey link, which took them to the questionnaire cover letter and survey. The sample is similar to the Canadian population in terms of sociodemographic factors but, like the majority of online panels, it is a nonprobability sample.49 The sample sociodemographics are summarized in Table 3.1.                                                     48 Survey companies often offer financial incentives to motivate people to register in their online panels (Baker et al., 2010).  49 A nonprobability sample is one where all members in the population do not have equal probability of being selected, as it does not use random selection (Babbie, 2007).   89  Table 3.1 Sociodemographic characteristics of survey respondents.   Sociodemographics n Valid % Gender Female 531 53.4 Male 463 46.6 Political identification Conservative Party 238 24.4 Liberal Party 351 36.0 New Democratic Party (NDP) 147 15.1 Green Party 61 6.3 None 177 18.2 Age Under 25 74 7.6 25-34 222 22.7 35-44 170 17.4 45-54 213 21.7 55-64 195 19.9 65 or Above 105 10.7 Income Under $40,000 243 28.0 $40,000 to 75,000 226 26.1 $75,000 to $100,000 153 17.6 $100,000 to $150,000 158 18.2 $150,000 and over 87 10.0 Education Less than high school 33 3.3 Completed high school 164 16.4 Some college or university 180 18.0 College or technical school certificate 233 23.4 Received a bachelor's degree 269 27.0 Received a graduate university degree 119 11.9 Ethnicity   White 743 75.7 Chinese 56 5.7 South Asian 23 2.3 Black 18 1.8 Indigenous 22 2.2 Other 120 12.2    90  The surveys were programmed in FluidSurvey software. Sections 1 and 2 were the first within a longer survey and will be the focus of this chapter (the other survey sections are described in Chapter 4 of this dissertation). Section 1 consisted of 12 Likert-type statements, shown in Table 3.3, which explored participant’s views on: economic growth and its costs; human ingenuity and technology; limits to growth; humanity’s place within nature; impact of industrialized vs. less industrialized nations; and sustainability. In this survey, economic growth refers to the increase in GDP. This was explained to respondents before introducing the questions (see Appendix F)50. The main purpose of these statements was to segment the audience based on the five mental models identified in Chapter 2, so the wording for half of these (six out of 12 statements) were based on the data obtained in that chapter. However, three statements (3, 8 and 9) were adapted from the anthropocentric and ecocentric scales of Thompson and Barton (1994), two (4 and 5) were obtained from the revised NEP scale (Dunlap et al., 2000) and one (11) was adapted from the DSP scale (Shafer, 2006). Higher agreement with items 1, 2, 3, 4, 5 and 6 indicate a more expansionist worldview, whereas greater agreement with items 7, 8, 9, 10, 11 and 12 point to a more ecological worldview. For minimizing question order effects, items were presented in randomized order for each respondent. Section 2 of the survey measured participant’s level of issue involvement with three questions. The wording of these questions was as follows: 1) Generally speaking, how concerned are you about the state of the natural environment? 2) Generally speaking, how concerned are you about the state of the Canadian economy? 3) How often do you think about how the economy and the environment affect each other? Response options for questions 1 and 2 included: not at all; a little; somewhat; very; and extremely concerned. Response options for question 3 included: never; not very much; a fair amount; and, a great deal. Sociodemographic data was collected in the survey’s last section.                                                   50 The phrase that clarified the meaning of economic growth was as follows: “In this survey, economic growth is understood to mean the annual increase in value of all goods and services produced in an economy (commonly measured as Gross Domestic Product or GDP).”   91  3.2.2 Data analysis Data were exported from FluidSurvey into IBM SPSS Statistics 23 software. In order to maintain data quality, low quality surveys (i.e. the same response in most or all questions) and surveys from participants who filled it in too quickly (i.e. speeders) were removed. A total of 1001 surveys remained and each was granted a unique code (between 1 and 1001). In order to carry out general analyses, data from Section 1 were numerically coded such that strongly disagree=1, moderately disagree=2, neither disagree nor agree=3, moderately agree=4, strongly agree=5 and missing=-99. Similarly, data from Section 2 were numerically coded so that not at all concerned=1, a little concerned=2, somewhat concerned=3, very concerned=4 and extremely concerned=5 and never=1, not very much=2, a fair amount=3 and a great deal=4.  In order to identify the different segments, the 12 statements from Section 1 were initially used as input for Latent Class Analysis (LCA) using Latent Gold 5.1.51 The choice of methodology was informed by Hine et al. (2014) comprehensive review of audience segmentation about climate change.  LCA is a model-based clustering approach that allows for the identification of latent classes based on a set of observed categorical, ordinal or continuous variables (Bartholomew et al., 2008; Nylund, Asparouhov, & Muthén, 2007). “The objective is to categorize people into classes using the observed items and identify items that best distinguish between classes” (Nylund et al., 2007, p. 539). The analysis was run using 50 sets of random starting values52 (Maibach et al., 2011; Nylund et al., 2007). Missing values were excluded from the analysis.                                                   51 Before deciding on using LCA, SPSS’s K-Means and TwoStep clustering techniques were employed. These two methods have been widely used in the social sciences (e.g. marketing research), especially K-Means (Schreiber & Pekarik, 2014). However, they did not work with great success because they are susceptible to data ordering (Norusis, 2009), so results changed every time the analysis was run. Nonetheless, K-Means results were more consistent (between 80% and 92% of participants were consistently assigned to the same group) than those of TwoStep clustering (between 68% and 88% were consistently assigned to the same group). 52 “One potential problem in estimating latent class models is the possibility of obtaining a local maximum solution rather than a globally-based solution: an estimation algorithm may converge on a local maximum solution, which is the best solution in a neighborhood of the parameter space, but not necessarily the best global maximum. As models become more complex this problem increases. To guard against local maximum solutions, the estimation algorithm should be run several times with different parameter start values” (Maibach et al. 2011, p. 8).   92  Models with solutions between 2 and 6 clusters were tested.53 L2 is a measure assessing the quality of the models generated by LCA. It indicates the amount of unexplained association between variables that remains after estimating the model. A lower value of L2 indicates a better fit of the model to the data (Magidson & Vermunt, 2004). When assessing model fit, attention was paid to the following additional indicators: 1) L2 bootstrap p-values,54 as models with p-values > 0.05 are deemed as better models55 (Magidson & Vermunt, 2004, p. 176-178; Vermunt & Magidson, 2005, p. 171); 2) information Criteria (IC) like the Bayesian Information criteria (BIC), Akaike’s Information Criterion (AIC) and the consistent AIC (CAIC),56 with BIC being recommended as the superior measure (Nylund et al., 2007); 3) plots of ICs against the number of clusters in order to determine the point at which the curve starts levelling off (Nylund et al., 2007). Table 3.2 provides goodness of fit statistics for the initial model. Table 3.2 Goodness-of-fit statistics for 2- to 6-cluster solutions using 12 variables (initial model).  Model L2 Bootstrap p-value BIC 2 clusters 18043.47 0.3220 31205.44 3 clusters 17746.55 0.3860 30997.44 4 clusters 17605.19 0.3780 30945.02 5 clusters 17464.19 0.2860 30892.93 6 clusters 17356.41 0.3060 30874.58  Although all models provided an adequate fit after running this initial LCA (indicated by non-significant L2 bootstrap p-values in Table 3.2), the values of L2 could be further reduced and the model could be improved by reducing the number of variables included in the model (Magidson & Vermunt, 2004). Thus, the variables with lower R2 values (less                                                  53 A maximum of 6-cluster solutions were tested, mainly because these were the number of clusters identified in Chapter 2 of this dissertation and similarly, other audience segmentation research in related topics (reviewed in Hine et al. 2014) often uncover between 4 and 6 clusters. 54 The bootstrap p-value is recommended as it relaxes the assumption that the L2 statistic follows a Chi-square distribution (Vermunt & Magidson, 2005).   55 See Magidson and Vermunt (2004, p. 176-178) for information and examples on assessing model fit. 56 Of these, BIC has been suggested as the best indicator (Nylund et al., 2007). Lower values indicate a better model (Vermunt & Magidson, 2005), although when the BIC continuously decreases with additional clusters, it has been recommended that subjective criteria be used in order to select the best model (Nagin 2005 in Hine et al. 2013).   93  than 0.15) were removed from the analysis, leaving a reminder of eight variables that were used as input in the improved model (see Table 3.3). The goodness of fit statistics for the improved model are presented in Table 3.4. These show that the value of L2 was reduced substantially from the initial model, indicating a better fitting model.  Table 3.3 R2 values for each variable in the initial model.  Variables R2 1. Economic growth is largely a good thing 0.1992 2. There are no limits to the capacity of the economy to keep expanding 0.1898 3. Technology will eventually solve our problems with scarce natural resources 0.2061 4. Human ingenuity will ensure that we do not make the Earth unlivable 0.3096 5. The so called 'ecological crisis' facing humankind has been greatly exaggerated 0.1890 6. Economic growth and environmental sustainability are compatible 0.1165 7. The world is currently not environmentally sustainable 0.2115 8. Humans are as much a part of nature as other animals 0.1713 9. Humans depend on nature to survive 0.2332 10. Economic growth will eventually be limited by the availability of natural resources 0.1056 11. The negative consequences of economic growth are greater than its benefits 0.0757 12. Developing countries have a lower impact on the environment than developed nations 0.0026 Note: R2 indicates how much of the variance in each variable is explained by the model. Table 3.4 Goodness-of-fit statistics for 2- to 6-cluster solutions using 8 variables (improved model). Model L2 Bootstrap p-value BIC 2 clusters 7694.98 0.036 20837.94 3 clusters 7458.04 0.114 20662.77 4 clusters 7380.72 0.102 20647.19 5 clusters 7300.01 0.086 20628.24 6 clusters 7250.57 0.082 20640.56  Among the models with a good fit, the 3-cluster model was deemed the most appropriate according to different indicators. Notably, as shown in Figure 3.1, ICs (especially BIC)   94  started to level off after three clusters and even increased after five clusters, indicating that the 4- and 5-cluster solution did not offer a substantial improvement. In addition, the model classification errors augmented with an increased number of clusters and the standard R2 of the whole model did not improve substantially with more clusters.  Lastly, in order to explore if there were significant differences between each cluster’s level of environmental and economic concern, a Kruskal-Wallis non-parametric test was employed comparing the three clusters. A non-parametric test was used as these variables were not normally distributed. Also, Chi-square and Kruskal-Wallis tests were carried out in order to determine any statistical associations between the clusters and sociodemographic characteristics (e.g. gender, age, income).  Figure 3.1 Plot of log-likelihood values (CAIC, BIC and AIC) for 2- to 6-cluster solutions for the improved model.   201002020020300204002050020600207002080020900210002 3 4 5 6Log-Likelihood ValuesCluster solutionsCAICBICAIC  95  3.3 Results Of the total survey participants, 53% reported being very or extremely concerned about the environment, while 59% reported the same about the national economy. Less than 3% reported not being at all concerned about these issues. About 70% of participants reported having thought a fair amount or a great deal about how the economy and the environment affect each other. Figure 3.2, Figure 3.3 and Figure 3.4 show these results.       Figure 3.2 Reported concern about the state of the Canadian economy.   Figure 3.3 Reported concern about the state of the natural environment. Not at all concerned1%A little concerned12%Somewhat concerned27%Very concerned39%Extremely concerned21%Not at all concerned3%A little concerned14%Somewhat concerned30%Very concerned38%Extremely concerned15%  96                                  Figure 3.4 Frequency with which participants think about how the economy and the environment affect each other.  As shown in Table 3.5, a majority of participants (more than 80%) agreed that humans depend on nature for survival and that we are as much a part of it as other animals (items 8 and 9). Similarly, there was high agreement (close to 70%) that economic growth is largely a good thing (item 1); although slightly more than 60% of participants concurred that, eventually, growth will be limited by the availability of resources (item 10). Respondents were particularly ambivalent about the balance of benefits and costs of growth, as reflected in the response pattern for item 11.  3.3.1 Respondent segmentation results The proportion of participants classified into each cluster, based on each participant’s highest posterior probability value57 is detailed in Figure 3.5. Cluster 1 was the largest segment, containing about 41% of participants, followed by Cluster 2 composed of about 36% of participants and, finally, Cluster 3 with about 23%.                                                     57 Posterior probability values indicate each participant’s likelihood of belonging to each cluster. Never2%Not very much27%A fair amount56%A great deal15%  97  Table 3.5 Relative frequencies, means and standard deviations for 12 items for all participants.  Statement % Distribution Mean SD SD MD ND/NA MA SA 1. Economic growth is largely a good thing 1.8 7.0 23.5 49.5 18.3 3.75 0.894 2. There are no limits to the capacity of the economy to keep expanding 14.3 28.7 25.2 24.5 7.3 2.82 1.169 3. Technology will eventually solve our problems with scarce natural resources 14.4 27.1 28.5 23.8 6.3 2.80 1.14 4. Human ingenuity will ensure that we do not make the Earth unlivable 9.4 21.8 27.4 32.1 9.3 3.10 1.132 5. The so called 'ecological crisis' facing humankind has been greatly exaggerated 25.4 31.3 23.1 15.2 5.0 2.43 1.165 6. Economic growth and environmental sustainability are compatible 4.7 16.5 26.5 39.1 13.2 3.40 1.056 7. The world is currently not environmentally sustainable 3.8 12.6 22.7 38.7 22.1 3.63 1.076 8. Humans are as much a part of nature as other animals 3.1 5.1 10.8 34.0 47.0 4.17 1.016 9. Humans depend on nature to survive 3.0 3.3 7.9 29.7 56.2 4.33 0.966 10. Economic growth will eventually be limited by the availability of natural resources 3.4 11.7 21.8 42.9 20.2 3.65 1.035 11. The negative consequences of economic growth are greater than its benefits 7.3 24.4 37.8 23.6 6.9 2.98 1.025 12. Developing countries have a lower impact on the environment than developed nations 20.1 27.7 23.2 21.0 8.1 2.69 1.233   Figure 3.5 Proportion and numbers of participants classified into each cluster. Cluster 141.1%n=392Cluster 236.3%n=347Cluster 322.6%n=216  98  Figure 3.6, Table 3.6 and Table 3.7 provide profile data for each segment. Based on these data, participants in Cluster 1 showed the greatest affinity with expansionist items, as shown in the higher probabilities of agreeing with items 1 to 5 (see Table 3.6) and higher means (see Table 3.7). On average, members of this cluster were significantly more positive towards economic growth than participants in other groups (see Kruskal-Wallis test results for item 1 in Table 3.7). Also, they were more likely to believe in the possibilities of indefinite growth (see higher probabilities of agreement with item 2 in Table 3.6 and significant Kruskal-Wallis test results in Table 3.7). Along similar lines, they were significantly more optimistic regarding human ingenuity and more likely to believe that technology will solve our problems with scarce resources (see higher probabilities of agreement for items 3 and 4 in Table 3.6 and Kruskal-Wallis results in Table 3.7). Participants in this group were more likely to believe that the ecological crisis has been exaggerated (see higher probabilities of agreement than other groups for item 5 in Table 3.6 and Kruskal-Wallis test results in Table 3.7). Interestingly, members of this group also tended to agree with ecological items. For instance, they recognized that the world is currently not sustainable, although to a lesser extent than participants in other groups (see lower mean for item 7 in Table 3.7). Regarding humanity’s place within nature, there was high agreement that humans are as much a part of nature as other animals and that we depend on nature for our survival (see high probabilities of agreement for items 8 and 9 in Table 3.6).  Participants in Cluster 2 were more likely to gravitate to the middle of the scale and did not express strong opinions about any statement (see Figure 3.6). On average, they were slightly optimistic towards economic growth (see agreement probabilities for item 1 in Table 3.6), although they were somewhat likely to disagree with the possibilities of indefinite expansion (see disagreement probabilities for item 2 in Table 3.6). Similarly, they showed skepticism towards ingenuity and technology as reflected in the conditional probabilities for items 3 and 4 in Table 3.6. There was some recognition that the world is currently not sustainable (see agreement probabilities for item 7 in Table 3.6) and they tended to disagree that the crisis has been exaggerated (see disagreement probabilities for item 5 in Table 3.6). The most striking result from this group is that, although many did   99  believe that humans are a part of nature and depend on it for our survival, they did significantly less so than participants in other clusters (see lower means and Kruskal-Wallis test results for items 8 and 9 in Table 3.7).  Of all groups, participants in Cluster 3 were the most likely to report the least agreement with expansionist statements and the greatest agreement with ecological items (see means in Table 3.7). Like Cluster 2, participants in this group were slightly positive towards economic growth, reflected in the agreement probabilities for item 1 in Table 3.6. However, they were significantly more likely than any other group, to disagree with the possibilities of indefinite economic growth (see higher disagreement probabilities for item 2 in Table 3.6 and Kruskal-Wallis test results in Table 3.7). Moreover, they tended to disagree that technology and human ingenuity will solve our problems (see disagreement probabilities for items 3 and 4 in Table 3.6). Correspondingly, they were significantly more likely than participants in other clusters to recognize human unsustainability and the reality of the ecological crisis (see Kruskal-Wallis test results for items 5 and 7 in Table 3.7). As expected, they largely agreed that humans are a part and depend on nature (see high agreement probabilities for items 8 and 9 in Table 3.6 and Kruskal-Wallis test results in Table 3.7). In summary, as shown in Table 3.7, the differences between the three cluster mean ranks were statistically significant for items 2, 5, 8 and 9 on the questions of limits to growth, the ecological crisis and humanity’s place within nature. Clusters 2 and 3 were significantly different from Cluster 1 (but not between each other) on items 1, 3 and 4, as Cluster 1 was significantly more optimistic about growth, technology and ingenuity. Finally, Clusters 1 and 2 were significantly different from Cluster 3 (but not from each other) on item 7 regarding sustainability. R2 values showed that the LCA model used in this analysis had varying success at fitting each statement. The overall R2 value for the model was 0.629758.                                                 58 R2 statistics for the overall model indicate how well the variables can predict cluster membership. Values closer to 1 are preferred as they indicate better predictions (Vermunt & Magidson, 2005). However, due to the exploratory nature of this research, great attention was not put in R2 values.    100   Figure 3.6 Profile plot of means for 3-cluster model. Values closer to 5 indicate more agreement with statement and values closer to 1 indicate more disagreement. To convey differences between clusters more clearly, items that measured expansionist attitudes are presented first (i.e. first five items), while items that measured ecological attitudes (i.e. three last items) are presented second. 12345Economic growth is largely a good thingThere are no limits to the capacity of theeconomy to keep expandingTechnology will eventually solve ourproblems with scarce natural resourcesHuman ingenuity will ensure that we do notmake the Earth unlivableThe so called 'ecological crisis' facinghumankind has been greatly exaggeratedThe world is currently not environmentallysustainableHumans are as much a part of nature as otheranimalsHumans depend on nature to surviveCluster 1 Cluster 2 Cluster 3EXPANSIONIST ITEMS ECOLOGICAL ITEMS   101  Table 3.6 Conditional probabilities for each cluster.  Conditional probabilities Expansionist items CL SD MD ND/NA MA SA 1. Economic growth is largely a good thing  1 0.00 0.01 0.11 0.55 0.33 2 0.03 0.11 0.32 0.46 0.08 3 0.03 0.11 0.32 0.46 0.08 2. There are no limits to the capacity of the economy to keep expanding  1 0.04 0.18 0.27 0.37 0.14 2 0.15 0.34 0.27 0.21 0.04 3 0.32 0.41 0.18 0.08 0.01 3. Technology will eventually solve our problems with scarce natural resources 1 0.04 0.15 0.30 0.38 0.13 2 0.21 0.35 0.28 0.14 0.02 3 0.24 0.36 0.26 0.12 0.01 4. Human ingenuity will ensure that we do not make the Earth unlivable 1 0.01 0.06 0.22 0.52 0.20 2 0.13 0.31 0.32 0.21 0.02 3 0.20 0.36 0.29 0.14 0.01 5. The so called 'ecological crisis' facing humankind has been greatly exaggerated  1 0.11 0.28 0.30 0.23 0.09 2 0.19 0.36 0.26 0.15 0.04 3 0.63 0.30 0.06 0.01 0.00 Ecological items CL SD MD ND/NA MA SA 7. The world is currently not environmentally sustainable  1 0.06 0.17 0.29 0.38 0.10 2 0.04 0.14 0.27 0.42 0.12 3 0.00 0.00 0.03 0.34 0.62 8. Humans are as much a part of nature as other animals 1 0.01 0.02 0.08 0.35 0.53 2 0.07 0.11 0.18 0.37 0.26 3 0.00 0.01 0.04 0.27 0.69 9. Humans depend on nature to survive 1 0.01 0.02 0.06 0.35 0.56 2 0.07 0.08 0.14 0.39 0.32 3 0.00 0.00 0.00 0.05 0.95 Note: SD stands for strongly disagree, MD for moderately disagree, ND/NA for neither disagree nor agree, MA for moderately agree and SA for strongly agree.      102  Table 3.7 Overall R2 and means for all participants, means for each cluster profile and Kruskal-Wallis results for comparisons between cluster mean ranks for each survey statement included in the model. Expansionist items Overall Cluster Means Kruskal-Wallis R2 Mean 1 2 3 1. Economic growth is largely a good thing  0.167 0.89 4.19 3.45 3.44 MC1 significantly differenta than MC2 & MC3 2. There are no limits to the capacity of the economy to keep expanding  0.198 1.17 3.38 2.65 2.05 All significantly differenta 3. Technology will eventually solve our problems with scarce natural resources 0.200 1.14 3.41 2.41 2.31 MC1 significantly differenta than MC2 & MC3 4. Human ingenuity will ensure that we do not make the Earth unlivable 0.312 1.13 3.84 2.68 2.41 MC1 significantly differenta than MC2 & MC3 5. The so called 'ecological crisis' facing humankind has been greatly exaggerated  0.234 1.17 2.92 2.48 1.44 All significantly differenta Ecological items  7. The world is currently not environmentally sustainable  0.232 1.08 3.28 3.43 4.58 MC3 significantly differenta  than MC1 & MC2 8. Humans are as much a part of nature as other animals 0.166 1.02 4.38 3.63 4.63 All significantly differenta 9. Humans depend on nature to survive 0.201 0.97 4.44 3.81 4.95 All significantly differenta Note: Scale coding: Strongly disagree = 1, moderately disagree = 2, neither disagree nor agree = 3, moderately agree = 4, strongly agree = 5 a p < 0.001  3.3.2 Concern with environmental and economic issues  The level with which participants are concerned about a certain issue has been shown to be an important factor in determining how they process information. If participants are more involved with an issue, they often process information in more effortful and conscious ways, whereas, if their involvement is low, they may use more heuristics and peripheral processing (Maheswaran & Meyers-Levy, 1990; Perloff, 2010). In order to explore differences in terms of issue involvement between segments, Kruskal-Wallis non-parametric tests were employed using cluster membership as the independent variable and the three measures of issue involvement as the dependent variables. As expected, environmental concern was statistically higher in Cluster 3 than in the other   103  groups (p = 0.000). Similarly, the frequency with which participants in Cluster 3 reported having thought about the relationship between the economy and the environment was significantly higher than in the other two clusters (p = 0.000). Concern for the Canadian economy was not significantly different between groups. Table 3.8 provides the means and standard deviations for these items. Table 3.8 Concern for the state of the natural environment and the Canadian economy classified by cluster.  Items Cluster 1 Cluster 2 Cluster 3 Scale Points Mean SD Mean  SD Mean  SD How concerned are you about the state of the natural environment? 3.25 1.032 3.39 0.94 4.06 0.803 5 How concerned are you about the state of the Canadian economy? 3.68 1.007 3.62 0.976 3.59 0.942 5 How often do you think about how the economy and the environment affect each other? 2.78 0.692 2.76 0.641 3.02 0.703 4 Note: Scale coding for the first two questions: Not at all concerned = 1, a little concerned = 2, somewhat concerned = 3, very concerned = 4 and extremely concerned = 5. Scale coding for last question: Never = 1, not very much = 2, a fair amount = 3 and a great deal = 4.  3.3.3 Sociodemographic characteristics of segments A sociodemographic characterization of each of the identified clusters was carried out, as sociodemographic factors have shown to be significant predictors of multiple environmental attitudes (Dunlap et al., 2000; Heath & Gifford, 2006; Maibach et al., 2011). Table 3.9 presents sociodemographic characteristics for each segment. Based on Chi Square tests, gender and political affiliation were significantly associated with the clusters (p = < 0.05) (see Table 3.10 and Table 3.11). Specifically, Cluster 3 comprised a larger proportion of women than men, a lower share of Conservatives and a larger share of Greens59 than the other clusters. Liberals were evenly distributed across groups. Based on a Kruskal-Wallis non-parametric test, education and income were significantly different (although marginally) between clusters, while no difference was found for age.                                                  59 Of all participants identified as Conservatives, 49.8% were classified into Cluster 1, while only 12% were assigned to Cluster 3. Of the ones who identified with the Green Party, 25% were assigned to Cluster 1, 26.7% to Cluster 2 and 48.3% were assigned to Cluster 3.   104  Specifically, Cluster 3 had a higher mean rank in educational achievement than Cluster 1 (p = 0.004) and Cluster 2 (p = 0.001), and a higher income mean rank (p = 0.25) than Cluster 2. Figure 3.7 shows the results for the Kruskal-Wallis pairwise comparisons. Table 3.9 Sociodemographic characteristics for each cluster in terms of gender, political affiliation, age, income and education.  Demographics Cluster 1 Cluster 2 Cluster 3 Gender (% male) 50.60% -- 46.10% -- 40.10% -- Political   Conservatives 29.20% -- 25.60% -- 12.90% -- Liberals 35.70% -- 33.30% -- 37.80% -- NDP 13.30% -- 15.20% -- 19.10% -- Green 3.80% -- 4.80% -- 13.90% -- None 18.00% -- 21.10% -- 16.30% --   Mean SD Mean SD Mean  SD Agea 3.54 1.44 3.57 1.53 3.5 1.43 Incomeb 2.57 1.38 2.47 1.28 2.8 1.33 Educationc 3.84 1.38 3.78 1.33 4.2 1.35 a Age: Less than 25 = 1; 25 to 34 = 2; 35 to 44 = 3; 45 to 54 = 4; 55 to 64 = 5; 65 or above = 6 b Income: Under 40,000 = 1; 40,000 to 75,000 = 2; 75,000 to 100,000 = 3; 100,000 to 150,000 = 4; 150,000 and over = 5 c Education: Less than high school = 1; Completed high school = 2; Some college or university = 3; Received a college or technical school certificate = 4; Received a university's bachelor degree = 5; Received a graduate degree = 6  Table 3.10 Chi-square test for gender and cluster. Chi-Square Tests   Value DF Asymptotic Significance                     (2-Sided) Pearson Chi-Square 6.205a 2 0.045 Likelihood Ratio 6.231 2 0.044 Linear-by-Linear Association 6.155 1 0.013 N of Valid Cases 948     a 0 cells (0.0%) have expected count less than 5. The minimum expected count is 98.84.    105  Table 3.11 Chi-square test for political identification and cluster. Chi-Square Tests    Value Df Asymptotic Significance (2-Sided) Pearson Chi-Square 43.961a 8 0.000 Likelihood Ratio 42.225 8 0.000 Linear-by-Linear Association 5.901 1 0.015 N of Valid Cases 929     a 0 cells (0.0%) have expected count less than 5. The minimum expected count is 13.50.                                   Figure 3.7 Kruskal-Wallis pairwise comparisons for income (left figure) and education (right figure). Gray lines show significant differences, while black or no lines indicate non-significant differences.  3.4 Discussion This chapter set out to explore the spectrum of expansionist and ecological attitudes among the Canadian population and to segment the audience based on these attitudes. Three clusters were selected and described. Figure 3.8 illustrates the relative size of each segment and its position in the expansionist-ecological spectrum.    106   Figure 3.8 Some perceptions that emerged in each cluster, represented in the expansionist-ecological spectrum. The size of the circles aims to reflect the relative size of each segment.  Before delving into the differences between clusters, common perceptions among all participants are discussed. Opinions on two themes indicated a relatively high degree of agreement. First, a large proportion of respondents agreed that economic growth is largely a good thing. These results provide further evidence that growth is still linked by many with positive initial associations – even in the greener segment (Cluster 3). This supports existing literature that points to growth being perceived as natural, unquestionable and inherently positive (Gustafsson, 2013). Most governments and capitalist democracies live under the promise of increased growth and it is often assumed in the media that “growth is good” (Dryzek, 2013), as increased economic growth is generally equated with a higher quality of life (Beddoe et al., 2009). Certainly, this will be a challenge for the communication of any economic proposal directly countering indefinite economic expansion.  Strategic messaging may help people to think in different ways about economic prosperity without relying on growth. In this sense, developing new and effective messages may be one of the ways to battle the hypocognition that currently exists in relation to economic progress (Gustafsson, 2013). Hypocognition refers to the lack of ideas, words or mental frames that help people to think in different or novel ways about Higher likelihood to believe that: There are no limits to growth Technology and ingenuity will save the day Ecological crisis has been exaggerated  Higher likelihood to believe that: There are limits to growth Technology and ingenuity may not save the day Ecological crisis has not been exaggerated  C 2          36.3% C 3   22.6% C 1            41.1% EXPANSIONIST ECOLOGICAL   107  an issue (Lakoff, 2010). For instance, Mignolo (2016) points out that when thinking about the economy, concepts like development, growth and progress are so deeply ingrained and linked with one another, that it is challenging and even impossible for many people to think about successful economies without these characteristics. Moreover, development is seen as natural and inevitable, and often, as the ‘only’ option (Mignolo, 2016). That said, it should be noted that most participants in this study were not blind to the costs of economic growth, as some agreed that the negative consequences are large and many were unsure and possibly conflicted, about the balance of benefits and costs.  A second topic that generated high agreement among respondents is the recognition that humans depend on nature to survive and that we are as much a part of nature as other animals (although Cluster 2 had the lowest proportion of participants that strongly agreed with these statements). These findings may indicate that Canadians hold certain ecocentric attitudes in terms of recognizing people as part of the natural cycle. Moreover, all clusters scored above the middle point in the scale (i.e. above 3) in the ecological items, indicating a tendency to agree with these statements. This pattern did not emerge with expansionist items. This could reinforce Dunlap et al.’s (2000) position that new ecological worldviews are indeed emerging, although more longitudinal studies are required to establish any definite conclusions.  This finding may provide some support to other research that indicates the high importance granted to the environment by Canadians. For instance, according to the World Values Survey (2005-2009 wave), 68% of Canadians prioritized environmental protection over economic growth and jobs (World Values Survey, 2015), while another study shows that this percentage rose to 88% in 2012 (Environics Institute, 2012).60 However, our study goes one step beyond by revealing that 52% of participants agreed                                                  60 The question wording varied in the two studies. While the World Values Survey asked participants to prioritize between two mutually exclusive options (i.e. protecting the environment or economic growth and creating jobs), the Environics Institute survey asked participants their level of agreement or disagreement with a 4-point Likert-type scale on the statement: “Canada should be a country where the environment is protected, even if it slows down economic development.” A neutral option (e.g. neither agree nor disagree) was not offered.   108  that economic growth and environmental sustainability are compatible. In other words, when given the opportunity, many respondents do not necessarily see these two goals as incongruous.61 Another study reveals that Canadians are more concerned with environmental issues compared to their counterparts in the United States and Europe, with over 70% reporting being moderately or strongly concerned with the environment (Pyman & Pammett, 2010). Our study supplements these findings by showing that more than 80% of all participants reported being somewhat, very or extremely concerned with environmental issues. However, these proportions increased to 95% when results were analyzed for Cluster 3 only. Pyman and Pammett (2010) also show that Canadians reported the least agreement with the notion that economic growth is needed to protect the environment (40% in comparison to 50% in Europe and 60% in the U.S.) and that modern science will solve environmental problems (15% agreement in Canada compared to about 25% in Europe and the U.S.). Similarly, 70% of all participants in our study were skeptical that technology will alleviate natural resource scarcity. These findings are likely good news for some environmentalists and ecological economists in that there appears to be a significant fraction of the population that already seems to be aligned with some aspects of an ecological worldview and, above all, claim to be highly concerned about the current state of the natural environment. Nonetheless, as mentioned above, more longitudinal studies are indeed required. Despite the commonalities mentioned, there were significant differences between segments. The largest segment composed of 41.1% of participants – Cluster 1 – showed the greatest alignment with expansionist attitudes. Members of this group were the most optimistic about economic growth, technology and human ingenuity. Although it was partly recognized that the world is unsustainable, participants in this cluster were also less likely than those in other clusters to believe in limits to growth. Interestingly, mean scores were above the scale mid-point for the majority of statements (see Figure 3.6)                                                  61 Similarly, based on a survey on public opinion in Spain, Drews and van den Bergh (2016) show that the majority of respondents preferred green growth among other economic options (including ‘growth at all costs’). One reason for this result, as argued by the authors, could be that green growth reduces people’s cognitive dissonance as they are not forced to choose between two positive, but seemingly contradictory, goals – the economy and the environment.   109  indicating a tendency to agree with both expansionist and ecological items. These results may indicate that environmental and sustainability issues are, indeed, becoming more evident to people, even to those with expansionist tendencies. Also, this may reflect a move away from ‘growth at all costs’ and Promethean discourses, which tend to neglect and dismiss all human impacts on nature (Dryzek, 2013). Although no other study (to my knowledge) has attempted an audience segmentation exercise specifically on people’s views of economic growth and the environment, other studies about climate change have reported comparable findings (Hine et al., 2016; Leiserowitz, Maibach, & Roser-Renouf, 2007; Maibach et al., 2011). For instance, participants in Cluster 1 resemble the Dismissive and/or Doubtful groups identified by Maibach et al. (2011) in the U.S. and Hine et al. (2016) in Australia, in that they are more skeptical and less concerned about climate change in those studies and about the ecological crisis in ours. Also, our study supports the finding that these groups (i.e. Cluster 1 and the Dismissive and Doubtful) tend to align more with right wing politics (Leiserowitz et al., 2007). The second largest segment – Cluster 2 – composed by 36.3% of participants, had a tendency to be in the middle of the scale on most issues and did not express any strong beliefs one way or another; not even about humans being a part of and dependent on nature. Participants in this group resemble the Uncomitted identified by Hine et al. (2016) and the Cautious and the Disengaged identified by Maibach et al. (2011) in that they are less convinced and unsure about many issues. It is possible that many members of this cluster had actually put little thought into these topics beforehand and, thus, do not have solid and well-formed opinions. Certainly, this was often revealed during the interviews described in Chapter 2 of this dissertation, mainly from participants classified into Clusters B, C and D. Although environmental topics like climate change are currently much more widely discussed in the media (Schmidt, Ivanova, & Schafer, 2013) and people are generally more aware and cognisant about them, issues related to environmental limits to growth are less prominently discussed in current media, if at all.  If anything, the goal of continuous economic growth has been implicitly and explicitly taken for granted as something desirable and achievable. Therefore, due to the lack of prominence and discussion of these issues, it is likely that many people have not thought   110  about them in much depth, leading to a high proportion of participants in this study who indicated not having a definite or strong position on this. Participants with these ‘ambivalent’ perceptions, may offer an opportunity for strategic post-growth communications, as it is possible that they may be more open to new ideas and thus sway more easily in their views, as they do not have well-established opinions yet. For instance, people with low involvement have shown to be more influenced by gain-framed messages (Nan, 2007).62 Future research could determine the types of messages that could be more appealable and effective with this large segment. Nonetheless, it would be worthwhile to differentiate between participants that were ambivalent about issues because they genuinely do not have well formed opinions versus those that did not understand the statement because they are not familiar with the terminology or concepts used.63  The smallest segment – Cluster 3 – composed of less than one-quarter of participants, leaned towards more ecological attitudes. Participants in this group gave greater recognition to sustainability issues and the ecological crisis, and disagreed that the economy can grow indefinitely. This cluster resembles the Alarmed and the Concerned groups identified by Hine et al. (2016) and Maibach et al. (2011) in their high levels of concern and realization that ecological threats are real, among other aspects. Similar to the findings uncovered by Maibach et al. (2011), members of this group are more likely to be women and to identify with more leftist political parties. Due to their higher level of concern and pro-environmental attitudes, it is likely that members of this group may be quite open and receptive towards some of the ideas and proposals of ecological economics, like transitioning to steady state economies. Nonetheless, future research is needed to confirm this theory.  This study provides further evidence of the relationship between environmental attitudes and sociodemographic factors. Specifically, participants in Cluster 3, who are more                                                  62 The desirability of end states was a crucial factor influencing the persuasiveness of gain or loss framed messages (Nan, 2007). 63 The survey in this study did not include a ‘don’t know’ option. Consequently, it is not possible to differentiate between these two types of participants.   111  aligned with an ecological worldview, were more likely to be women and to identify with the New Democratic (NDP) and Green parties. Correspondingly, participants more closely aligned with an expansionist worldview (as represented in Cluster 1) were more likely to be associated with the Conservative party. Furthermore, members of Cluster 3, had marginally higher levels of formal education that their counterparts in Clusters 1 and 2. Other studies have reported similar findings. For instance, people that score higher on the belief in limits to growth have more leftist political tendencies (Drews & van den Bergh, 2016). Similarly, studies in BC and Canada show that women and people identifying with left-wing political parties show more environmental concern (Blake, Guppy, & Urmetzer, 1997; Pyman & Pammett, 2010). Also, people with higher levels of formal education have also been found to have higher levels of environmental concern (Pyman & Pammett, 2010).  A purpose of this research was to test if the clusters identified in Chapter 2 replicated in a larger sample. Although the six clusters identified in Chapter 2 did not fully replicate, there were some similarities with the segments identified here. Cluster 1 parallels the ‘Expansionist’ (Cluster A) and to some extent the ‘Hesitant’ (Cluster B), in the positive views expressed about economic growth, the faith in technology and ingenuity, and the little recognition granted to human unsustainability and the ecological crisis. Cluster 2 resembles a combination of ‘Indifferent’ (Cluster C) mainly in that participants did not seem to hold strong opinions about the subject matter of this research. Cluster 3 resembles the ‘Green’ (Cluster D) and ‘Ecological’ (Cluster E) in the acknowledgment of human dependence on nature and the recognition of the sustainability crisis, among other aspects. Despite these similarities, there were some differences. For example, participants in Cluster 1 agreed with many ecological items (e.g. humanity’s dependence on nature), making this group less expansionist than the corresponding groups in Chapter 2. Additionally, the proportions of the clusters varied, since fewer people were classified in Cluster 3 (i.e. the ecological segment) in this chapter, compared to the ‘Green’ and ‘Ecological’ segments in Chapter 2. Similarly, a larger share of respondents were classified into Cluster 2 (i.e. in the ‘middle of the scale’) in this chapter, in comparison with the ‘Indifferent’ or ‘Uninformed’ in Chapter 2.      112  Although the results presented in this chapter challenge in some way the existence of the six clusters identified in Chapter 2, it is important to note that the inconsistency may be partly due to the samples not being fully comparable and differences in the methodologies employed. The qualitative methods used in Chapter 2 allowed the researcher to explore, deconstruct and understand participant’s thinking with greater depth and detail; so, respondents that could have been classified as ambivalent or undecided initially, were largely not, as their actual pattern of thinking was clarified through probing and deeper questioning. In addition, most participants devoted about one hour to think and answer questions about these topics, many clarifying their own thinking during the process. On the contrary, a survey methodology like the one employed in this chapter, does not allow exploring participant’s thinking with such depth or confirm if the questions are being understood in the intended way. Moreover, participants devoted less time to answering, and possibly thinking, about these questions. Having said that, a survey methodology is essential for reaching larger samples. This highlights the importance of using multiple methodologies with different approaches and techniques, to tackle research questions. More importantly, it underscores the need for replication to determine the validity and reliability of the results presented here.  The findings presented here represent a new way of identifying and differentiating audiences based on the spectrum of expansionist and ecological tendencies. The questions used in this study may complement traditional questionnaires on environmental attitudes, in that they tackle more directly people’s views about the economy-environment interactions, and more importantly, they delve into participants perceptions about limits to economic growth. This could be an important contribution, as these issues are not necessarily explored in environmental attitudes surveys. However, it is important that in future research, participants are offered a ‘Don’t know’ option in survey questions, so as to differentiate between participants who do not have a solid and well-formed opinion (i.e. those that neither agree nor disagree) from those that genuinely do not know (possibly in part due to a lack of understanding of the terminology or concepts included in the question).   113  Despite the novelty of the findings presented here, these should not be thought of as representative of the Canadian population in a statistical sense, due to the nonprobability sampling strategy used (Baker et al., 2010). In addition, opinions shift with time and circumstances; therefore, it is likely that the compositions of clusters would change if other studies with new methodologies are implemented (see Leiserowitz et al., 2007; Maibach et al., 2011). Moreover, if these issues start being more widely discussed in the media or political circles, the structure of the identified segments could over time be altered, with the emergence of new groups and the disappearance of old ones (Hine et al., 2014).  Future research could improve and add new indicators (e.g. based on policy preferences, behaviours). Moreover, more research is needed to explore and ratify the presence of these segments among the Canadian population using probability sampling strategies. In addition, new studies could identify policy priorities and behavioural preferences of each segment, and delve into how different communication strategies may influence and interact with the dominant mental model held by each cluster. For example, studies could look at the perceptions of each group towards different concepts and messages, and explore the development of message frames that resonate more effectively with each target audience. This could contribute towards improving the effective communications and wide dissemination of some ideas of ecological economics among the general population.  3.5 Conclusions The dominant western worldview has often been called out as one of the main drivers of the current ecological crisis. Understanding the different ways in which people think about economic growth, sustainability and the environment, is crucial for eventually moving society away from unsustainable worldviews. Results indicates that the majority of participants in this research have certain ecological attitudes, mainly when it comes to recognizing that humans are a part of nature and depend on it for our survival. Many recognized that economic growth will eventually be limited by the availability of resources, although a majority still believed that economic growth is largely positive.   114  This research also distinguished three different segments among the sampled population. Participants classified into Cluster 1 (41.1%) expressed a high degree of agreement with all expansionist items. For instance, they were the most positive towards economic growth, technology, human ingenuity and the possibilities of indefinite economic expansion. Participants in Cluster 2 (36.3%) were often in the middle of the spectrum on all items and did not express strong opinions one way or another about any particular issue. Participants in Cluster 3 (22.6%) expressed the least agreement with expansionist items and the highest agreement with ecological ones. Respondents in this group revealed high awareness of human unsustainability, disagreed strongly with the possibilities of indefinite growth and reported significantly higher levels of concern for the environment. Certain sociodemographic characteristics – gender, political identification, education and income – were shown to be associated with particular segments. Specifically, Cluster 3 comprised a larger proportion of women, a lower proportion of Conservatives (and, therefore, a larger share of Greens), and higher levels of education compared to the other groups, and higher levels of income compared to Cluster 2. Further research is needed to determine if the segments identified here replicate. Furthermore, new studies could identify the behavioural and policy preferences of each group and inform the development of effective messages and communication strategies that could aid in transitioning towards an ecological macroeconomy.    115  4. Framing of Ecological Economics This chapter examines how different messages related to transitioning into a new economic paradigm, influence people’s thoughts, emotions and attitudes. In addition, it explores if different audience segments (uncovered in Chapter 3 of this dissertation), participants level of involvement with the subject matter of this research, and sociodemographic make ups, are potential moderating variables of message framing effects.  This chapter is structured as follows: 1) the introduction provides a brief overview of ecological economics and message framing theory; 2) the methodology presents the data collection and analysis approaches; 3) the results section shows framing effects (if any) on cognitive responses, emotional reaction and attitudes, and identifies significant moderating variables and covariates; 4) the discussion examines the main implications of the findings and their connections to other research; and, 5) the conclusions synthesize the key findings of this research.  4.1 Introduction  For more than two decades, the transdiscipline of ecological economics has espoused the need to move into a different economic paradigm, one that is more sustainable and equitable (Daly & Farley, 2011; Røpke, 2004). Currently, a similar call is being echoed by a larger number of people, institutions and governments around the world (Klein, 2014; OECD, 2011; UNEP, 2011). Ecological economics differs from mainstream economics in that it acknowledges the existence of biophysical constraints and recognizes that global carrying capacity is limited (Daly, 1991; Daly & Farley, 2011; Rees, 2011). “Where conventional economics espouses growth forever, ecological economics envisions a steady state economy at optimal scale” (Daly & Farley, 2011, p. 23). To be sustainable, humanity should use the flow of services and resources generated by natural capital (i.e. natural income), but should not consume the capital itself.    116  Central to ecological economics is the recognition of the opportunity costs of economic growth. Ecological economics theory (as presented by Daly and Farley, 2011) acknowledges diminishing marginal returns to growth; that is, further growth in already large economies does not generate large marginal social benefits. Moreover, globally, we may well be in a period of uneconomic growth, where the costs of economic expansion (e.g. social, environmental) exceed the benefits (Daly, 1991, 2013). For instance, while global GDP has risen exponentially, genuine progress may actually be declining (Kubiszewski et al., 2013). This aligns with studies that show that increases in income in industrialized countries are not reflected in large gains in well-being and life satisfaction (Easterlin, 2001; Jackson, 2011; Kahneman, 2011). Therefore, many ecological economists propose putting the focus on improving people’s overall quality of life, rather than on increasing output (Daly & Farley, 2011; Dietz & O’Neill, 2013).  Nonetheless, transitioning into a new economic paradigm is a major undertaking and will not come without significant challenges. Scholars have primarily focused on discussing and designing the theory that would support the macroeconomics, institutions and politics for this new ecological macroeconomy, but empirical studies have received less attention (Berg & Hukkinen, 2011) and the level of public support for such an ambitious endeavour is still very unclear and not well researched (Drews & van den Bergh, 2016). Furthermore, challenging the status quo (of endless economic growth) requires research to identify effective ways in which these concepts can be framed, communicated and accepted by the public. This chapter aims to fill some of these research gaps by exploring ways in which some concepts of ecological economics can be successfully conveyed to the public. Specifically, the main objective of this study is to assess whether the way a message is framed influences people’s thoughts, emotions and attitudes about economic growth and related issues. In addition, it aims to explore if other variables (e.g. audience segments, sociodemographic) are moderators of message framing effects.  4.1.1 Framing theory Framing effects happen when different responses arise and shifts of choice occur, depending on the formulation of an issue. In other words, “[…] the way in which   117  information is presented can influence how people understand, evaluate, and act on a problem or issue” (Nabi, 2003, p. 255). Framing influences people’s interpretation and the underlying considerations when assessing an issue (Chong & Druckman, 2007). Framing research has been implemented in multiple fields (e.g. marketing, health, political communications) and, consequently, has various conceptual and operational definitions (Chong & Druckman, 2007; Weaver, 2007). This investigation will use the concept of frames in communication which “occur when (often small) changes in the presentation of an issue or an event produce (sometimes large) changes of opinion” (Chong & Druckman, 2007, p. 104).  Frames in communication have been classified as equivalency or issue framing (Chong & Druckman, 2007). Equivalency or valence framing is when equivalent or identical options are framed in different ways. For example, a surgery with a 90% survival rate is preferred by more than 80% of respondents; however, when it is framed as a 10% mortality rate, support falls to 50% (note that both options are identical) (Levin et al., 1998; Tversky & Kahneman, 1981). Issue or emphasis framing occurs when qualitatively different aspects or dimensions of the same issue are highlighted or made more salient (Chong & Druckman, 2007; Entman, 1993; Weaver, 2007). By highlighting or omitting certain aspects of a communication, people’s underlying considerations and lines of thought are influenced (Chong & Druckman, 2007; Nisbet, 2009). For example, climate change mitigation could be framed towards its health benefits instead of being framed by its economic impacts.  Although framing research has most often been operationalized as some form of equivalency framing, there is still little consensus as to which type of message frames are the most effective (Levin et al., 1998). In an effort to clarify the literature, Levin et al. (1998) classified equivalence framing as: 1) attribute framing; 2) goal framing; and 3) risky-choice framing. Attribute framing focuses the attention of the recipient on different characteristics of an item (e.g. success rate vs. failure rate, percentage of fat vs. percentage lean). Results show that positive messages are consistently more effective than negative ones. Goal framing focuses on the positive effects or gains of performing a   118  behaviour, or in the negative outcomes or losses of not performing it (Levin et al., 1998). Results show that negative frames have stronger effects than positive frames, although evidence is not fully consistent. Finally, in risky-choice framing, outcomes of varying riskiness are presented in a positive light as gains (e.g. lives saved) or in a negative light as losses (e.g. lives lost). Negatively framed messages tend to be more effective for risky outcomes, whereas positive frames are more effective for outcomes that are certain, although experimental results are still quite mixed (Levin et al., 1998).  Like equivalence framing, studies involving issue framing have also been applied to several topics, such as education, consumption, marketing, health and others (Kolandai-Matchett, 2009; Lockwood, 2011; Price et al., 1997; Shen, 2004). A good deal of research has also been applied to environmental issues and, more recently, to the communication of climate change (Cheng et al., 2011; Davis, 1995; Gifford & Comeau, 2011; Loroz, 2007; Schuldt & Roh, 2014; Spence & Pidgeon, 2010). Many studies have used the traditional gain versus loss frames. Gain frames focus on the benefits of implementing a certain behaviour or action, while loss frames centre on the losses of not performing that behaviour or action (Loroz, 2007). Loss frames have shown to be slightly more effective (Cheng et al., 2011), but results are mixed (Loroz, 2007). Other frames that have been used (especially on climate change) are: motivational versus sacrifice frame (Gifford & Comeau, 2011), local versus global (Sheppard, 2012), personal versus social (Spence & Pidgeon, 2010), public health, national security, economic development, environmental concern, ethics and morality, among others (Bain, Hornsey, Bongiorno, & Jeffries, 2012a; Hine et al., 2016; Lockwood, 2011; Myers et al., 2012; Nisbet, 2009).  Framing effects indicate that people often do not hold stable and consistent attitudes about many issues, as opinions can easily be shifted depending on how issues are framed (Chong & Druckman, 2007; Shen, 2004). The psychological mechanisms behind framing have been explained by a number of theories. The availability heuristic theory argues that framing, by making some issues more salient in the message, makes these more accessible in people’s minds. “Much of the work on framing seems to regard it as an   119  extension of the priming literature, with accessibility as the main theoretical explanation for framing effects” (Gross & D’Ambrosio, 2004, p. 3). In this sense, framing effects occur only when people already hold in mind the considerations invoked by the message (Chong & Druckman, 2007). For example, a message on artificial intelligence may not shift opinions if people do not have a mental frame or basic understanding about this issue. Nelson, Oxley, et al. (1997) add that framing works by affecting the weight that is given to some dimension of our attitudes, and not only by making some issues more accessible. “Frames appear to activate existing beliefs and cognitions, rather than adding something new to the individual’s beliefs about the issue” (Nelson, Oxley, et al., 1997, p. 235-236).  Multiple factors have been shown to moderate framing effects. Some of these include the strength of the message, individual characteristics of the receiver (e.g. values, prior opinions) and contextual factors (Cesario et al., 2013; Chong & Druckman, 2007; Myers et al., 2012; Tversky & Kahneman, 1981). For instance, levels of sophistication and prior knowledge are important (Price et al., 2005). Framing effects seem to be stronger when individuals already hold previous information about an issue. As framing makes some considerations more available, knowledgeable individuals are likely to have those considerations in mind (regardless of whether they agree with them or not) (Druckman & Nelson, 2003; Nelson, Oxley, et al., 1997). In this sense, frames are recommended for targeting according to the stage of awareness or characteristics of the audience (Cheng et al., 2011; Hine, Phillips, et al., 2013; Pelletier & Sharp, 2008). Nonetheless, there is evidence that framing effects can be weaker, cancelled out and even backfire when people hold strong attitudes about an issue (Levin et al., 1998; Rothman & Salovey, 1997) and when the frame is inconsistent with previous beliefs (Lakoff, 2004; Myers et al., 2012; Shen, 2004). In this sense, determining the mental models of the audience is a very important step in a framing exercise (Cheng et al., 2011; McDonald, 2009; Shome & Marx, 2009) and various messages should be developed to reach diverse audiences (Entman, 1993; Gifford & Comeau, 2011; Myers et al., 2012). In addition, studies have   120  shown that the level of involvement in information processing64 is important. Some evidence suggests that loss-framed messages are more effective under high involvement and gain-framed messages are more persuasive under low involvement65 (Maheswaran & Meyers-Levy, 1990). At the same time, sociodemographic characteristics, such as gender, age and education, have influenced results in some studies (Gifford & Comeau, 2011; Lockwood, 2011; Van de Velde et al., 2010). The same holds true for political affiliation (Gross & D’Ambrosio, 2004; Hardisty et al., 2010; Iyengar, 1991; Lockwood, 2011; Price et al., 2005; Shen, 2004). Although framing studies have been carried out extensively, research gaps still remain. For one, emotional reactions to different message frames have been largely overlooked, as studies have mainly focused on cognitive aspects (Gross, 2008; Gross & D’Ambrosio, 2004; Lecheler et al., 2013; Myers et al., 2012). This is an important area of future research, as emotions and affective reactions have been shown to influence thought processes and decision-making (Gross, 2008; Lodge & Taber, 2013). In addition, more research is needed about the relationship between framing and personal characteristics such as issue involvement, prior attitudes and knowledge (Van de Velde et al., 2010). Moreover, research on how mental models and different audience segments moderate framing effects has not been deeply explored (Hine et al., 2016; Shen, 2004). More importantly, to my knowledge, little research to date has researched framing effects on topics related to post-growth communications.  In summary, this investigation aims to uncover more effective ways in which the idea of transitioning into economies not primarily centered on economic growth, can be better communicated to the public. The particular objectives are to determine the effects of different message frames (about moving into economies not centered on growth) on people’s attitudes, emotions and thoughts, and to explore potential moderating variables                                                  64 According to the Elaboration Likelihood Model, message processing occurs through two alternative routes: 1) Central processing, which is more cognitive and systematic; or 2) Peripheral processing which is faster, involves less elaboration and uses more heuristic cues. Individual’s motivation (e.g. involvement in the topic) and ability (e.g. time, knowledge) will influence the processing route that is taken (Perloff, 2010).  65 This is explained by our negativity bias, as negative information is overweighed when making judgements under high involvement, and by the associative model, as we use heuristic cues under low involvement (such as the positive feelings associated with a message).   121  of framing effects. It is important to note that this study is largely exploratory and it aims to look for a wide range of patterns without strong predictions about which ones will be found. Specifically, the research questions that this chapter addresses are: 1) What are the effects of different message frames on participants’: a) cognitive responses, b) emotional reactions, and c) attitudes towards economic growth in Canada? 2) Are framing effects moderated by audience segments, level of involvement with the issue and sociodemographic factors? Figure 4.1 illustrates the different effects that are explored in this chapter.           Figure 4.1 Main and interaction effects explored in this chapter. 4.2 Methodology 4.2.1 Data collection methods 1,250 Canadian residents participated in an online survey in January 2016, implemented by the ResearchNow market research agency. Participants were randomly contacted by Research Now from their Canada-wide pre-enlisted online panel and were provided with basic information about the study (e.g. subject matter, estimated length of exercise). Interested participants clicked on the survey link, which took them to the questionnaire cover letter and survey. For maintaining data quality, surveys with low quality (i.e. same response in most or all questions) and from participants who completed it too quickly (i.e. speeders) were removed. After this process, a total of 1001 surveys remained. Results Message frames Thoughts Emotions Attitudes •Audience segments • Issue involvement •Sociodemographics   122  from this sample cannot be generalized to the Canadian population at large, because like the majority of online panels, it is a non-probability sample. Table 4.1 details the sociodemographic characteristics of the sample.  Table 4.1 Sociodemographic characteristics of survey respondents.  Variable n Valid % Gender Female 531 53.4 Male 463 46.6 Political identification Conservative 238 24.4 Liberal 351 36.0 NDP 147 15.1 Green 61 6.3 None 177 18.2 Age Under 25 74 7.6 25-34 222 22.7 35-44 170 17.4 45-54 213 21.7 55-64 195 19.9 65 or Above 105 10.7 Income Under $40,000 243 28.0 $40,000 to 75,000 226 26.1 $75,000 to $100,000 153 17.6 $100,000 to $150,000 158 18.2 $150,000 and over 87 10.0 Education Less than high school 33 3.3 Completed high school 164 16.4 Some college or university 180 18.0 College or technical school certificate 233 23.4 Received a bachelor's degree 269 27.0 Received a graduate university degree 119 11.9    123  Surveys were programmed in FluidSurvey software. Five versions of the same survey (i.e. four treatment and one control version) 66 were used for testing if different ways in which a message is framed influence cognitive, emotional and attitudinal responses. The survey had the following sections:67 1) opinions and beliefs about economic growth and related issues; 2) concern and interest about economic and environmental issues; 3) framed message; 4) thoughts and opinions about the framed message; 5) assessment of the framed message (e.g. credibility, strength); 6) emotional reaction to the framed message: 7) attitudes towards economic growth in Canada; 8) voting likelihood, priorities between economic growth, environmental issues and social well-being, and opinions about the desired rate of economic growth in the next decade; 9) awareness about economic terms; and 10) sociodemographics. As mentioned in Chapter 2, the survey questions that ask about economic growth refer to increases in GDP. This was explained to respondents before introducing the first battery of questions (see Appendix F). Most sections contained close-ended items only, although Sections 4, 6, 8 and 10 included open-ended questions. The survey itself is seen in Appendix F. A between-subjects design was used. Each sampled individual received only one of five randomly distributed conditions. About 200 participants were assigned in each condition. The four treatment surveys were identical, except for Section 3 which contained the framed message. The control condition did not include Sections 3, 5 and 6, as no message was provided, while Section 4 was modified to request participant’s general opinions, rather than their specific views about the framed message. The average completion time for the treatment surveys was 17 minutes, while for the control survey, it was 13 minutes.                                                  66 In this dissertation, the term ‘condition’ is used for referring to the four frames and the control condition, whereas the term ‘treatment’ is used for referring to one or more of the four frames (excluding the control condition). 67 Sections 1 and 2 of the survey are analyzed and discussed in Chapter 3, while Section 9 is not analyzed nor discussed in this dissertation.   124  4.2.1.1 Variables 4.2.1.1.1 Independent Variable Framed message (Section 3): Each framed message consisted of one paragraph up to 180 words that emphasized one of four different angles related to transitioning into an economic model not based on indefinite economic growth. The message versions were focused on: 1) potential environmental gains of transitioning into a new economic model; 2) potential environmental losses of not transitioning into a new economic model; 3) potential gains in well-being of transitioning into a new economic model; and 4) potential losses in well-being of not transitioning into a new economic model. Table 4.2 illustrates the four treatments and Table 4.3 provides the wording for each. The control version did not include any framed message. Hereafter, each condition will be referred as Frame 1, Frame 2, Frame 3, Frame 4 and Control.  Table 4.2 Description of message frames used in this study.  Types of framed messages Gains Losses Environment Frame 1 Environmental message focused on the gains of transitioning into a new model Frame 2 Environmental message focused on the losses of not transitioning into a new model Well-Being Frame 3 Well-being message focused on the gains of transitioning into a new model Frame 4 Well-being message focused on the losses of not transitioning into a new model   Environmental and well-being themes were chosen for the messages, because these highlight different aspects of the transition to sustainability. Also, the environmental frame has been commonly used by environmentalists, while the well-being frame has been employed to a much lesser extent in environmental and sustainability communications. Another distinction between treatments is that Frames 1 and 2 (messages focused on the environment) focus more explicitly on limits to growth and the feasibility of indefinite growth, while Frames 3 and 4 (messages focused on well-being) put more attention on the desirability of continuous growth.   125  Table 4.3 Wording for each treatment condition. Frame Text 1 Economic growth is a main goal for most governments. During the past century, economic production increased fifteen-fold while, at the same time, humanity’s use of fossil fuels and other natural resources has increased at unprecedented levels. Natural resources and a healthy environment are the foundation of economic activity and a flourishing society, but around the world they have become scarce. As a result, some experts propose that, over the next decade, developed countries should stabilize their levels of production and consumption and find an alternative to economic growth as a major policy goal.  Making this transition will eventually lead to less global pollution and the restoration of vital resources, may reduce human mass migrations caused by environmental scarcity, and will likely increase social order and harmony around the world. The message is clear: Moving beyond perpetual economic growth is not only possible, but could, if founded on a healthy environment, lead to the creation of more stable societies and ways of life. 2 Economic growth is a main goal for most governments. During the past century, economic production increased fifteen-fold while, at the same time, humanity’s use of fossil fuels and other natural resources has increased at unprecedented levels. Natural resources and a healthy environment are the foundation of economic activity and a flourishing society, but around the world they have become scarce. As a result, some experts propose that, over the next decade, developed countries should stabilize their levels of production and consumption and find an alternative to economic growth as a major policy goal. Failing to make this transition will eventually lead to greater global pollution and the depletion of vital resources, may increase human mass migrations due to environmental scarcity, and will likely cause greater social disorder and public unrest around the world. The message is clear: Perpetual economic growth is not only impossible, but will soon lead to the destabilization and collapse of our societies and ways of life. 3 Economic growth is a main goal for most governments. During the past century, economic production increased fifteen-fold; however, people in developed countries like Canada, are not becoming much happier. Studies show that once basic needs are met, more income and consumption do not seem to improve our real welfare. As a result, some experts propose that over the next decade, developed countries should stabilize their levels of production and consumption and find an alternative to economic growth as a major policy goal. Making this transition would liberate us from the burden of pursuing material excesses and may well increase our personal well-being and life satisfaction. The possibilities of creating truly worthwhile and happy lives would increase, with more time for family and community, stronger quality of relationships and connections and possibly, even better levels of physical and mental health, because we will be devoting less of our time, energy and resources to more consumption. The message is clear: Moving beyond a culture focused on material growth could provide a better path to happiness and genuine human improvement. 4 Economic growth is a main goal for most governments. During the past century, economic production increased fifteen-fold; however, people in developed countries like Canada, are not becoming much happier. Studies show that once basic needs are met, more income and consumption do not seem to improve our real welfare. As a result, some experts propose that over the next decade, developed countries should stabilize their levels of production and consumption and find an alternative to economic growth as a major policy goal. Failing to make this transition could mean losses in our personal well-being and life satisfaction, as we would be increasingly burdened by the pursuit of material excesses. The possibilities of creating truly worthwhile and happy lives would decrease, with less time for family and community, lower quality of relationships and connections, and possibly, even diminishing physical and mental health, because we will be placing too much of our time, energy, and resources to more consumption. The message is clear: The culture of material growth is a poor path to happiness and genuine human improvement.   126  4.2.1.1.2 Confounding variables Assessment of the framed message (Section 5): Using three close-ended interval scale-type questions, this section measured participant’s assessment on the credibility, convincingness and strength of the message. This section aimed to verify that all frames were perceived with equal strength and quality, so that this would not be a confounding factor in the experimental condition.    4.2.1.1.3 Dependent Variables Cognitive response to frames (Section 4): Using an open-ended question, this section aimed to capture respondent’s thoughts after reading the message. The wording of this statement was based on Price et al. (1997).  Emotional reaction to frames (Section 6): Using three close-ended interval scale-type questions, this section measured emotional reactions towards the framed message, specifically hope, fear and anger. Fear and anger were chosen because they have been frequently studied (Angie, Connelly, Waples, & Kligyte, 2011; Gross & D’Ambrosio, 2004; Nabi, 2003) and have been identified as ‘basic emotions’ due to their universality across cultures (Jack, Garrod, & Schyns, 2014). Hope was included due to its positive emotional state and its relation to agency68 (Snyder, Rand, & Sigmon, 2005), as it can be important for encouraging action and countering despair (Stevenson & Peterson, 2016). For capturing the reasoning behind these feelings, three open-ended questions were also included. Question wording was based on Gross and D’Ambrossio (2004).  Attitudinal responses to frames (Sections 7 and 8): With a matrix of 10 Likert scale-type statements, participant’s attitudes towards economic growth and related issues in Canada were explored. Three items (1, 3 and 4) were based on Drews’ (2016) survey on economic growth. All items, except 5 and 9 related directly to economic growth (as measured in GDP), while items 5 and 9 address people’s views about material consumption. Section 8 included questions related to participants’ voting likelihood (for a                                                  68 Agency is defined as “the perceived capacity to use one’s pathways so as to reach desired goals” (Snyder et al. 2005, p. 258).    127  politician that does not pursue economic growth as a primary goal); economic, social and environmental priorities; and the desired level of economic growth for the next decade. Finally, one open-ended question explored the reasoning behind voting preferences.  4.2.1.1.4 Moderator Variables Audience segments (Section 1) (i.e. pre-existing attitudes): With a matrix of 12 close-ended Likert scale-type statements, participant’s views about economic growth and the environment were measured, with the purpose of classifying respondents into distinct audience segments or clusters. An in-depth description of the different segments uncovered are described in Chapter 3 of this dissertation. Issue involvement (Section 2): Participant’s involvement in environmental and economic issues was measured with three close-ended interval scale-type questions. Question wording was based on Schuldt and Roh (2014) and Maibach et al. (2011).  Sociodemographics (Section 9): Using a combination of close-ended multiple choice items and open-ended questions, information regarding gender, age, education, ethnicity, place of residence, occupation, income and political affiliation were collected. 4.2.2 Data analysis 4.2.2.1 Qualitative analysis Data from open-ended questions were loaded into NVivo 10 qualitative software for the corresponding coding and analysis. A total of four open-ended questions were analyzed, corresponding to two dependent variables: cognitive responses and emotional reaction. These four questions were related to: 1) cognitive responses to the frames, and explanations for feelings of 2) hope, 3) fear and 4) anger.69 Qualitative coding was carried out by two coders: the principal researcher and one undergraduate student at the University of British Columbia. Data from cognitive responses were inductively coded70                                                  69 For the wording of these questions, see Sections 4 and 5 of the survey in Appendix F.  70 In inductive coding, the codes emerge from the data, while in deductive coding, the codes are pre-determined by theory.   128  and re-coded by the principal researcher, who determined the 10 to 15 most referenced codes. Then, the student coder deductively coded data into these 10 or 15 codes.71 Data regarding emotional reaction, in contrast, was inductively coded by both coders. Each coder classified data into general and specific thematic codes and developed a codebook with the definition and meaning of each code and subcode. Codebooks were compared to ensure consistency in coding.72 Inter-coder agreement for all codes (for cognitive and emotional responses) was measured with NVivo’s coding comparison query function, which provides a Kappa coefficient as a measure of agreement. Coders discussed any discrepancies and re-coded data until average Kappa coefficients of 0.8 or higher were reached.73 After this process of coding and re-coding, queries and matrices were developed in NVivo 10 to determine the most mentioned themes in each question for each of the treatment conditions. This allowed for a qualitative description of the data and for establishing comparisons between the frames.  4.2.2.2 Quantitative analyses Data from close-ended items were transferred from FluidSurvey into IBM SPSS Statistics 23 software for the corresponding quantitative analysis. Each survey was granted a unique code (between 1 and 1001) and responses were numerically coded. Table 4.4 summarizes the numerical coding for each section of the survey. Lower scores indicated less agreement with the corresponding variable. Missing values were coded with a -99. After checking for normality using a Kolmogorov-Smirnov test, all dependent variables were deemed to be not normally distributed. Moreover, many violated the assumption of homogeneity of variance (measured using Lavene’s test). Consequently, only non-parametric tests were used. A Kruskal-Wallis test was used to examine for significant differences between the multiple dependent variables across the treatment conditions. A Mann-Whitney U-test was employed for checking significant differences between gain-                                                 71 Data from cognitive responses was not coded inductively by the student coder, due to the large amounts of data and due to time constraints. 72 Coders discussed the different codes created to ensure consistency. For example, some codes were named differently by each coder, although they were meant to capture the same type of responses.  73 Kappa coefficients of 0.8 or higher are generally deemed acceptable (Hruschka et al., 2004).     129  framed messages (Frames 1 and 3) and loss-framed messages (Frames 2 and 4). In addition, ordinal logistic regressions were carried out in SPSS to explore potential covariates and moderating variables (using the two-way model effects). The SPSS split file command allowed for further exploration of interactions and for determining the framing effects on specific subgroups.  Table 4.4 Numerical coding for different sections of the survey, categorized by the type of variable. Confounding Variables Section 5 Credibility, persuasiveness and strength of message frames  Not credible/convincing at all=1; somewhat credible/convincing=2; slightly credible/convincing=3; and, very credible/convincing=4. Very weak=1; somewhat weak=2; somewhat strong=3; and, very strong=4. Dependent Variables Section 6 Hope, fear and anger Not at all hopeful/fearful/angry=1; a little hopeful/fearful/angry=2; somewhat hopeful/fearful/angry=3; very hopeful/fearful/angry=4; and, extremely hopeful/fearful/angry=5. Section 7 & 8 Attitudinal questions Voting likelihood Growth in the next decade  Strongly disagree=1; moderately disagree=2; neither disagree nor agree=3; moderately agree=4; and, strongly agree=5. Very unlikely=1; somewhat unlikely=2; somewhat likely=3; and, very likely=4. Less=1; About the same=2; More=3; and, I don’t know=-7. Moderating Variables Section 1 Pre-existing attitudes  Strongly disagree=1; moderately disagree=2; neither disagree nor agree=3; moderately agree=4; and, strongly agree=5. Section 2 Issue involvement Frequency of thought  Not at all concerned=1; a little concerned=2; somewhat concerned=3; very concerned=4; and, extremely concerned=5. Never=1; not very much=2; a fair amount=3; and, a great deal=4. Section 9 Sociodemographics  Gender: Female=1; Male=2; and, Other=3. Age: Less than 25=1; 25 to 34=2; 35 to 44=3; 45 to 54=4; 55 to 64 = 5; and, 65 or above=6. Education: Less than high school=1; Completed high school=2; Some college or university=3; Received a college or technical school certificate=4; Received a university's bachelor degree=5; Received a graduate degree=6. Income: Under 40,000=1; 40,000 to 75,000=2; 75,000 to 100,000=3; 100,000 to 150,000=4; and, 150,000 and over=5. Political ideology: Conservative Party=1; Liberal Party=2; NDP=3; Green=4; Other=5; and None=6.    130  4.3 Results This section describes framing effects (if any) on each dependent variable: 1) cognitive responses; 2) emotional reaction; and 3) attitudinal responses. In addition, when framing effects are found to be significant, it explores possible covariates and moderating variables.  4.3.1 Confounding variables No statistically significant differences were found between the message frames in terms of their credibility, persuasiveness and strength. These results indicate that all frames were perceived of similar quality.  4.3.2 Framing effects on cognitive responses In order to measure cognitive responses to the frame, participants were asked to write down up to three thoughts or ideas that they were currently thinking (after reading the message). In the control condition, participants were asked to write down up to three thoughts or ideas that they currently had about the economy, society or the environment, and/or how they affect each other. Results from the qualitative analysis show that each condition generated several cognitive responses. For simplicity, only the three most mentioned themes in each condition (eight in total) are shown in Table 4.5.74 The most striking differences emerged on the themes of ‘overconsumption and materialism’ and ‘happiness and well-being’. These topics were largely mentioned by participants that received Frames 3 and 4 (i.e. messages focused on well-being), while they were scarcely referenced in any of the other conditions. The following quotes provide some examples of the specific issues mentioned: “Consumption level should be decreased. Consumption doesn't provide any happiness” (Frame 3, ID 602); “It is true that there is too much focus on economic growth. We don’t need much to be happy” (Frame 4, ID 826).                                                  74 Table 4.5 shows the most mentioned themes. Directionality or content of opinions is not self-evident in most themes included in this table (e.g. whether participants thought that consumerism is positive or negative).   131  Similar results emerged for the ‘natural resources, environment and sustainability’ theme. These were mentioned with a higher frequency in the control condition and then in Frames 1 and 2 (i.e. messages focused on the environment), while they were referenced much less in Frames 3 and 4. Interestingly, the topic of climate change was brought up much more frequently in the control condition than in any other condition. Also, ideas related to ‘negative outcomes and collapse’ were much more referenced in Frame 2 (i.e. message focused on the environment and losses) and the control condition than in any other frames. Some illustrations of this theme included: “We can easily eat up the ground we live on and collapse” (Control, ID 39); “Armaggedon in the works. If not extremely difficult, then impossible to reverse the paths & damages” (Frame 2, ID 415). Topics related to the ‘connection between the economy and the environment’, were, by far, mentioned more often in the control condition, followed by Frames 1 and 2. In this theme, many respondents made reference to the impact of the economy on the environment and pointed out the dependency of the former on the latter. Nonetheless, a few participants also expressed the view that it is possible to harmonize economic growth and sustainability.   Table 4.5 Most frequent cognitive responses that emerged in each condition.  Themes Control Frame 1 Frame 2 Frame 3 Frame 4 n=180 n=187 n=186 n=186 n=188 # % # % # % # % # % Natural resources, environment and sustainability 126 70.0 100 53.5 108 58.1 40 21.5 34 18.1 Issues related to the economy 66 36.7 44 23.5 36 19.4 45 24.2 50 26.6 Connection between the economy and the environment 74 41.1 39 20.9 42 22.6 6 3.2 4 2.1 Overconsumption and materialism 8 4.4 14 7.5 13 7.0 78 41.9 68 36.2 Happiness and well-being 3 1.7 8 4.3 2 1.1 73 39.2 69 36.7 Negative outcomes and collapse 16 8.9 8 4.3 23 12.4 3 1.6 4 2.1 Positive comments about the message frame (e.g. agree with ideas) 0 0.0 44 23.5 37 19.9 64 34.4 46 24.5 Negative comments about the message frame (e.g. disagree, unrealistic) 0 0.0 41 21.9 23 12.4 34 18.3 23 12.2 Note: The # column indicates the number of participants that mentioned each theme and the % column shows the proportion of respondents within a condition that mentioned each theme. Some responses were classified into more than one code.   132  ‘Issues related to the economy’ were mentioned much more frequently in the control condition that in any other condition. Topics within this theme included inflation and currency, jobs and the importance of economic stability and diversification, among others. Many participants in the control condition expressed negative views about the current economic condition. For instance, some comments included: “Economy is not doing so well” (Control, ID 19); “The Canadian economy is in the tank caused by the environment society” (Control, ID 177).  Although positive comments about the message frames emerged in all treatment conditions, Frame 3 (i.e. message focused on well-being and gains) was the one with the greatest number of affirmative remarks. More than 30% of the participants in this group expressed their agreement with the ideas proposed in the message. Some of the comments included: “This is a great idea.  People have become endless consumers […]” (Frame 3, ID 623); “I believe this article is bang on. Letting go of material things and focusing on what we have right now will greatly improve our lives and make us happier” (Frame 3, ID 633); “I agree fully. Consumerism has become a necessity of life” (Frame 3, ID 647). Even so, between 10% and 13% of the participants in Frames 1 and 3 also mentioned that the ideas in the message were unrealistic and unfeasible.  4.3.3 Framing effects on emotional reaction 4.3.3.1 Quantitative analysis: Differences between conditions The message frames had different effects on the emotional reaction that they generated among participants. Figure 4.2 illustrates Kruskal-Wallis pairwise comparisons. Frame 2 (i.e. message focused on the environment and losses) was significantly different from the other frames in that it had a significantly lower mean rank for hope and a higher mean rank for fear than all the other frames (p < 0.05). Also, Frame 2 had a higher mean rank for anger, especially compared to Frames 1 and 3 (i.e. gain-framed messages), while differences were not statistically significant with Frame 4 (i.e. message focused on well-being and losses). Similarly, when analyzed in combination, loss-framed messages (Frames 2 and 4) were significantly different from gain-framed messages in that they   133  generated less hope, more fear and more anger (p < 0.000). The frequency proportions, means and standard deviations for each emotion, categorized by each frame are depicted in Table 4.6, Table 4.7 and Table 4.8. These tables confirm that the percentage of not at all hopeful was significantly higher for Frame 2 than for the other conditions, while the percentages of not at all fearful and not at all angry were significantly lower for Frame 2. In contrast, Frame 3 generated the opposite reactions (i.e. more hope, less fear and less anger), although the mean ranks were not significantly different from those of Frames 1 and 4.     Figure 4.2 Kruskal-Wallis pairwise comparisons for a) hope, b) fear and c) anger. Gray lines show significant differences (p < 0.05), while black lines or no lines indicate non-significant differences between frames. Table 4.6 Frequency proportions, means and standard deviations for reported feelings of hope between treatments.  Frame % distribution for hope Mean SD n Not at all hopeful A little hopeful Somewhat hopeful Very hopeful Extremely hopeful 1 30.3 28.8 29.8 8.6 2.5 2.24 1.07 198 2 46.7 27.1 20.6 5.0 0.5 1.85 0.95 199 3 28.0 24.0 34.0 13.0 1.0 2.35 1.06 200 4 33.5 25.5 32.0 8.0 1.0 2.18 1.02 200   a) Hope  b) Hope  c) Hope  d) Hope b) Fear  b) Fear  b) Fear  b) Fear c) Anger  c) Anger  c) Anger  c) Anger   134  Table 4.7 Frequency proportions, means and standard deviations for reported feelings of fear between treatments. Frame % distribution for fear Mean SD n Not at all fearful A little fearful Somewhat fearful Very fearful Extremely fearful 1 42.6 32.5 17.3 6.1 1.5 1.91 0.99 197 2 27.5 33.5 23.5 12.5 3.0 2.30 1.09 200 3 48.2 27.1 22.1 2.0 0.5 1.79 0.89 199 4 38.0 33.0 24.0 4.0 1.0 1.97 0.94 200  Table 4.8 Frequency proportions, means and standard deviations for reported feelings of anger between treatments. Frame % distribution for anger Mean SD n Not at all angry A little angry Somewhat angry Very angry Extremely angry 1 61.6 17.7 14.1 3.0 3.5 1.69 1.05 198 2 47.0 20.5 23.5 7.0 2.0 1.97 1.08 200 3 67.8 17.1 10.6 2.5 2.0 1.54 0.93 199 4 56.5 22.0 16.0 3.5 2.0 1.73 0.99 200  4.3.3.2 Covariates and moderating variables for emotional reaction In order to identify moderating variables of framing effects on emotional response and other covariates,75 cumulative odds ordinal logistic regressions were carried out in SPSS.76 The dependent variables were level of hope, level of fear and level of anger. One regression was carried out for each dependent variable. The independent variables included in the regressions were: message frame;77 audience segments (i.e. detailed in Chapter 3); issue involvement (as measured in level of concern for environmental and                                                  75 Data from the control condition was excluded from the analysis as participants in this condition did not answer this section of the survey. 76 Ordinal logistic regression allows having an ordinal variable as the dependent variable and categorical, ordinal or interval variables as predictors.  77 In ordinal logistic regression, SPSS uses the last category of a variable as the reference category. For message frame, it made sense to use Frame 2 as the reference category because results from Kruskal-Wallis non-parametric test indicated that this frame generated significantly different emotional responses in comparison with the other frames.   135  economic issues)78; gender; and, political affiliation. Collinearity diagnostics were obtained for all independent variables. The lowest value for tolerance was 0.548 and the highest VIF was 1.826, indicating no apparent multicollinearity problems.79 All regressions met the assumption of proportional odds, which was assessed by a test of parallel lines (p > 0.05).80 The ordinal logistic regression models for hope, fear and anger were statistically significant (hope: χ2(11) = 45.827, p < 0.001; fear: χ2(11) = 78.183, p < 0.001; anger: χ2(11) = 57.043, p < 0.001). Table 4.9 summarizes the model effects for each of the dependent variables. Consistently, message frame was the only independent variable that was a significant predictor for all regressions.  Table 4.9 Test of model effects for the dependent variables hope, fear and anger. Independent Variables Hope Fear Anger Wald Chi-Square df Sig. Wald Chi-Square df Sig. Wald Chi-Square df Sig. Frame 27.437 3 .000 25.215 3 .000 21.390 3 .000 Audience segment 6.561 2 .038 5.165 2 .076 2.616 2 .270 Concern Environment 2.001 1 .157 9.804 1 .002 7.609 1 .006 Concern Economy 0.850 1 .356 17.988 1 .000 7.919 1 .005 Gender 0.009 1 .925 6.044 1 .014 0.278 1 .598 Political 9.665 3 .022 1.054 3 .788 12.419 3 .006                                                    78 Due to the very low number of participants that reported being ‘not at all’ or ‘a little’ concerned about the state of the environment and the national economy, and in order to carry out the relevant statistical analyses, categories of environmental and economic concern were grouped into two main groups: the less concerned (which included the ‘not at all concerned’, ‘a little concerned’ and ‘somewhat concerned’) and the more highly concerned (which included ‘very concerned’ and the ‘extremely concerned’). 79 Tolerance values lower than 0.1 or VIF values greater than 10 may represent a problem. 80 The test of parallel lines compares the model fit between two cumulative odds models. The null hypothesis model assumes that the lines are parallel, while the second model (called general model) is for separate lines. If the assumption of proportional odds is met, then the difference between the two models is expected to be non-significant (p > 0.05) (Norusis, 2011).    136  4.3.3.2.1 Hope  Significant associations with the dependent variable hope (p = < 0.05) were found for message frame, audience segment and political affiliation, as shown in Table 4.9. Based on the parameters estimates shown in Appendix G.1,81 Frames 1, 3 and 4 had statistically significant higher odds of generating more hope than Frame 2. Specifically, the odds of Frame 3 generating more hope than Frame 2 were 2.86 (95% CI: 1.885 to 4.384, Wald χ2(1) = 24.060, p = 0.000), followed by the odds of Frame 1 that were 2.42 (95% CI: 1.580 to 3.702, Wald χ2(1) = 16.534, p = 0.000) and Frame 4 that were 1.95 (95% CI: 1.269 to 2.985, Wald χ2(1) = 9.304, p = 0.002). Also, the odds of participants in Cluster 1 reporting higher levels of hope than those in Cluster 3 was 1.69 (95% CI: 1.127 to 2.548), a statistically significant effect (Wald χ2(1) = 6.415, p = 0.011). Regarding political identification, respondents identified with the Green Party had higher odds of reporting a lower level of hope than those identified with the Conservative Party (odds at 0.53, 95% CI: 0.280 to 0.982, Wald χ2(1) = 4.064, p = 0.044). In order to explore the effects of any moderating variables, two-way interactions between variables were further explored with the ordinal regression.82 Table 4.10 shows interaction effects for the message frames and each of the dependent variables. For hope, statistically significant interactions were found only for message frame and audience segment (p = 0.31). For exploring this interaction in detail, framing effects were analyzed for each cluster separately. Data revealed that framing effects were significant for members of Clusters 2 and 3 only (see parameter estimates table in Appendix H.1). No significant effects for hope were evidenced for members of Cluster 1. Effects were greater for participants in Cluster 3, as Frames 1 (p = 0.000), 3 (p = 0.000) and 4 (p = 0.001) generated significantly higher levels of hope than Frame 2, while for participants                                                  81 Significance values can be obtained from the significance column (under hypothesis test) in the Parameter Estimates tables in Appendix G. The odds ratio can be obtained from the Exp(B) column in the same tables.   82 Two-way interactions were tested for: frame*gender, frame*audience segment, frame*political and frame* economic concern.    137  in Cluster 2, only Frame 3 was significantly different (p = 0.006) from Frame 2 (in that it had higher odds of generating greater hope).  Table 4.10 Test of interaction effects for the dependent variables hope, fear and anger. Independent Variables Hope Fear Anger Wald Chi-Square df Sig. Wald Chi-Square df Sig. Wald Chi-Square df Sig. Frame * Audience segment 13.886 6 0.031 8.668 6 0.193 8.415 6 0.209 Frame * Concern environment 6.601 3 0.086 3.999 3 0.262 2.496 3 0.476 Frame * Concern economy 0.761 3 0.859 2.881 3 0.410 1.146 3 0.766 Frame * Gender 5.482 3 0.140 11.972 3 0.007 3.796 3 0.284 Frame * Political 13.063 9 0.160 5.218 9 0.815 12.885 9 0.168  4.3.3.2.2 Fear  Significant associations with the dependent variable fear (p = < 0.05) were found for message frame, concern for the environment, concern for the economy and gender, as shown in Table 4.9. Specifically, Frame 1 (Wald χ2(1) = 19.949, p = 0.000), Frame 3 (Wald χ2(1) = 17.734, p = 0.000) and Frame 4 (Wald χ2(1) = 9.704, p = 0.002) were significantly more likely of generating lower levels of fear than Frame 2. Less concern for the state of the environment (Wald χ2(1) = 9.804, p = 0.002) and the economy (Wald χ2(1) = 17.988, p = 0.000) were associated with higher probabilities of reporting lower levels of fear. Regarding gender, the odds ratio of women reporting higher levels of fear than men was 1.47 (95% CI: 1.081 to 2.002), a statistically significant effect (Wald χ2(1) = 6.044, p = 0.014). Although the variable audience segment only approximated significance (p = 0.76) as shown in Table 4.9, when giving a closer look at the parameter estimates table (Appendix G.2), data shows that members of Cluster 1 were more likely to report lower levels of fear (Wald χ2(1) = 5.100, p = 0.024) than members of Cluster 3. Two-way interactions between message frame and other variables were tested and the only significant interaction found was between message frame and gender (p = 0.007) as shown in Table 4.10. Framing effects were analyzed for men and women separately. Data   138  shows that framing effects for fear only occurred among women, while there were no significant effects among men (as shown in the parameter estimates table in Appendix H.2). Specifically, Frame 1 (Wald χ2(1) = 15.842, p = 0.000), Frame 3 (Wald χ2(1) = 25.443, p = 0.000) and Frame 4 (Wald χ2(1) = 7.204, p = 0.007) generated lower levels of fear than Frame 2. Frame 3 had the higher odds of generating the least fear.  4.3.3.2.3 Anger Significant associations with the dependent variable anger (p = < 0.05) were found for message frame, concern for the environment, concern for the economy and political identification, as shown in Table 4.9. Similar to the effects for fear, Frame 1 (Wald χ2(1) = 10.651, p = 0.001), Frame 3 (Wald χ2(1) = 18.984, p = 0.000) and Frame 4 (Wald χ2(1) = 5.317, p = 0.021) were significantly more likely of generating less anger than Frame 2. Also, less concern for the state of the environment (Wald χ2(1) = 7.609, p = 0.006) and the economy  (Wald χ2(1) = 7.919, p = 0.005) were associated with higher probabilities of reporting lower levels of anger. Regarding political identification, those identified with the Liberal party and NDP have higher odds at reporting lower levels of anger than those identified with the Conservative Party (Wald χ2(1) = 9.218, p = 0.002 and Wald χ2(1) = 4.259, p = 0.039, respectively). Differences were non-significant between Greens and Conservatives. Two-way interactions between message frame and other variables were tested, but no significant interactions were found for anger (as shown in Table 4.10).  4.3.3.3 Qualitative analysis for emotional reaction Responses to open-ended questions were used to further clarify framing effects on the emotional responses of hope, fear and anger. Participants provided multiple explanations to justify the feelings they encountered while reading the message. Only the three to five most mentioned themes in each frame will be presented, with the purpose of establishing some comparisons. For each emotion, responses were coded into two main categories:83 1) those that directly mentioned or reacted towards the framed message specifically, and                                                  83 Some responses, especially for anger, did not fit into these two main categories. For example, some referred to other emotions (e.g. sadness), while others were deemed too ambiguous (e.g. “the truth hurts”, “angry at everyone including myself”).    139  2) those that mentioned or reacted towards the broader context. Table 4.11 summarizes the number of responses for each of these two main categories, illustrating that most comments were about the general context. As expected, Frame 2 generated the largest amount of comments for fear and anger, and the least for hope. The opposite trend emerged for Frame 3. The references made in relation to each emotion will be discussed in turn.  Table 4.11 Number of responses pertaining to the content of the message and the general context.  Frame Hope Fear Anger Message Context Message Context Message Context 1 37 87 16 82 10 48 2 18 69 22 110 15 74 3 44 97 11 78 11 39 4 21 98 22 78 19 55 Note: There was minor overlap between these two main categories (i.e. message and content), with only a few references being coded in both. 4.3.3.3.1 Hope  Participants expressed hopeful comments regarding the messages presented, especially for Frames 1 and 3 (as shown in Table 4.12). Many expressed their agreement towards the ideas in the paragraph, especially for Frame 3, where a few participants explicitly mentioned that the message offered them hope (e.g. “It made it seem like we could actually change the feelings of sadness in our society” [Frame 3, ID 124]; “Because the way it was worded made me feel very hopeful for mine and my children's future” [Frame 3, ID 163]). Frame 3 received the most supportive comments, while Frame 2 received the least. Some disapproving comments also emerged out of all frames. Some participants expressed that the messages were simplistic, unappealing and insubstantial, while others mentioned that the ideas espoused in the message would be difficult and unrealistic to implement. A few participants expressed uneasiness about Frames 2 and 4 deliberately causing fear.     140  Table 4.12 Most commonly mentioned themes that emerged for hope for each frame, in relation to the message and context.  Themes (Hope) Frame Total 1  2 3 4 Related to the message Supportive comments 28 6 34 11 79 Critical comments 9 12 13 12 46 Related to the context Reasons for being hopeful  61 54 73 65 253 Other people know or are concerned 23 14 22 19 78 Change is posible 8 15 16 19 58 Monetary and material priorities are changing 0 1 22 14 37 Sustainability 14 7 5 4 30 Trust in human capacity 5 10 2 4 21 Reasons for being less hopeful  29 17 29 42 117 Difficulty of change 5 1 6 9 21 Difficulty of generating awareness and buy-in 4 4 7 5 20 Human traits (e.g. greed) 10 3 2 4 19 Feeling of despair or disempowerment 6 3 3 6 18 Note: Each cell represents the number of participants that mentioned each theme. Some participants mentioned more than one theme, so there is some overlap. Participants brought up many contextual issues that made them feel more or less hopeful. As shown in Table 4.12, some participants felt hopeful because they believed that others are already aware, concerned or thinking along similar lines. The following quote reflects this feeling: “Reading this article allowed me to see that there are others in the world who believe in another definition of progress and are interested in changing the world” (Frame 4, ID 978). Some participants also mentioned that it is not too late to bring about change (e.g. “Because it's never too late to change” [Frame 3, ID 610]). Other references in Frames 3 and 4 showed hope for a shift away from material and monetary values to other more meaningful goals. Finally, a few participants were hopeful about human’s capacity for adaptation, innovation and problem-solving (e.g. “It would seem to me/when the time comes to replace a resource that has been depleted/MAN will do so” [Frame 2, ID 516]). Other participants felt less hopeful due to the difficulties of generating awareness and buy-in from people (e.g. “Because people do not listen” [Frame 2, ID 596]; “Most people are still in denial” [Frame 4, ID 842]). Additionally, some were pessimistic about human   141  nature (e.g. “I like the idea but fear human nature would get in the way” [Frame 1, ID 253]; “I believe the human species is not capable of ignoring (or suppressing) our basic greed and self-serving nature” [Frame 3, ID 729]). Some participants also mentioned feeling disempowered and despaired. 4.3.3.3.2 Fear  Some participants were fearful due to the framed message (as shown in Table 4.13). For instance, some disagreed with the argument as they thought it was false or exaggerated (e.g. “If this is what “experts” believe, we are in big trouble…” [Frame 1, ID 217]). Others were fearful about the economic consequences of implementing such a plan (e.g. impacts on jobs), while a few mentioned, especially for Frame 2, that the message used scare tactics (e.g. doom and gloom messaging) and was designed to generate fear.  Table 4.13 Most commonly mentioned themes that emerged for fear for each frame, in relation to the message and context.  Themes (Fear) Frame Total 1 2 3 4 Related to the message Reject message 12 13 9 15 49 Disagree with stance 5 3 5 6 19 Economic consequences 4 3 2 4 13 Fear mongering 1 8 0 4 13 Related to the context Environment and natural resources 25 34 11 10 80 Difficulty of change 19 22 24 14 79 Future and future generations 9 21 6 12 48 Collapse and other negative outcomes 7 23 8 9 47 Inaction 13 12 7 6 38 Discontent and distrust in government 8 11 12 2 33 Economy 9 13 6 5 33 Greed, selfishness and money focus 6 6 14 7 33 Materialism and loss of other values 1 0 3 11 15 Note: Each cell represents the number of participants that mentioned each theme. Some participants mentioned more than one theme, so there is some overlap.   142  Most participants expressed fear with respect to contextual issues (as shown in Table 4.13). Fear for the state of the natural environment was the most commonly mentioned theme, especially in Frames 1 and 2. Some of the issues brought up included pollution, climate change and resource depletion, although some respondents were much more general in their explanations (e.g. “Who wants to live on a dying planet” [Frame 1, ID 232]; “Because we are ruining our earth” [Frame 2, ID 460]). Another commonly mentioned theme was related to the difficulty of generating change. One participant explained: “I fear that we won't be able to change our ways before irreversible damage has been done to our way of life” [Frame 2, ID 510]). A few pointed out the difficulties of achieving global cooperation. Other participants, especially in Frame 2, were fearful about the future, be it their own or that of future generations. Moreover, some mentioned the possibilities of heading to a very negative state of affairs and even systemic collapse. Common phrases in this theme included: “Because if we continue as we are, societal unrest is a very real possibility” [Frame 2, ID 6]; “Chaos!” [Frame 2, ID 8]; “A collapse of society as we know it is inevitable unless change is made” [Frame 2, ID 13]. Finally, ideas related to greed and selfishness emerged with more prominence in Frame 3, while those related to materialism emerged more in Frame 4.  4.3.3.3.3 Anger  Some participants were angry at the messages because they did not agree with the arguments presented, perceived them to be biased and unfair (e.g. with a hidden agenda) or considered the ideas to be naïve and absurd. The following quote reveals some of the suspicion generated by the message: “It reflects the tepid support by the federal government and the outright hostility shown by the environmental lobbies (funded largely outside of Canada) against the oil industry in Canada” (Frame 4, ID 967); “Why are smart people being paid good money to come up with such ridiculous ideas?” (Frame 1, ID 17).  Many more respondents were angry towards the general context (as shown in Table 4.14). Frustration regarding the inaction of individuals, government and society was the most mentioned theme. The following quote reflects this frustration: “This is not   143  something that just "creeped" up on us....it has been going on for years and years....it seems "we" have been dragging our collective feet […]” (Frame 2, ID 438). Other participants were angry about environmental issues, which were more prominently mentioned in Frame 2. Some mentioned pollution and overuse of resources, while others were more general in their statements (e.g. “Humans are destroying our planet for future generations” [Frame 1, ID 318]; “I'm angry because any future children are going to grow up in a post-apocalyptic wasteland” [Frame 2, ID 562])”. Additionally, participants expressed their anger towards other topics like human greed, ignorance and politics (including the government). Some revealed having feelings of disempowerment and hopelessness: “Because this model in which we're living cannot go on; yet, there is nothing we can do about it” (Frame 4, ID 802). Anger towards materialism and consumerism were mainly mentioned in Frame 4, while no references to these issues were made in Frames 1 and 2. Table 4.14 Most commonly mentioned themes that emerged for anger for each frame, in relation to the message and context.  Themes (Anger) Frame Total 1 2 3 4 Related to the message Disagree with message 0 7 1 5 13 Message is biased 3 2 2 5 12 Message is absurd 4 1 3 4 12 The way its presented 0 4 4 4 12 Related to the context Inaction 18 20 7 6 51 Environmental issues 10 26 1 11 48 Greed and selfishness 6 10 6 10 32 Ignorance 3 10 4 8 25 Discontent with government 8 8 4 4 24 Disempowerment 3 7 3 7 20 Difficulty of change 6 5 4 5 20 Unfairness and inequality 3 4 8 2 17 Materialism 0 0 2 9 11 Note: Each cell represents the number of participants that mentioned each theme. Some participants mentioned more than one theme, so there is some overlap.   144  4.3.4 Framing effects on attitudinal responses No statistically significant differences were found between conditions in any of the attitudinal questions, providing no evidence that the message frames had a significant influence on participant’s attitudes. These results were further confirmed by creating an index using the Likert-scale type attitudinal statements (items 1 to 10 in Table 4.15), which showed no significant differences between conditions. Table 4.15 depicts the means and standard deviations for each statement per condition. For offering an overview of aggregate results, Table 4.16 provides data for each statement (regardless of experimental condition).  Table 4.15 Means and standard deviations for attitudinal questions categorized by five experimental conditions.  Item or Statement Frame 1 Frame 2 Frame 3 Frame 4 Control Mean SD  Mean SD  Mean SD Mean   SD  Mean  SD 1. Continued economic growth is essential for improving people’s quality of life. 3.35 1.06 3.18 1.13 3.24 1.04 3.23 1.09 3.44 1.08 2. Economic growth is the best measure of social progress. 2.79 1.11 2.83 1.04 2.80 1.01 2.92 1.13 2.84 1.00 3. Politicians should give less priority to economic growth as a major public policy goal. 3.00 1.15 3.03 1.11 3.06 1.04 3.07 1.15 2.95 0.96 4. A 'good life' is possible without continuous economic growth. 3.46 1.06 3.42 1.08 3.42 1.14 3.56 1.11 3.33 1.03 5. In view of limited natural resources, people should figure out ways to increase quality of life while reducing overall material consumption. 4.18 0.92 4.24 0.87 4.21 0.83 4.04 0.92 4.16 0.88 6. Economic growth will not be limited by the availability of natural resources. 2.99 1.13 2.87 1.14 2.78 1.10 2.94 1.14 2.79 1.13 7. The benefits of economic growth outweigh its negative consequences. 2.77 1.09 2.72 1.10 2.86 1.04 2.79 1.15 2.69 1.01 8. We should continue growing our economy despite any large negative consequences. 2.37 1.15 2.43 1.13 2.43 1.13 2.47 1.09 2.35 1.08 9. We should eventually transition into an economic model based on reduced levels of consumption. 3.75 0.98 3.86 0.91 3.78 1.01 3.78 1.00 3.89 0.92   145  Item or Statement Frame 1 Frame 2 Frame 3 Frame 4 Control Mean SD  Mean SD  Mean SD Mean   SD  Mean  SD 10. A sustainable economic model will only be possible if we stabilize the size of our population. 3.18 1.24 3.19 1.06 3.24 1.11 3.28 1.09 3.27 1.04 11. How likely or unlikely are you to support a Canadian politician that does NOT pursue economic growth as a major policy goal? 2.27 0.89 2.32 0.84 2.37 0.90 2.36 0.88 2.36 0.81 12. Importance granted to issues:                       A. Economic growth,  31.08  15.42  31.42  14.45  31.44  16.22  29.85  15.82  30.81  16.53 B. Environmental issues, 33.87 14.53 33.88 14.07 33.63 13.89 34.27 14.97 35.36 17.07 C. Social well-being. 35.35 14.97 35.71 13.29 35.64 16.04 35.65 14.81 33.87 17.04 13. What level of economic growth do you think the government should aim for in the next 10 years?a 2.06 0.69 2.07 0.69 2.08 0.70 1.98 0.72 2.14 0.65 Note: No significant differences were found between conditions. A lower mean value indicates less agreement with the statement, while a higher mean value indicates greater agreement. Statements 1 to 10 used the following scale: Strongly disagree=1; moderately disagree=2; neither disagree nor agree=3; moderately agree=4; and, strongly agree=5.  Statement 11 used the following scale: Very unlikely=1; somewhat unlikely=2; somewhat likely=3; and, very likely=4.  Statement 12 used a numerical ranking (total points adding up to 100, indicating the importance granted to economic growth, environmental issues and social well-being).  Statement 13 used the following scale: Less than in the previous decade=1; About the same than in the previous decade =2; More than in the previous decade=3; and, I don’t know=-7.  a The ‘I don’t know’ option was excluded from these calculations.      146  Table 4.16 Relative frequencies, means and standard deviations for attitudinal statements for all participants.  Item or Statement % Distribution Mean SD Strongly disagree Moderately disagree Neither disagree nor agree Moderately agree Strongly agree 1. Continued economic growth is essential for improving people’s quality of life 5.8 19.0 27.2 36.4 11.6 3.29 1.081 2. Economic growth is the best measure of social progress 10.2 29.3 32.7 22.2 5.5 2.83 1.058 3. Politicians should give less priority to economic growth as a major public policy goal 9.5 21.6 34.1 27.1 7.7 3.02 1.084 4. A 'good life' is possible without continuous economic growth 4.9 15.7 26.4 36.9 16.1 3.44 1.086 5. In view of limited natural resources, people should figure out ways to increase quality of life while reducing overall material consumption 1.5 2.9 14.5 40.1 41.1 4.16 0.884 6. Economic growth will not be limited by the availability of natural resources 12.8 26.1 28.6 26.1 6.4 2.87 1.129 7. The benefits of economic growth outweigh its negative consequences 12.1 31.3 29.9 21.5 5.2 2.77 1.080 8. We should continue growing our economy despite any large negative consequences 24.2 32.6 25.4 13.8 4.0 2.41 1.115 9. We should eventually transition into an economic model based on reduced levels of consumption 2.8 6.0 22.9 43.8 24.4 3.81 0.965 10. A sustainable economic model will only be possible if we stabilize the size of our population 7.1 17.3 34.6 27.3 13.7 3.23 1.107 11. How likely or unlikely are you to support a Canadian politician that does NOT pursue economic growth as a major policy goal? Very unlikely Somewhat unlikely Somewhat likely Very likely   Mean SD 17.9 39.1 34.8 8.2  2.33 0.863 13. What level of economic growth do you think the government should aim for in the next 10 years? Less Same More Don't know Mean SD 17.0 42.9 22.4 17.5 2.07 0.689 12. Importance granted to issues:                         A. Economic growth 30.91 15.68 B. Environmental issues 34.22 14.94 C. Social well-being 35.24 15.26 Note: The order of items 12 and 13 has been exchanged to improve the format of the table.   147  The following figures provide an easier visualization of attitudinal items results. Each figure (between Figure 4.3 to Figure 4.15) represents each item from Table 4.16.   Figure 4.3 Percentage responses for all participants to attitudinal item #1: Continued economic growth is essential for improving people’s quality of life. Strongly disagree6%Moderately disagree19%Neither disagree nor agree27%Moderately agree36%Strongly agree12%  148   Figure 4.4 Percentage responses for all participants to attitudinal item #2: Economic growth is the best measure of social progress.  Figure 4.5 Percentage responses for all participants to attitudinal item #3: Politicians should give less priority to economic growth as a major public policy goal. Strongly disagree10%Moderately disagree29%Neither disagree nor agree33%Moderately agree22%Strongly agree6%Strongly disagree10%Moderately disagree22%Neither disagree nor agree34%Moderately agree27%Strongly agree8%  149   Figure 4.6 Percentage responses for all participants to attitudinal item #4: A 'good life' is possible without continuous economic growth.  Figure 4.7 Percentage responses for all participants to attitudinal item #5: In view of limited natural resources, people should figure out ways to increase quality of life while reducing overall material consumption. Strongly disagree5%Moderately disagree16%Neither disagree nor agree26%Moderately agree37%Strongly agree16%Strongly disagree2%Moderately disagree3%Neither disagree nor agree14%Moderately agree40%Strongly agree41%  150   Figure 4.8 Percentage responses for all participants to attitudinal item #6: Economic growth will not be limited by the availability of natural resources.  Figure 4.9 Percentage responses for all participants to attitudinal item #7: The benefits of economic growth outweigh its negative consequences. Strongly disagree13%Moderately disagree26%Neither disagree nor agree29%Moderately agree26%Strongly agree6%Strongly disagree12%Moderately disagree31%Neither disagree nor agree30%Moderately agree22%Strongly agree5%  151   Figure 4.10 Percentage responses for all participants to attitudinal item #8: We should continue growing our economy despite any large negative consequences.  Figure 4.11 Percentage responses for all participants to attitudinal item #9: We should eventually transition into an economic model based on reduced levels of consumption. Strongly disagree24%Moderately disagree33%Neither disagree nor agree25%Moderately agree14%Strongly agree4%Strongly disagree3% Moderately disagree6%Neither disagree nor agree23%Moderately agree44%Strongly agree24%  152   Figure 4.12 Percentage responses for all participants to attitudinal item #10: A sustainable economic model will only be possible if we stabilize the size of our population.  Figure 4.13 Percentage responses for all participants to item #11: How likely or unlikely are you to support a Canadian politician that does NOT pursue economic growth as a major policy goal? Strongly disagree7%Moderately disagree17%Neither disagree nor agree35%Moderately agree27%Strongly agree14%Very unlikely18%Somewhat unlikely39%Somewhat likely35%Very likely8%  153   Figure 4.14 Percentage responses for all participants to item #12: What level of economic growth do you think the government should aim for in the next 10 years?  Figure 4.15 Point assigned to economic growth, environmental issues and social well-being in order of importance (the total of all three adds to 100) by all participants. The more important an issue is, the more points it is assigned. Less17%Same43%More22%Don't know18%282930313233343536Economic growth Environmental issues Social well-being  154  4.4 Discussion This chapter investigated the effects of different message frames (related to transitioning into a new economic paradigm), on participant’s cognitive responses, emotional reaction and attitudes. In addition, it explored if framing effects are moderated by different audience segments, participant’s level of involvement with the subject matter of this research and sociodemographic factors. This discussion synthesizes the main findings, relates them to the literature on framing effects and environmental messaging, discusses some theoretical and practical implications and points to possible future research avenues. It is important to interpret the statistically significant findings of this study with caution. Due to the multiple tests performed (for exploring main and interaction effects), there could be a multiple comparisons problem (i.e. finding significant results from random data patterns). Therefore, this investigation is meant to serve as an initial exploration on this topic. Important differences emerged in the cognitive and emotional reactions of participants between the experimental conditions. With respect to cognitive responses, the environmentally-focused messages (Frames 1 and 2) generated more references to sustainability, the environment and the relationship between the economy and the environment than the messages focused on well-being (Frames 3 and 4). Likewise, well-being messages (Frames 3 and 4) generated many more comments related to overconsumption and happiness than any of the other conditions. In the control condition, thoughts related to the environment and its relationship to the economy were very prominent, even more so than in the environmentally-focused messages (Frames 1 and 2). This is not completely surprising given that this was explicitly asked to respondents in the survey question.84 These results are consistent with previous research which shows that frames do exert an influence on the focus of participant’s thoughts and associations (Price et al., 1997; Schuldt & Roh, 2014), and that messages tend to elicit frame-consistent                                                  84 The question was phrased as follows: “Please write down up to 3 thoughts or ideas that you currently have about the economy, society or the environment, and/or how they affect each other.” In hindsight, it would have been more appropriate to ask a more neutral question, so as not to prime respondents to think in one issue or another. An example of a neutral question could be: “Please write down any 3 thoughts that you currently have in your mind.”   155  thoughts. These findings support the accessibility theory of framing in that message framing makes some issues more accessible and available, while suppressing others (Gross & D’Ambrosio, 2004). Certainly, this does not preclude participants from thinking about other topics not directly referenced in the message (e.g. government, technology), although this seemed infrequent in this study.  With respect to emotional reactions, the message focused on the environment and losses (Frame 2) generated significantly less hope, more fear and more anger than the other conditions (based on the ordinal regression results). Frame 2 also generated many more references to collapse and negative outcomes than the other conditions, more mentions of fear and anger for the state of the natural environment, and more discussion about the future and future generations. Interestingly, a few respondents stated that this message (i.e. Frame 2) was intentionally generating fear (i.e. fear mongering); a comment that rarely emerged for gain-framed messages (Frames 1 or 3). Similarly, loss-framed messages (Frame 2 and 4) generated significantly less hope, more fear and more anger than gain-framed messages (Frames 1 and 3). These results support the finding that message framing can indeed affect emotional reaction (Gross & D’Ambrosio, 2004; Hornsey & Fielding, 2016; Myers et al., 2012). These findings are consistent with other environmental research that shows that loss-framed messages generate more fear and anger (Cheng et al., 2011; Spence & Pidgeon, 2010), while gain-framed messages can generate more positive attitudes (Spence & Pidgeon, 2010) like joy and contentment (Cheng et al., 2011). Research on the effectiveness of loss- versus gain-framed messages are mixed, as findings vary according to subject matter, type of behaviour, individual differences and other related variables (Cesario et al., 2013; Cheng et al., 2011; Rothman & Salovey, 1997). Some studies have found negative frames to be more effective than positive ones (Davis, 1995), while other studies have found the opposite (Van de Velde et al., 2010).  Provoking negative emotions in an audience can generate desirable, as well as harmful, outcomes. Previous research shows that, using fear in messaging may be useful for attracting an audience’s attention, increasing their level of concern (Hine & Gifford,   156  1991), the importance of the topic (O’Neill & Nicholson-Cole, 2009) and for temporarily influencing attitudes (Hastings & Stead, 2004; Moser & Dilling, 2004) and behavioural intentions (Feldman & Hart, 2015). However, research has shown that fear is not associated with more support for climate policies (Smith & Leiserowitz, 2014), that it can be ineffective at generating real engagement and action (O’Neill & Nicholson-Cole, 2009) and, consequently, may be counterproductive and backfire (Feinberg & Willer, 2011; Moser & Dilling, 2004). Results from this study did not provide evidence that negative emotions ultimately influenced attitudes and behavioural intention,85 in that Frame 2 (which provoked more fear and anger) did not generate significantly different attitudinal and behavioural responses in comparison with the other conditions. Having said that, these data should be interpreted with caution, as effects of message framing on emotions (although statistically significant), may have not been large enough to make any significant influence in attitudes. Moreover, this study was not explicitly designed for testing the main effects of provoking certain emotions (i.e. treating emotional reaction as an independent or mediator variable). Rather, emotional reaction was treated as a dependent variable in this study. Previous research offers some guidance as to when using fear appeals may be disadvantageous. For instance, it has been shown that, if fear appeals are used continuously on the same topic, people may become desensitized and habituated to them (see review in Hastings & Stead, 2004). In this sense, using fear appeals in the early stages of communication of ecological economics to unfamiliar audiences, may be useful for attracting their attention and increasing their level of concern. However, due to the reasoning mentioned above it may not suffice as a long-term strategy for generating genuine engagement and action towards transforming the current paradigm. Furthermore, it is crucial to consider that the effectiveness of fear appeals for motivating people into action depends on certain conditions (Stern, 2012), like self-efficacy. If there is little feeling of control (i.e. low self-efficacy) over a threat or problem, the individual will likely reduce his or her fear with maladaptive reactions like apathy, denial, fatalism or                                                  85 Behavioural intention was measured in terms of participant’s voting likelihood (see statement 11 in Table 4.15).   157  helplessness (Hastings & Stead, 2004; O’Neill & Nicholson-Cole, 2009). In this sense, Stevenson and Peterson (2016) highlight the importance of communicating the seriousness of the threat, but doing so in a way that avoids despair and, instead, encourages a sense of agency. This is crucial advice for the communication of ecological economics, in that, while pointing out the great challenges facing society, it may be also crucial to transmit a sense of hope about the feasibility of effecting change.  Although anger and fear are both considered negative-valenced emotions, research has shown that they influence information-processing and risk perception in different ways (Lerner & Keltner, 2001). For instance, under certain conditions, angry individuals will make more optimistic assessments than fearful people, who are consistently more pessimistic (Lerner & Keltner, 2001). Anger has been identified as a powerful, influential and even functional emotion, so long as it is conveyed in socially appropriate ways (Angie et al., 2011). For instance, anger can evoke feelings of individual power and instigate people to act against the source of the emotion (i.e. activation response) (Angie et al., 2011; Feldman & Hart, 2015). In this study, societal inaction and environmental issues were the most mentioned causes of anger in the environmentally-focused messages (Frames 1 and 2). Accordingly, some individuals that received these messages may be more motivated to act against these sources in order to reduce their irritation. This could produce positive outcomes for ecological economics like inducing more engagement and action in relation to environmental issues. Nonetheless, generating anger through messaging can be a double-edge sword and may backfire, as anger can also limit people from incorporating new knowledge, relies more on heuristic processing and can make people less attuned to their environment (Druckman & McDermott, 2008). In this study, the loss-framed messages, especially the one focused on the environment (Frame 2), generated lower levels of hope than the other messages. At the individual level, hope has been associated with positive outcomes like academic success and good health (Snyder et al., 2005). On a broader level, research has shown that hopeful individuals are more likely to support climate policies (Smith & Leiserowitz, 2014) and engage in intended climate activism (Feldman & Hart, 2015), and that hope is a good   158  predictor of pro-environmental behaviour among adolescents (Stevenson & Peterson, 2016). Consequently, hope has been proposed as a powerful promoter of action – if aligned with high self-efficacy (Smith & Leiserowitz, 2014) – and many authors have called for a movement away from fear-based communication to more hopeful messages about climate change and other environmental issues (Diaz-Meneses, 2010; Stevenson & Peterson, 2016). Nonetheless, caution should be exercised in promoting hope as a communication strategy, as research has shown that hopeful messages can, in some instances, be less effective than pessimistic and sad messages (e.g. in increasing motivation for climate change mitigation (Hornsey & Fielding, 2016) and in promoting information seeking and policy support for a marine disease (Lu, 2016)). Hornsey and Fielding (2016) explain that higher levels of hope decreased risk perception, which, in turn, reduced concern on the topic. Interestingly, this same study did not find evidence of a strong correlation between hope and feelings of efficacy (Hornsey & Fielding, 2016). Indeed, more research is needed in order to suggest clear pathways in this regard, as the effects of some emotional appeals (like sadness or hope) have been understudied in environmental communication (Lu, 2016) and not at all studied in ecological economics.  It is interesting to note that the well-being frame focused on gains (Frame 3) generated more references that explicitly mentioned participant’s agreement with the message than any other condition (see Table 4.5) and, although results were not statistically significant (in comparison with Frames 1 and 4), it generated the most hope, the least fear and the least anger of all messages (see Figure 4.2). Although not conclusive in any way, this may indicate some of the positive aspects of moving beyond traditional fear-based messages to bring other issues to the fore that could resonate with a broader audience, like quality of life, well-being and happiness. This is not to say that messages related to the environment and natural resources should not be used, but rather, that people interested in spreading the ideas and ethos of ecological economics and the environmental movement, could have a messaging strategy better designed to connect with a larger and far-reaching set of issues. For example, in their study of discourses in Canadian environmental NGOs, Haluza-DeLay and Fernhout (2011) conclude that the “environmentalist” frame is still very prevalent and that it only marginally includes issues   159  of social justice and equity. Similarly, climate change has also been often framed as an environmental issue, although this is starting to change (Myers et al., 2012), with some authors encouraging a re-framing strategy (Bain et al., 2012; Nisbet, 2009; Stern, 2012). In this sense, Bain et al. (2012) examined the potential benefits of re-framing climate action in terms of positive societal outcomes and show that it may increase pro-environmental action intentions, especially among certain segments of the public. This advice is highly pertinent for the dissemination of ecological economics, in that this transdiscipline is as much about ecology and sustainability, as about social issues, just distribution and other matters. Thus, incorporating these others topics in relevant messages and discourses may resonate with a broader audience. Results of this study show that framing effects were moderated by audience segment and gender. Participants in Cluster 3 (who aligned more with ecological attitudes) were the more susceptible to framing effects influencing their sense of hope; for them, the environmental message focused on losses (Frame 2) resulted in lower feelings of hope. Although framing effects on hope were also evidenced for Cluster 2, they occurred less prominently; while these effects were not significant at all for Cluster 3. These results are coherent in the sense that members of Cluster 3 are likely more aware and concerned about environmental issues, making them more susceptible to messages that are consistent with their own pre-existing mental schemas. On the other hand, it makes sense that members of Clusters 1 and 2 would be less affected in that they are not as worried about this issue. These results also reinforce the idea that, although fearful frames may be useful for drawing people’s attention, they could backfire with individuals who are already concerned, perhaps inducing some level of despair. Nonetheless, more research is needed to establish any definitive conclusions and recommendations in this regard. Curiously, audience segment was not a significant moderating factor for the dependent variables fear and anger. This is interesting in as much as pre-existing attitudes have been shown to moderate framing effects (Rothman & Salovey, 1997) and ‘knowing your audience’ is often a crucial advice in messaging strategies (Hine, Phillips, et al., 2013). Nonetheless, further research is needed in order to dig deeper about the role of audience   160  segments in moderating different post growth messages, possibly including a wider diversity of messages, emotions and variables. With respect to gender, Frame 2 generated higher levels of fear, but only among women. Research on the impact of gender in moderating framing effects is mixed. Some studies have shown that men are less vulnerable to framing (Druckman & McDermott, 2008; Ellingsen, Johannesson, Mollerstrom, & Munkhammar, 2013; Fagley & Miller, 1997; Van de Velde et al., 2010), while others have found the opposite (Gifford & Comeau, 2011). Some results vary according to the frame employed (Lockwood, 2011), while others have found no significant associations (Levin, Gaeth, Schreiber, & Lauriola, 2002). Our study is not conclusive in this issue, as it offers mixed results in the sense that gender was a significant moderator for feelings of fear, but it was not significant for other emotions like anger and hope. Findings in this study also show that audience segments, issue involvement (related to environmental and economic concern), gender and political identification, were important covariates associated with feelings of hope, fear and anger. Audience segments were significant for feelings of hope, as members of Cluster 1 (expansionist segment) reported being more hopeful than those in Cluster 3 (ecological segment). This optimism could be due to the greater faith of this group concerning human ingenuity and technology (see Chapter 3). Higher environmental and economic concerns were associated with more reported fear and anger. This was an expected result, as concern has been shown to be significantly associated with fear (Smith & Leiserowitz, 2014). Women reported higher levels of fear than men, which is consistent with literature that shows that women report greater vulnerability to anxiety and number of fears than men (see review in McLean & Anderson, 2009). Finally, political identification was associated with feelings of hope and anger, as those identified with the Conservative Party were more hopeful than those identified with the Green Party and were angrier than those identified with the Liberal Party and NDP.  The results of this study may also point to the limits of message framing, as no significant effects were found on attitudinal items. One of the issues that could be at play here is the   161  possibility that people had not thought much about the issues of this research beforehand, especially in relation to questioning economic growth as a main policy goal. As mentioned earlier, framing works by making some issues more salient and accessible than others. However, “[…] framing effects are, in part, derived from already held stores of knowledge made accessible based on a message’s perspective. If such knowledge stores, or schemas, do not exist or are poorly developed, we cannot reasonably expect strong framing effects to occur” (Nabi, 2003, p. 230). Moreover, the idea of a non-growing economy is such an unconventional one, that regardless of message it will be likely challenging changing people’s current attitudes on this issue with such a short intervention. In addition, the framing experiments in this research had other limitations. For instance, all participants were presented with a battery of initial attitudinal questions (see survey sections 1 and 2 in Appendix F) before they were exposed to the framed message. These initial questions served to identify the audience segments described in Chapter 3. However, it cannot be ruled out that, by asking these questions, participants (including those in the control condition) could have been influenced in their thinking about the subject matter of this research, thus confounding potential framing effects. For future research, it would be ideal to first present the frames, thus avoiding any of these potential confounding effects.  Furthermore, framing experiments in general tend to have other limitations. For instance, they tend to focus on one dimension per frame, whereas actual conditions (e.g. media, news articles) tend to offer multiple dimensions per frame and even present counter frames (Bolsen, Druckman, & Cook, 2013). The long-term influence of framing is also uncertain as most research only measures immediate effects and some data suggest that impacts may weaken over time (Druckman & Nelson, 2003). Some authors have even suggested that psychological studies may overstate the impact of framing, as these often draw on individuals as the unit of analysis (Scheufele & Nisbet, 2007). Research indicates that effects may be even weaker in group settings and in more ‘real-life’ situations than in controlled and context-free environments (Druckman, 2004; Nelson, Bryner, & Carnahan, 2011; Price et al., 2005). In addition, the possibility exists of having a ‘file drawer problem’, where mostly positive results are published. Even if this were not   162  the case, many framing experiments often show significant (but small) differences between frames with some even showing no significant differences (Bernauer & McGrath, 2016).  Regardless of framing effects, it is important to note that many participants agreed with the statements related to reducing consumption (see Figures 4.7 and 4.11). For instance, more than 80% of respondents moderately or strongly agreed that, in view of limited natural resources, people should try to figure out ways to increase quality of life while reducing overall material consumption. Actually, less than 5% of respondents disagreed with this statement. Similarly, almost 70% moderately or strongly agreed that we should eventually transition into an economic model based on reduced levels of consumption, while only 9% disagreed with this. The answers to these statements reveal a high level of support, at least ‘in theory’, for the basic tenets of the steady state economy and sustainable degrowth.  The items that explored participants’ views on economic growth were more nuanced. For instance, although 53% agreed that a good life is possible without continuous economic growth (see Figure 4.6), 48% also agreed that economic growth is essential for improving people’s quality of life (see Figure 4.3). Nonetheless, less than 20% of participants moderately or strongly agreed with the idea that we should continue growing our economies despite any large negative consequences (see Figure 4.11), indicating that many disagree with the goal of ‘growth at all costs’. Interestingly, the items that asked about economic growth being the best measure of social progress (see Figure 4.4) and politicians giving less priority to economic growth as a main policy goal (see Figure 4.5) had a large proportion of answers in the middle of the scale, with 33% and 34% of respondents, respectively, neither disagreeing nor agreeing with these statements. This may indicate that many people are ambivalent, unsure and maybe even indifferent about economic growth being a central policy goal and measure of progress.   Remarkably, 43% of participants reported being somewhat or very likely to vote for a politician that does not pursue economic growth as a main policy goal (see Figure 4.13). These findings suggest that, although economic growth is often linked with initial   163  positive associations, it is not an inevitable and inescapable goal. Indeed, it may even be possible to move away from it as a main policy priority if greater public support and awareness is built around this endeavour (especially considering the large segment of people that is undecided in relation to this, as explained earlier). These findings may reflect a disconnect between the dominant political discourse in Canada (frequently focused on economic growth as a central policy goal) and people’s perceptions about the actual need of focusing so heavily on this one indicator. Having said that, the findings presented here underscore the need for more cross-sectional and longitudinal studies on public opinion about these issues, to more accurately assess the political feasibility of moving away from the growth paradigm. If supported by further research, these initial findings could be a significant ‘wake-up call’ to politicians in their enduring framing of continued economic growth as a sine qua non of political discourse.   This investigation is an initial attempt at exploring questions around framing of some concepts of ecological economics (as conceptualized by Daly and Farley, 2011 and others in the field), thus many research gaps remain. Future research could systematically explore the most prevalent frames and discourses being used by individuals and organizations involved in the communication of ecological and post-growth economics, and assess their cognitive, emotional, attitudinal and behavioural effects, if any. This may allow the development of more effective messaging strategies. Future studies are encouraged to include other emotions not explored in this research, such as concern, sadness and enthusiasm, and evaluate more directly the effects of experiencing these emotions. For instance, worry and concern have been shown to be important predictors of climate policy support (Smith & Leiserowitz, 2014) and pro-environmental behaviour intentions (Stevenson & Peterson, 2016). Finally, further research could explore with more depth the moderating influence of different audience segments and sociodemographic factors on framing effects. The findings in this chapter may contribute to the communication and education of ecological and post-growth economics, by providing some data about the state of public opinion in Canada and by showing that some message frames matter at the time of   164  influencing cognitive associations and emotional reactions. Although this research did not provide evidence that frames were successful at influencing attitudes or voting intention, this does not preclude that cognitive and emotional reactions may have an influence on information processing and judgements (e.g. risk perception), as this has been evidenced in previous studies (Angie et al., 2011; Lu & Schuldt, 2015; Smith & Leiserowitz, 2014). As mentioned earlier, the novelty of the topic (i.e. post-growth communications) and the primacy of growth as something positive, will likely make it difficult to change attitudes with such a short experimental intervention. Nonetheless, communicators should be cognizant and aware about the feelings and affective reactions that they can cause in their audiences and possibly even evaluate the ethical and moral implications of this. As discussed here, fear-based communication may initially work to draw people’s attention and increase concern, but, due to the drawbacks mentioned, the effectiveness of fear appeals in the longer term appear to be doubtful. Moreover, invoking other emotions may prove to be more fruitful, but more research is needed in this regard. In this sense, this investigation has opened many more questions than what it was able to answer, thus future research in the areas mentioned earlier is warranted.   4.5 Conclusions Moving to a post-growth economic paradigm will require research to identify compelling ways in which key concepts of this transition can be more effectively communicated to the public. This chapter explored the effects of different message frames on people’s thoughts, emotions and attitudes, and examined potential moderating variables. Results show that the frames influenced participants’ cognitive and emotional reactions. Specifically, the messages focused on the environment (Frames 1 and 2) primed participants to think about natural resources, the environment and sustainability, while the messages focused on well-being (Frames 3 and 4) generated more references about overconsumption, materialism and happiness, thus confirming that the messages tend to generate frame-consistent thoughts. Regarding emotional reaction, the message that focused on the environment and losses (Frame 2) generated significantly less hope, more fear and more anger than the other conditions. The results for hope were moderated by   165  audience segment, as effects were more prominent for members of Cluster 3 (i.e. more ecological segment) and were not significant for participants in Cluster 1 (i.e. more expansionist segment). The results for fear were moderated by gender, as effects were only significant for women. When analyzed in combination, loss-framed messages (Frames 2 and 4) generated more negative emotions than the gain-framed messages (Frames 1 and 3). This study did not find evidence to suggest that the message frames influenced participant’s attitudes. Future research could replicate this study and confirm the findings presented here. Also, different frames and narratives could be further explored assessing their impacts and effectiveness. Ultimately, this exploratory study aims to contribute, even if indirectly, to inform the development of more deliberate and strategic communications and narratives about the transition to a sustainable economic paradigm that does not rely on indefinite growth.      166  5. Conclusions Transforming the current economic paradigm is crucial in the transition to sustainability and the transdiscipline of ecological economics provides valuable insights and pathways for embarking on this transition. However, understanding people’s thinking about economic growth, sustainability and the environment may be necessary for eventually moving society away from unsustainable worldviews. Furthermore, challenging the status quo of limitless economic growth requires research to identify ways in which post-growth concepts can be more effectively conveyed to the public. In this context, this doctoral dissertation aims to contribute to the effective communication of some tenets of ecological economics (as conceptualized by Daly and Farley, 2011 and others in the field) by exploring the realms of public opinion and message delivery.  This research had three main objectives: first, to describe and bring to light mental models on how people perceive the relationship between economic growth, society and the environment, and to explore which other variables (e.g. cultural cognition, sociodemographics) relate most strongly with the identified mental models; second, to explore the prevalence of expansionist and ecological attitudes among the general population, and to segment the audience based on these attitudes (a secondary aim was to explore the correlation between each segment and other variables, such as sociodemographics and issue involvement); and third, to test the effects of different message frames (related to the transition to economies not primarily centered on growth) on people’s cognitive responses, emotional reactions and attitudes, as well as to explore potential moderating variables of framing effects (e.g. audience segments, issue involvement, sociodemographics). These three objectives were highly related and interdependent. The findings from the first objective (described in Chapter 2) informed the development of data collection methods for addressing the second objective. In addition, the second objective (described in Chapter 3) tested if the mental models identified from the first objective could be replicated in a larger sample. Finally, one of the moderating variables (i.e. audience   167  segments) deployed for addressing the third objective (described in Chapter 4) was obtained through analyzing findings from the second objective.   This mixed methods study was implemented in Canada. For tackling the first objective of exploring people’s mental models, 60 semi-structured interviews were carried out in the Province of British Columbia (BC). Data was inductively and deductively analyzed with NVivo 10 qualitative software. Specifically, cluster analysis was used for uncovering the distinct mental models. For addressing the second objective on exploring people’s attitudes and segmenting the audience, more than 1,000 Canadian residents participated in an online survey. Descriptive and correlational analyses ware carried out with IBM SPSS Statistics 23 software and the different audience segments were uncovered with Latent Class Analysis (LCA) using Latent Gold 5.1. For dealing with the third objective on message framing effects, data were obtained with the same online survey using an experimental design approach. Data were analyzed with IBM SPSS Statistics 23 software and NVivo 10 qualitative software. This concluding chapter describes the main results found for each research objective, elucidates how they relate to the relevant scientific literature, synthesizes these findings and discusses their theoretical and practical significance. In addition, it describes the strengths and limitations of this doctoral research and points to future research directions.   5.1 Summary of Empirical Findings 5.1.1 Objective 1: Mental models regarding the economy, society and the environment  Based on the interviews carried out in BC, five clusters of participants were described (although six were identified), each representing a distinct mental model of relationships between the economy, society and the environment. The five mental models, labeled A (The Expansionist), B (The Hesitant), C (The Indifferent), D (The Green) and E (The Ecological), sat in a spectrum of views anchored to an expansionist or to an ecological worldview. On one side of the spectrum, members of Cluster A (18%) embraced a more expansionist perspective that the economy was unrestrained from natural limits and   168  expressed faith and optimism that human ingenuity and technology could overcome any constraints to economic growth. On the other side of the spectrum, participants in Cluster E (15%) embraced an ecological perspective, with limits to growth being fully acknowledged due to resource constraints and ecological reasons. This was the only group that gave clear indications of being more eco-centric and concerned about the well-being of other non-human species. Clusters B (28%), C (12%) and D (13%) sat in the middle of the spectrum. Participants in Cluster B were often ambivalent about the existence of limits to growth. Members of Cluster C believed in limits, but the primary reasoning had to do with debt and other explanations not related to natural resources. Members of Cluster D believed in constraints to economic growth mainly due to natural resource scarcity. In this sense, perspectives about limits to economic growth and the corresponding reasoning related to this, marked the most remarkable differences between each of the mental models. Sociodemographic characteristics, like gender and political affiliation, were related with the mental models. A larger proportion of men and participants who identified with the Conservative Party were associated with expansionist ideas (as reflected in the membership of Cluster A), while a larger proportion of participants who identified with the Green Party were more associated with ecological ideas (as reflected in the membership of Cluster E). This supports existing literature about the relationship between sociodemographic factors and environmental views (Hunter et al., 2004; Shafer, 2006; Zelezny et al., 2000). However, contrasting what previous studies have suggested about the relation between cultural cognition of risk and environmental views, this study did not find evidence of a significant correlation between the mental models and cultural cognition. These findings confirm that the expansionist worldview is still present in society today, although results show that most respondents did grasp, albeit to different degrees, the idea of ecological limits, possibly indicating that perceptions could be shifting away from expansionism. The mental models described here do resemble the worldviews upheld by the Dominant Social Paradigm (DSP), especially those expressed in Cluster A and, to a   169  lesser extent, in Cluster B, and the New Ecological Paradigm (NEP), especially those expressed in Clusters E and D (Dunlap et al., 2000; Shafer, 2006). In addition, the results are somewhat comparable to the audience segments identified in climate change communication research (Hine, Reser, et al., 2013; Maibach et al., 2011), possibly indicating that the mental models described in this investigation – although focused more closely on economic issues – are indeed similar to segments emerging from other environmental and sustainability studies. Nonetheless, it is difficult to establish more useful comparisons with other studies due to the novelty of this research.  Regardless of cluster, most participants identified similar strengths and challenges related some of the ideas and tenets of ecological economics and the steady state economy. Many thought that the concepts were sound and logical, while most criticisms revolved around issues of implementation and were not directly targeted at the ideas per se; except that participants in Cluster A were much less open to these (the ideas were considered to underestimate human ingenuity and technological progress). These findings suggest that, for most people, the basic worldview of ecological economics seems coherent and reasonable, and they point to the importance of focusing on positive and achievable aspects of the transition, rather than on the massive implementation challenges and difficulties laying ahead. 5.1.2 Objective 2: Prevalence of expansionist and ecological attitudes in the Canadian population and audience segmentation The second objective of this research expanded and added to the findings above. Results show that most of the surveyed participants held certain ecological attitudes, such as recognizing that humans are a part of nature and depend on it for our survival. In addition, most reported being highly concerned for the state of the natural environment, as well as for the Canadian economy. These findings are consistent with literature that points to the high importance granted to the environment by Canadians (Environics Institute, 2012; Pyman & Pammett, 2010; World Values Survey, 2015). Although many respondents recognized the eventual resource limitations of indefinite economic expansion, a majority still believed that growth is good, complementing existing   170  literature that argues that economic growth is largely perceived as natural, unquestionable and inherently positive (Gustafsson, 2013; Dryzek, 2013). Moreover, many respondents did not see the goals of environmental sustainability and economic growth as contradictory, confirming that many people may see both goals as achievable. This stage of research identified three different segments among the sample of more than 1,000 Canadian residents. Participants fell into three clusters based on their responses to attitudinal questions measuring expansionist and ecological perspectives. Cluster 1 (41.1%) expressed a high degree of agreement with all expansionist items; they had the most positive attitudes towards economic growth, technology, human ingenuity and the possibilities of indefinite economic expansion. Interestingly, members of this group also agreed with ecological statements; for example, they recognized that humans depend on the environment for our survival and that we are as much a part of nature as other animals. Participants in Cluster 2 (36.3%) were often in the middle of the spectrum for all survey items and did not express strong opinions one way or another about any particular issue. Participants in Cluster 3 (22.6%) expressed the least agreement with expansionist items and the highest agreement with ecological ones; they strongly agreed with the notion of human unsustainability and strongly disagreed with the possibilities of indefinite growth. Moreover, they reported significantly greater concern for the environment (indicating higher issue involvement) than participants in the other two clusters. Establishing accurate comparisons with other studies is difficult as little research, to my knowledge, has attempted an audience segmentation exercise based specifically on people’s views about economic growth and the environment. Nonetheless, like the previous objective, some comparisons can be drawn with research on climate change communication (Hine et al., 2016; Maibach et al., 2011). For instance, Hine et al. (2016) uncovered three comparable segments in Australia, which varied in terms of their beliefs and concern regarding climate change.  Gender, political identification, education and income were significantly associated with particular segments, supporting research that has found similar correlations (Blake et al.,   171  1997; Drews & van den Bergh, 2016; Leiserowitz et al., 2007; Maibach et al., 2011; Pyman & Pammett, 2010). Specifically, Cluster 3 comprised a larger proportion of women, a lower proportion of respondents identified with the Conservative Party (and, therefore, a larger share of people identified with the Green Party), and higher levels of education compared to the other groups. In addition, it reported significantly higher income levels compared to Cluster 2.  Although the segments for the first (Chapter 2) and second objectives (Chapter 3) did not fully replicate, there are some similarities. Cluster 1 resembles Clusters A and B in the prevalence of more expansionist perspectives. Cluster 2 resembles Cluster C in the lack of strong opinions about the subject matter of this research. Cluster 3 is similar to Clusters D and E in their alignment with an ecological worldview. The inconsistencies in findings may be due to differences in the samples and methodologies employed, indicating the importance of experimental replication and of tackling research questions with different methodological approaches. The findings presented here may represent a new way of identifying audiences and differentiating the public based on the spectrum of expansionist and ecological tendencies, and provide baseline data that could be useful for tracking how these attitudes change and evolve across time.  5.1.3 Objective 3: Message framing effects about transitioning to economies not centered on growth  Results for the third and last objective show that, although there was no evidence that the four different message frames (focused on transitioning to economies not primarily focused on growth) affected significantly participant’s attitudes, they did influence their cognitive and emotional responses. Environmentally-focused messages (Frames 1 and 2) generated more references about the environment and sustainability, while messages focused on well-being (Frames 3 and 4) generated many more comments related to overconsumption and happiness. These results provide evidence for the claim that framing influences the focus of participants’ thoughts and associations (Price et al., 1997; Schuldt & Roh, 2014) and makes some issues more accessible and available than others (Gross & D’Ambrosio, 2004).   172  Regarding emotional reaction, loss framed-messages (Frame 2 and 4) generated significantly less hope, more fear and more anger than gain-framed messages (Frames 1 and 3). More specifically, the message focused on the environment and losses (Frame 2) generated significantly less hope, more fear and more anger than the other conditions. The message focused on well-being and gains (Frame 3) generated the opposite reactions (more hope, less fear and less anger), although differences were not statistically significant. These findings confirm that message framing can, indeed, affect emotional reaction (Gross & D’Ambrosio, 2004; Hornsey & Fielding, 2016; Myers et al., 2012) and that loss-framed messages often generate more negative-valenced emotions than gain-framed messages (Cheng et al., 2011; Spence & Pidgeon, 2010).  Audience segment and gender were significant moderators of framing effects. The frames, especially the environmental message focused on losses (Frame 2), influenced Cluster’s 3 reported level of hope to a greater extent than it did for members of Cluster 2. Moreover, no evidence of significant framing effects for hope were found for participants in Cluster 1. It is surprising that audience segment was not a significant moderating variable for fear and anger, as previous research has shown the moderating influence of pre-existing attitudes (Bain et al., 2012; Rothman & Salovey, 1997). With respect to gender, Frame 2 generated higher levels of fear, but only among women. This supports some research that shows that men are less vulnerable to framing (Druckman & McDermott, 2008; Ellingsen et al., 2013; Fagley & Miller, 1997; Van de Velde et al., 2010), although these findings are not very conclusive in that gender was not a significant moderator for other emotions like anger and hope.  Responses to attitudinal questions (regardless of experimental condition) show that a majority of respondents agree with statements related to reducing consumption. Participants views on economic growth were more nuanced, although surprisingly, more than 40% reported some likelihood of voting for a politician that does not pursue growth as a main policy goal. These findings are important in that they shed light about the state of public opinion regarding growth as a main policy goal in Canada and show that economic growth may not be an inevitable and unescapable goal. Moreover, these data   173  may point to some potential disconnect between prevalent political discourses (heavily centered on economic growth as a main policy priority) and people’s actual thinking and prioritization of this goal.  5.2 Synthesis of Empirical Findings, Theoretical Significance and Practical Implications This research has uncovered different mental models on how people perceive the relationship between the economy, society and the environment. Although participants in this study seemed to possess some aspects of an ecological worldview, a large part of the sampled population remained with expansionist traits. In addition, this study showed that different message frames about the transition to economies not primarily centered on growth influenced participants’ cognitive responses and emotional reactions, but there was no evidence that they significantly influence attitudes about this topic.  Results from this study suggest that many people have not thought in much depth about the relationship between economic growth and the environment; more specifically, about limits to growth. The interviews in Chapter 2 showed that multiple participants were ambivalent about their stances on these topics, with some admitting that they had never thought about these issues beforehand. In Chapter 3, a large proportion of respondents (more than 35%) were classified into the ambivalent segment (i.e. Cluster 2), suggesting that many people have mixed perspectives and possibly indicating that little thought has been granted to these issues. This is not surprising as it has been continuously pointed out throughout this dissertation that economic growth is a ubiquitous goal that is rarely ever questioned either by politicians, businesspeople or the general population; it is assumed to be inherently good and positive. Another important result that emerged from Chapter 2 is that, when people were exposed to the basic principles of ecological economics (such as that the economy is a subsystem of the ecosphere) and confronted with the idea that indefinite economic growth is unsustainable, these ideas became logical and self-evident for many (but only after exposure). This data may suggest that mental models on continuous growth are not   174  inevitable and irreversible, at least among a significant proportion of the public sampled here. This may also be a prime opportune time to renew the discussion about ecological limits, as there is increasing evidence that, despite all of our ingenuity and technological wizardry, limits are becoming more apparent and proximate (e.g. crossing planetary boundaries like climate change, biodiversity loss and others). Moreover, the findings from Chapter 4 indicate that a significant proportion of the sampled population reported being open to the idea of supporting a model not based on economic growth as a primary goal. Paradigms are slow to change, but they do shift by continuously pointing out the weaknesses, contradictions and anomalies that we see in current paradigms (Donella Meadows, 1997). Thus, the issue of limits to growth could be more broadly and consistently discussed in the public and in political and academic circles, so as to bring this issue to the fore and encourage more people to question the growth imperative.  Additionally, the findings from Chapter 4 suggest that diversifying sustainability and post-growth messages beyond environmental and resource issues to more frequently include matters of personal relevance such as quality of life and well-being, may resonate with a broader population and encourage a more widespread diffusion of these ideas. This suggestion emerges from the fact that the message focused on well-being and gains (Frame 3) generated the most explicit statements of agreement of any frame (revealed in the answers to open-ended questions). Moreover, the findings presented here show that many participants agree with moving towards a system that promote lower levels of consumption. In this sense, concepts like ‘sustainable well-being’ and ‘genuine progress indicator’ (Kubiszewski et al., 2013) are steps in the right direction.  Regardless of the theoretical and practical importance granted to frames and narratives, this investigation did not provide a strong evidence that message frames were relevant at the time of influencing attitudes. One of the reasons for this finding could be the novelty of the topic of this research, as the idea of an economy not centered primarily on growth is so unconventional that, regardless of message, it will likely be challenging to change people’s existing attitudes on this issue (especially with a short experimental intervention like the message frames employed here). People’s attitudes about economic growth may   175  change, but possibly only through the continuous and long-term exposure to messages countering this goal and suggesting alternatives to growth (thus, battling the hypocognition that currently exists about this issue). Nonetheless, cognitive and emotional reactions should not be dismissed, as previous research has shown the importance of emotions in moderating information processing and risk assessments. Moreover, taking into account emotional reactions could be especially relevant in current times, where facts and scientific evidence do not seem to influence policy as much as before, especially in countries like the United States.  This doctoral dissertation makes multiple theoretical contributions. First, it provides data on topics that have been little explored, such as describing mental models on people’s views of economic growth and the environment, and identifying audience segments based on expansionist and ecological attitudes. These data provide new insights into people’s thought processes and attitudes regarding the current economic model and may shed light on the state of public opinion about the possibilities of transitioning to a different economic paradigm. This dissertation brings attention to this line of research and could lead to inspire future related studies (see Section 5.4). In addition, this investigation tests an existing theory in a new setting, as it is uses message framing theory in the context of ecological economics. This investigation provided some support for framing theory in so much as the evidence supported framing effects for cognitive associations and emotional reaction, but not for attitudinal responses. However, as mentioned earlier, possibly due to the novelty of the research approach or aspects of experimental design, this study is not strongly conclusive with respect to framing effects. In this sense, it has created many more questions than it has answered, underscoring the need for more research. Finally, an important contribution of this empirical study is that it linked theory between two forms of heterodox economics: behavioural and ecological economics. The transition to sustainability will require using foundations and models from both fields. In addition to the theoretical relevance above, this study has some practical significance. This research can be useful for policy-makers, communicators, educators and other individuals interested in the transition to a new economic paradigm (not based on   176  indefinite growth), as understanding people’s attitudes and mental models about these issues may be a crucial step in delivering effective communications. For instance, by designing post-growth communications targeted for each specific audience, some segments of the public may be more successfully engaged. In this sense, diversifying communications away from fear-based messaging may contribute to eliciting more positive emotions (such as hope and empowerment), which could be especially pertinent for segments of the public that are already highly aware and concerned. In addition, these findings confirm the importance of considering the cognitive and emotional reactions provoked in the audience by communicators, as emotions can affect information processing and judgements. Also, being cognizant of the audience’ characteristics, such as gender composition and pre-existing views, is essential, as it will likely influence how a message is perceived. Ultimately, it is hoped that the findings presented in this research will aid practitioners in improving the rhetorics of some concepts of ecological economics and their diffusion to broader segments of the population. 5.3 Strengths and Limitations of this Research This research presents a novel way of identifying and differentiating audiences based on the spectrum of expansionist and ecological tendencies, and is an initial attempt at studying message framing in relation to some concepts of ecological economics. Exploring these research questions from different angles using a mixed methods approach provides a robust means by which results could be compared, contrasted and triangulated. Despite these advantages, this research has multiple limitations. First, the results presented here should not be thought of as representative of the Canadian or British Columbian populations in a statistical sense due to the non-probability strategies used for sampling (Baker et al., 2010). Also, some self-selection bias could have occurred as people who were more interested and involved in issues relating to the environment and the economy may have been more likely to participate in the interviews and survey. Second, regarding the identification of audience segments, opinions shift with time and circumstances; therefore, it is likely that the compositions of the identified clusters would   177  change if other studies with new methodologies and new variables were implemented (see Leiserowitz et al., 2007; Maibach et al., 2011). Third, the results from this study also point to the limits of framing effects and framing experiments in particular. In the specific case of this research, it is possible that the initial attitudinal questions presented before the message frames (survey sections 1 and 2 in Appendix F) could have generated some confounding effects which influenced participants’ thinking about these issues early on. Thus, as mentioned in Chapter 4, for future research, it would be ideal to present the frames prior to any attitudinal questioning, thus avoiding any cofounding effects. Furthermore, framing experiments tend to have low external validity because they often focus on one dimension per frame, whereas real conditions (e.g. media, news articles) tend to offer multiple dimensions per frame and even present counter frames (Bolsen et al., 2013). They also tend to focus on the individual in an isolated or controlled context, whereas decision-making often occurs in groups or social settings. The long-term influence of framing is also uncertain, as most studies only measure immediate effects, and some data suggest that impacts may weaken over time (Druckman & Nelson, 2003). Fourth, it is important to be aware that many framing experiments often show significant (but small) differences between frames and some even show no significant differences at all (Bernauer & McGrath, 2016). Finally, this investigation is exploratory in the sense that it is only scratching the surface of people’s thinking and reactions on the complex topic of public perceptions about post-growth economics. By no means does this study aspire to provide a definitive summary of people’s attitudes and actual behaviours towards these issues, nor does it intend to change these in the long-term.  5.4 Future Research Directions Owing to the novelty and exploratory nature of this investigation, many future research avenues are warranted. Specifically, further studies are needed to better understand people’s thought patterns about economic growth and the environment, and to expand these studies to other populations outside of British Columbia and Canada. Future research could also replicate the audience segmentation carried out here, so as to confirm the existence and consistency of the segments identified and possibly explore their   178  prevalence in other populations. New variables could be included in the segmentation exercise that go beyond attitudinal ones to include, for example, values, behavioural intentions and policy preferences.  In relation to message framing, future research could also systematically explore the most prevalent frames and discourses being used in the communication of ecological economics, and assess their cognitive, emotional, attitudinal and behavioural effects. This could allow for the identification of the most effective messages and those that backfire, thus contributing to the development of better communication strategies. Furthermore, by continuing to explore the interactions between audience segments and message frames, it is hoped that these future avenues of research will contribute to developing new frames and narratives that help people to think, in different ways, about the economy and the meaning of economic progress without growth. Finally, future studies could include and explore other emotions not included in this research, such as concern, sadness and enthusiasm.  5.5 Concluding Remarks The process of making a democratic transition to a sustainable economic model may well be one of the most relevant and debated issues of our times. This study aims to contribute to the field of ecological economics by offering data on dimensions that have not yet been widely explored, namely, public opinion and communications. Moreover, by exploring people’s initial reactions to these issues, this dissertation aspires to be part of a longer social learning process that may eventually lead to shifting and replacing some of the dominant unsustainable discourses currently upheld by society. Finally, it aims to advance knowledge on framing effects and to explore some of the conditions under which these occur. Ultimately, this doctoral dissertation aims to contribute, even indirectly, to improving the communication and rhetoric of ecological economics and especially of post-growth communications. It is hoped that the findings of this research are relevant and useful to policy-makers, communicators, educators and others interested in contributing towards the transition to a sustainable economy.   179  Bibliography Agnolucci, P., Flachenecker, F., & Söderberg, M. (2017). The causal impact of economic growth on material use in Europe. Journal of Environmental Economics and Policy, (May), 1–18.  Angie, A., Connelly, S., Waples, E., & Kligyte, V. (2011). The influence of discrete emotions on judgement and decision-making: A meta-analytic review. Cognition & Emotion, 9931(March), 1–30. Ariely, D. (2009). Predictably irrational: The hidden forces that shape our decisions. New York, USA: Harper Perennial. Assadourian, E. (2012). The path to degrowth in overdeveloped countries. In Worldwatch Institute (Ed.), State of the World 2012: Moving toward sustainable prosperity (pp. 22–37). Island Press. Babbie, E. R. (2007). The practice of social research. Belmont, CA: Wadsworth Pub Co. Bain, P. G., Hornsey, M. J., Bongiorno, R., & Jeffries, C. (2012). Promoting pro-environmental action in climate change deniers. Nature Climate Change, 2(8), 603–603. Baker, R., Blumberg, S. J., Brick, J. M., Couper, M. P., Courtright, M., Dennis, J. M., … Zahs, D. (2010). Research synthesis: AAPOR report on online panels. Public Opinion Quarterly, 74(4), 1–71. Bang, M., Medin, D., & Atran, S. (2007). Cultural mosaics and mental models of nature. Proceedings of the National Academy of Science, 104(35), 13868–13874. Bartholomew, D., Steele, F., Moustaki, I., & Galbraith, J. (2008). Analysis of multivariate social science data (2nd ed). Boca Raton, FL: Chapman & Hall. Beddoe, R., Costanza, R., Farley, J., Garza, E., Kent, J., Kubiszewski, I., … Woodward, J. (2009). Overcoming systemic roadblocks to sustainability: The evolutionary redesign of worldviews, institutions, and technologies. Proceedings of the National Academy of Sciences of the United States of America, 106(8), 2483–2489. Berg, A., & Hukkinen, J. I. (2011). The paradox of growth critique: Narrative analysis of the Finnish sustainable consumption and production debate. Ecological Economics, 72, 151–160. Bernauer, T., & McGrath, L. F. (2016). Simple reframing unlikely to boost public support for climate policy. Nature Climate Change, 6(March), 680–683.  Blake, D., Guppy, N., & Urmetzer, P. (1997). Canadian public opinion and environmental action: Evidence from British Columbia. Canadian Journal of Political Science, 30(3), 451–472. Bolsen, T., Druckman, J. N., & Cook, F. L. (2013). How frames can stunt support for scientific adaptations: Politicization and the status quo bias. In American Political Science Association Annual Meeting (pp. 1–43). Bostrom, A., Morgan, M. G., Fischhoff, B., & Read, D. (1994). What do people know about global climate change? Risk Analysis, 14(6), 959–970. Boulding, K. (1966). The economics of the coming spaceship Earth. In H. Jarrett (Ed.), Environmental Quality in a Growing Economy: Essays from the Sixth Resources for the Future for the Future Forum (pp. 3–14). Baltimore: Johns Hopkins Press. Byrch, C., Kearins, K., Milne, M., & Morgan, R. (2007). Sustainable “what”? A   180  cognitive approach to understanding sustainable development. Qualitative Research in Accounting & Management, 4(1), 26–52. Carley, K., & Palmquist, M. (1992). Extracting, representing and analyzing mental models. Social Forces, 70(3), 601–636. Cesario, J., Corker, K. S., & Jelinek, S. (2013). A self-regulatory framework for message framing. Journal of Experimental Social Psychology, 49(2), 238–249. Cheng, T., Woon, D. K., & Lynes, J. (2011). The use of message framing in the promotion of environmentally sustainable behaviors. Social Marketing Quarterly, 17(2), 48–62. Chong, D., & Druckman, J. N. (2007). Framing theory. Annual Review of Political Science, 10(1), 103–126. City of Vancouver. (2012). Greenest city: 2020 action plan. Vancouver, BC. Corbin, J., & Strauss, A. (2015). Basics of qualitative research: Techniques and procedures for developing grounded theory (4th ed). Thousand Oaks, CA: Sage Publications, Inc. Costanza, R., Arge, R., Groot, R. De, Farberk, S., Grasso, M., Hannon, B., … van den Belt, M. (1997). The value of the world’s ecosystem services and natural capital. Nature, 387(May), 253–260.  Costanza, R., Daly, H. E., & Bartholomew, J. A. (1991). Goals, agenda, and policy recommendations for ecological economics. In R. Costanza (Ed.), Ecological Economics: The science and management of sustainability (pp. 1–20). Columbia University Press. Costanza, R., Kubiszewski, I., Giovannini, E., Lovins, H., McGlade, J., Pickett, K., … Wilkinson, R. (2014). Time to leave GDP behind. Nature, 505, 283–285. Cox, R. (2013). Environmental communication and the public sphere (3rd ed). Thousand Oaks, CA: Sage Publications, Inc. Creswell, J. W. (2014). The selection of a research approach. In Research design: Qualitative, quantitative and mixed methods approaches (4th ed, pp. 3–23). Thousand Oaks, CA: SAGE Publications, Inc. Crutzen, P. J. (2002). Geology of mankind. Nature, 415, 23. Dale, A., Herbert, Y., Newell, R., & Foon, R. (2012). Action agenda: Rethinking growth and prosperity. Daly, H. (1991). A catechism of growth fallacies. In Steady-State Economics (2nd ed). Island Press. Daly, H. (1994). Operationalizing sustainable development by investing in natural capital. In A. Jansson, M. Hammer, C. Folke, & R. Constanza (Eds.), Investing in Natural Capital: The Ecological Economics Approach to Sustainability (pp. 22–38). Washington, D.C.: Island Press. Daly, H. (2013). A further critique of growth economics. Ecological Economics, 88, 20–24. Daly, H., & Farley, J. (2011). Ecological economics: Principles and applications (2nd ed). Washington D.C. Davis, J. J. (1995). The effects of message framing on response to environmental communications. Journalism & Mass Communication Quarterly, 72(2), 285–299. De Graaf, J., & Batker, D. (2011). What’s the economy for anyway? Why it’s time to stop chasing growth and start pursuing happiness. New York, USA: Bloomsbury Press.   181  Demaria, F., Schneider, F., Sekulova, F., & Martinez-Alier, J. (2013). What is degrowth? From an activist slogan to a social Movement. Environmental Values, 22(2), 191–215. Diaz-Meneses, G. (2010). Refuting fear in heuristics and in recycling promotion. Journal of Business Research, 63(2), 104–110.  Dietz, R., & O’Neill, D. (2013). Enough is enough: Building a sustainable economy in a world of finite resources. San Francisco, CA: Berrett-Koehler Publishers, Inc. Dixon, R., Griffiths, W., & Lim, G. C. (2014). Lay people’s models of the economy: A study based on surveys of consumer sentiments. Journal of Economic Psychology, 44, 13–20. Drews, S. (2016). Public and scientific opinion on climate policy, economic growth and the environment. PhD Thesis, Universitat Autonoma de Barcelona. Drews, S., & Antal, M. (2016). Degrowth: A “missile word” that backfires? Ecological Economics, 126, 182–187. Drews, S., & van den Bergh, J. C. J. M. (2016). Public views on economic growth, the environment and prosperity: Results of a questionnaire survey. Global Environmental Change, 39, 1–14. Druckman, J. N. (2004). Political preference formation: Competition, deliberation, and the (ir)relevance of framing effects. The American Political Science Review, 98(4), 671–686. Druckman, J. N., & McDermott, R. (2008). Emotion and the framing of risky choice. Political Behavior, 30(3), 297–321. Druckman, J. N., & Nelson, K. R. (2003). Framing and deliberation : How citizen’s conversations limit elite influence. American Journal of Political Science, 47(4), 729–745. Dryzek, J. S. (2013). The politics of the Earth: Environmental discourses (3rd ed). Oxford: Oxford University Press. Dunlap, R. E., & Van Liere, K. D. (1984). Commitment to the dominant social paradigm and concern for environmental quality. The Social Science Quarterly, 65(6278), 1013–1028. Dunlap, R. E., Van Liere, K. D., Mertig, A. G., & Jones, R. E. (2000). Measuring endorsement of the New Ecological Paradigm: A revised NEP scale. Journal of Social Issues, 56(3), 425–442. Durfee, J. L. (2006). “Social Change” and “Status Quo” framing effects on risk perception: An exploratory experiment. Science Communication, 27(4), 459–495. Easterlin, R. (2001). Income and happiness: Towards a unified theory. The Economic Journal, 111(473), 465–484. Ellingsen, T., Johannesson, M., Mollerstrom, J., & Munkhammar, S. (2013). Gender differences in social framing effects. Economics Letters, 118(3), 470–472. Entman, R. M. (1993). Framing: Toward clarification of a fractured paradigm. Journal of Communication, 43(4), 51–58. Environics Institute. (2012). The Common Good: Who Decides? A National Survey of Canadians. Everitt, B. S., Landau, S., Leese, M., & Stahl, D. (2011). Cluster Analysis. (W. A. Shewhart & S. S. Wilks, Eds.) (5th ed). Wiley Series on Probability and Statistics. Fagley, N. S., & Miller, P. M. (1997). Framing effects and arenas of choice: Your money   182  or your life? Organisational Behaviour and Human Decision Processes, 71(3), 355–373. Feinberg, M., & Willer, R. (2011). Apocalypse soon?: Dire messages reduce belief in global warming by contradicting just-world beliefs. Psychological Science, 22(1), 34–38. Feldman, L., & Hart, P. S. (2015). Using political efficacy messages to increase climate activism: The mediating role of emotions. Science Communication, 38(1), 99–127. Finucane, M. L., Slovic, P., Mertz, C. K., Flynn, J., & Satterfield, T. (2000). Gender, race, and perceived risk: the “white male” effect. Health, Risk & Society, 2(2), 159–172. Gentner, D., & Whitley, E. (1997). Mental models of population growth: A preliminary investigation. In M. Bazerman, D. Messick, A. Tenbrunsel, & K. Wade-Benzoni (Eds.), Environment, ethics and behavior: The psychology of environmental valuation and degradation. San Francisco, CA. Gifford, R., & Comeau, L. (2011). Message framing influences perceived climate change competence, engagement and behavioral intentions. Global Environmental Change, 21, 1301–1307. Gladwin, T., Newburry, W., & Reiskin, E. (1997). Why the northern elite mind is biased against community, the environment and sustainability. In M. Bazerman, D. Messick, A. Tenbrunsel, & K. Wade-Benzoni (Eds.), Environment, ethics and behavior: The psychology of environmental valuation and degradation. San Francisco, CA. Global Footprint Network. (2011). What happens when an infinite-growth economy runs into a finite planet? Annual Report. Global Footprint Network. (2012). Ecological Footprint 2012. Annual Report. Government of British Columbia. (2014). BC’s Green Economy: Growing Green Jobs. Government of British Columbia. (2016). Climate Leadership Plan. Gowdy, J. (2000). Terms and concepts in ecological economics. Wildlife Society Bulletin, 28(1), 26–33. Gowdy, J., & Erickson, J. D. (2005). The approach of ecological economics. Cambridge Journal of Economics, 29, 207–222. Greer, J. M. (2011). The wealth of nature: Economics as if survival mattered. Gabriola Island, Canada: New Society Publishers. Gross, K. (2008). Framing persuasive appeals: Episodic and thematic framing, emotional response, and policy opinion. Political Psychology, 29(2), 169–192. Gross, K., & D’Ambrosio, L. (2004). Framing emotional response. Political Psychology, 25(1), 1–29. Gustafsson, A. W. (2013). The metaphor challenge of future economics : Growth and sustainable development in Swedish media discourse. In M. Benner (Ed.), Before and Beyond the Global Crisis: Economics, Politics and Settlement (pp. 197–217). Edward Elgar Publishing Limited. Haluza-DeLay, R., & Fernhout, H. (2011). Sustainability and social inclusion? Examining the frames of Canadian English-speaking environmental movement organisations. Local Environment, 16(7), 727–745. Hamilton, C. (2010). Consumerism, self-creation and prospects for a new ecological consciousness. Journal of Cleaner Production, 18(6), 571–575.   183  Hardin, G. (1968). The tragedy of the commons. Science, 162 (December), 1243–1248. Hardisty, D. J., Johnson, E. J., & Weber, E. U. (2010). A dirty word or a dirty world?: Attribute framing, political affiliation, and query theory. Psychological Science, 21(1), 86–92. Hastings, G., & Stead, M. (2004). Fear appeals in social marketing: Strategic and ethical reasons for concern. Psychology and Marketing, 21(11), 961–986. Hawcroft, L. J., & Milfont, T. L. (2010). The use (and abuse) of the new environmental paradigm scale over the last 30 years: A meta-analysis. Journal of Environmental Psychology, 30(2), 143–158. Heath, Y., & Gifford, R. (2006). Free-market ideology and environmental degradation: The case of belief in global climate change. Environment and Behavior, 38(1), 48–71. Hedlund-de Witt, A. (2012). Exploring worldviews and their relationships to sustainable lifestyles: Towards a new conceptual and methodological approach. Ecological Economics, 84, 74–83. Heinberg, R. (2011). The end of growth: Adapting to our new economic reality. Gabriola Island, Canada: New Society Publishers. Hine, D., & Gifford, R. (1991). Fear appeals, individual differences, and environmental concern. Journal of Environmental Education, 23, 36–41. Hine, D., Phillips, W., Cooksey, R., Reser, J., Nunn, P., Marks, A., … Watt, S. (2016). Preaching to different choirs: How to motivate dismissive, uncommitted, and alarmed audiences to adapt to climate change? Global Environmental Change, 36 (January), 1–11. Hine, D., Phillips, W., Reser, J., Cooksey, R., Marks, A., Nunn, P., … Ellul, M. (2013). Enhancing climate change communication: Strategies for profiling and targeting Australian interpretive communities. Gold Coast, Australia. Hine, D., Reser, J., Morrison, M., Phillips, W., Nunn, P., & Cooksey, R. (2014). Audience segmentation and climate change communication: Conceptual and methodological considerations. Wiley Interdisciplinary Reviews: Climate Change, 5(July/August), 441–459. Hine, D., Reser, J., Phillips, W., Cooksey, R., Marks, A., Nunn, P., … Glendon, I. (2013). Identifying climate change interpretive communities in a large Australian sample. Journal of Environmental Psychology, 36, 229–239. Hoffman, M., Lubell, M., & Hillis, V. (2014). Linking knowledge and action through mental models of sustainable agriculture. PNAS, 111(36), 13016–21. Hopkins, R. (2008). The transition companion: Making your community more resilient in uncertain times. Totnes, United Kingdom: UIT Cambridge. Hornsey, M. J., & Fielding, K. S. (2016). A cautionary note about messages of hope: Focusing on progress in reducing carbon emissions weakens mitigation motivation. Global Environmental Change, 39, 26–34.  Hruschka, D. J., Schwartz, D., St.John, D. C., Picone-Decaro, E., Jenkins, R. A., & Carey, J. W. (2004). Reliability in coding open-ended data: Lessons learned from HIV behavioral research. Field Methods, 16(3), 307–331. Hunter, L., Hatch, A., & Johnson, A. (2004). Cross-national gender variation in environmental behaviors. Social Science Quarterly, 85(3), 677–694. IEA. (2008). Worldwide trends in energy use and efficiency: Key insights from IEA   184  ndicator analysis. Paris, France. Iyengar, S. (1991). Is anyone responsible? Chicago: The University of Chicago Press. Jack, R. E., Garrod, O. G. B., & Schyns, P. G. (2014). Dynamic facial expressions of emotion transmit an evolving hierarchy of signals over time. Current Biology, 24(2), 187–192. Jackson, T. (2011). Prosperity without growth: Economics for a finite planet. UK: Earthscan. Jeffrey, K., Wheatley, H., & Abdallah, S. (2016). The Happy Planet Index: 2016. A global index of sustainable well-being. London, U.K. Jepson, E. (2004). Human nature and sustainable development: A strategic challenge for planners. Journal of Planning Literature, 24(249), 3–15. Jones, N., Ross, H., Lynam, T., & Perez, P. (2014). Eliciting mental models: A comparison of interview procedures in the context of natural resource management. Ecology and Society, 19(1), 13. Jones, N., Ross, H., Lynam, T., Perez, P., & Leitch, A. (2011). Mental models: An interdisciplinary synthesis of theory and methods. Ecology And Society, 16(1). Kahan, D., & Braman, D. (2006). Cultural cognition and public policy. Yale Journal of Law and Public Policy, 24, 147–170. Kahan, D., Braman, D., Gastil, J., Slovic, P., & Mertz, C. K. (2007). Culture and identity-protective cognition: Explaining the white male effect in risk perception. Journal of Empirical Law Studies, 4(3), 465–505. Kahan, D., Jenkins-Smith, H., & Braman, D. (2011). Cultural cognition of scientific consensus. Journal of Risk Research, 14(2), 147–174. Kahneman, D. (2011). Thinking, Fast and Slow. New York, USA: Farrar, Straus and Giroux. Kallis, G. (2011). In defence of degrowth. Ecological Economics, 70(5), 873–880. Kearney, A. R., & Kaplan, S. (1997). Toward a methodology for the measurement of knowledge structures of ordinary people: The conceptual content cognitive map (3CM). Environment and Behavior, 29(5), 579–617. Kerschner, C. (2010). Economic de-growth vs. steady-state economy. Journal of Cleaner Production, 18(6), 544–551. Kilbourne, W. E., Beckmann, S. C., & Thelen, E. (2002). The role of the dominant social paradigm in environmental attitudes: A multinational examination. Journal of Business Research, 55(3), 193–204. Klein, N. (2014). This changes everything: Capitalism vs. the climate. New York, NY, USA: Simon and Schuster. Klitgaard, K., & Krall, L. (2012). Ecological economics, degrowth, and institutional change. Ecological Economics, 84, 247–253. Kolandai-Matchett, K. (2009). Mediated communication of “sustainable consumption” in the alternative media: A case study exploring a message framing strategy. International Journal of Consumer Studies, 33(2), 113–125. Koltko-Rivera, M. E. (2004). The psychology of worldviews. Review of General Psychology, 8(1), 3–58. Krausmann, F., Erb, K. H., Gingrich, S., Haberl, H., Bondeau, A., Gaube, V., … Searchinger, T. (2013). Global human appropriation of net primary production doubled in the 20th century. Proceedings of the National Academy of Sciences,   185  110(25), 10324–10329. Krausmann, F., Gingrich, S., Eisenmenger, N., Erb, K. H., Haberl, H., & Fischer-Kowalski, M. (2009). Growth in global materials use, GDP and population during the 20th century. Ecological Economics, 68, 2696–2705. Krautkraemer, J. A. (2005). Economics of Scarcity. In D. Simpson, M. Toman, & R. Ayres (Eds.), Scarcity and growth revisited: Natural resources and the environment in the new millennium (pp. 54–77). Washington D.C, U.S.A.: Resources for the Future. Kubiszewski, I., Costanza, R., Franco, C., Lawn, P., Talberth, J., Jackson, T., & Aylmer, C. (2013). Beyond GDP: Measuring and achieving global genuine progress. Ecological Economics, 93, 57–68. Lakoff, G. (2004). Don’t think of an elephant: Know your values and frame the debate. Vermont: Chelsea Green Publishing. Lakoff, G. (2010). Why it matters how we frame the environment. Environmental Communication: A Journal of Nature and Culture, 4(1), 70–81. Latouche, S. (2007). Degrowth: An electoral stake? The International Journal of Inclusive Democracy, 3(1). Lecheler, S., Schuck, A. R. T., & de Vreese, C. H. (2013). Dealing with feelings: Positive and negative discrete emotions as mediators of news framing effects. Communications - The European Journal of Communication Research, 38(2), 189–209. Leiserowitz, A. (2006). Climate change risk perception and policy preferences: The role of affect, imagery, and values. Climatic Change, 77(1–2), 45–72. Leiserowitz, A., Kates, R., & Parris, T. (2006). Sustainability values, attitudes, and behaviors: A review of multinational and global trends. Annual Review of Environment and Resources, 31(1), 413–444. Leiserowitz, A., Maibach, E., & Roser-Renouf, C. (2007). Global warming’s “Six Americas”: An audience segmentation. Lerner, J. S., & Keltner, D. (2001). Fear, anger and risk. Journal of Personality and Social Psychology, 81(1), 146–159. Levin, I., Gaeth, G., Schreiber, J., & Lauriola, M. (2002). A new look at framing effects: Distribution of effect sizes, individual differences, and independence of types of effects. Organizational Behavior and Human Decision Processes, 88(1), 411–429. Levin, I., Schneider, S., & Gaeth, G. (1998). All frames are not created equal: A typology and critical analysis of framing effects. Organizational Behavior and Human Decision Processes, 76(2), 149–188. Lockwood, M. (2011). Does the framing of climate policies make a difference to public support? Evidence from UK marginal constituencies. Climate Policy, 11, 1097–1112. Lodge, M., & Taber, C. S. (2013). The rationalizing voter. New York, USA: Cambridge University Press. Loroz, P. S. (2007). The interaction of message frames and reference points in prosocial persuasive appeals. Psychology and Marketing, 24(11), 1001–1023. Lu, H. (2016). The effects of emotional appeals and gain versus loss framing in communicating sea star wasting disease. Science Communication, 38(2), 143–169. Lu, H., & Schuldt, J. (2015). Exploring the role of incidental emotions in support for   186  climate change policy. Climatic Change, 131(4), 719–726. Luks, F. (1998). The rhetorics of ecological economics. Ecological Economics, 26(2), 139–149. Lynam, T., & Brown, K. (2012). Mental models in human – environment interactions: Theory, policy implications and methodological explorations. Ecology and Society, 17(3:24). Lynam, T., Mathevet, R., Etienne, M., Stone-Jovicich, S., Leitch, A., Jones, N., … Perez, P. (2012). Waypoints on a journey of discovery: Mental models in human-environment interactions. Ecology and Society, 17(3), 23–33. Magidson, J., & Vermunt, J. K. (2004). Latent class models. In D. Kaplan (Ed.), The Sage handbook of qualitative metholodogy for the social sciences (pp. 175–198). Thousand Oaks, California: Sage Publications, Inc. Maheswaran, D., & Meyers-Levy, J. (1990). The influence of message framing and issue involvement. Journal of Marketing Research, 27, 361–367. Maibach, E. W., Leiserowitz, A., Roser-Renouf, C., & Mertz, C. K. (2011). Identifying like-minded audiences for global warming public engagement campaigns: An audience segmentation analysis and tool development. PLOS ONE, 6(3). Mankiw, G., & Scarth, W. (2004). Macroeconomics: Canadian Edition (2nd ed). New York, USA: Worth Publishers. Martínez-Alier, J., Pascual, U., Vivien, F.-D., & Zaccai, E. (2010). Sustainable de-growth: Mapping the context, criticisms and future prospects of an emergent paradigm. Ecological Economics, 69(9), 1741–1747. Matthey, A. (2010). Less is more: The influence of aspirations and priming on well-being. Journal of Cleaner Production, 18(6), 567–570. McDonald, S. (2009). Changing climate, changing minds: Applying the literature on media effects, public opinion, and the issue-attention cycle to increase public understanding of climate change. International Journal of Sustainability Communication, 4, 45–63. McLean, C. P., & Anderson, E. R. (2009). Brave men and timid women? A review of the gender differences in fear and anxiety. Clinical Psychology Review, 29(6), 496–505. McNeill, J. (2000). Something new under the sun: An environmental history of the twentieth century world. New York, USA: Norton & Company. Meadows, D. (1997). Places to intervene in a system. Whole Earth, Winter, 78–84. Meadows, D. (1998). Indicators and information systems for sustainable development. Vermont. Meadows, D. (2006). Tools for the transition to sustainability. In M. Keiner (Ed.), The future of sustainability (pp. 161–178). Springer. Meadows, D. (2008). Thinking in Systems: A Primer. (D. Wright, Ed.). Vermont: Chelsea Green Publishing. Meadows, D. D., Meadows, D. D., Randers, J., & Behrens, W. (1972). The limits to growth. Universe Books. Meadows, D., Meadows, D., & Randers, J. (1992). Beyond the limits: Confronting global collapse, envisioning a sustainable future. Vermont: Chelsea Green Publishing. Meyers-Levy, J., & Maheswaran, D. (2004). Exploring message framing outcomes when systematic, heuristic, or both types of processing occur. Journal of Consumer Psychology, 14(1–2), 159–167.   187  Mignolo, W. (2016). Sustainable development or sustainable economies? Ideas towards living in harmony and plenitude. Miles, M., Huberman, M., & Saldana, J. (2014). Qualitative data analysis: A methods sourcebook (3rd ed). Thousand Oaks, CA: SAGE Publications, Inc. Millenium Ecosystem Assessment. (2005). Living Beyond Our Means: Natural Assets and Human Well-being - Statement from the Board. Washington D.C. Moore Lappe, F. (2013). EcoMind: Changing the way we think, to create the world we want. New York, USA: Nation Books. Morgan, G., Fischhoof, B., Bostrom, A., & Atman, C. (2002). Risk communication: A mental models approach. New York, USA: Cambridge University Press. Moser, S., & Dilling, L. (2004). Making climate hot: Communicating the urgency and challenge of global climate change. Environment, 34, 32–46. Myers, T., Nisbet, M. C., Maibach, E. W., & Leiserowitz, A. (2012). A public health frame arouses hopeful emotions about climate change. Climatic Change, 113(3–4), 1105–1112. Nabi, R. (2003). Exploring the framing effects of emotions: Do discrete emotions differentially influence information accessibility, information seeking, and policy preference? Communication Research, 30(2), 224–247. Nan, X. (2007). The relative persuasive effect of gain- versus loss-framed messages: Exploring the moderating role of the desirability of end-states. Journalism and Mass Communication Quarterly, 84(3), 509–524. Nelson, T., Bryner, S., & Carnahan, D. (2011). Media and politics. In J. N. Druckman, D. P. Green, J. H. Kuklinski, & A. Lupia (Eds.), Cambridge Handbook of Experimental Political Science (pp. 201–213). New York, USA: Cambridge University Press. Nelson, T., Clawson, R., & Oxley, Z. (1997). Media framing of a civil liberties conflict and its effect on tolerance. American Political Science Review, 91(3), 567–583. Nelson, T., Oxley, Z., & Clawson, R. (1997). Toward a psychology of framing effects. Political Behavior, 19(3), 221–246. New Economics Foundation. (2009). The Great Transition. Nikiforuk, A. (2012). The energy of slaves: Oil and the new servitude. Vancouver: David Suzuki Foundation & Greystone Books. Nisbet, M. (2009). Communicating climate change: Why frames matter for public engagement. Environment, 51(2), 12–23. Norgard, J. S., Peet, J., & Ragnarsdottir, K. V. (2010). The history of The Limits to Growth. Solutions Journal, 1(2), 59–63. Norusis, M. (2009). Cluster Analysis. In SPSS 16.0 Statistical Procedures Companion (pp. 361–391). Prentice Hall. Norusis, M. (2011). Ordinal regression. In IBM SPSS Statistics 19 Advanced Statistical Procedures Companion (pp. 69–89). Addison Wesley. Nylund, K. L., Asparouhov, T., & Muthén, B. O. (2007). Deciding on the number of classes in Latent Class Analysis and Growth Mixture Modeling: A Monte Carlo Simulation study. Structural Equation Modeling, 14(4), 535–569. O’Neill, S., & Nicholson-Cole, S. (2009). “ Fear Won’t Do It ” Promoting positive engagement with climate change through visual and iconic representations. Science Communication, 30(3), 355–379. Ockwell, D. G. (2008). Energy and economic growth: Grounding our understanding in   188  physical reality. Energy Policy, 36(12), 4600–4604.  OECD. (2011). Towards Green Growth. Ott, K., Muraca, B., & Baatz, C. (2011). Strong sustainability as a frame for sustainability communication. In J. Godemann & G. Michelsen (Eds.), Sustainability communication: Interdisciplinary perspectives and theoretical foundations (pp. 13–25). Springer. OXFAM. (2016). An Economy for the 1% (Summary). OXFAM Briefing Papers (Vol. 210). Pahl-Wostl, C., & Hare, M. (2004). Processes of social learning in integrated resources management. Journal of Community & Applied Social Psychology, 14, 193–206. Pearce, D. W., & Atkinson, G. D. (1993). Capital theory and the measurement of sustainable development: An indicator of “weak” sustainability. Ecological Economics, 8, 103–108. Pelletier, L. G., & Sharp, E. (2008). Persuasive communication and proenvironmental behaviours: How message tailoring and message framing can improve the integration of behaviours through self-determined motivation. Canadian Psychology, 49(3), 210–217. Perloff, R. (2010). The dynamics of persuasion: Communication and attitudes in the 21st century (4th ed). New York, USA: Routledge. Pezzey, J., & Toman, M. (2002). The economics of sustainability: A review of journal articles. Resources for the Future Discussion Paper 02-03. Poortinga, W., & Darnton, A. (2016). Segmenting for sustainability: The development of a Welsh model to engage the public in sustainability and sustainability-related issues. Journal of Environmental Psychology, 45, 221–232. Price, V., Lilach, N., & Cappella, J. (2005). Framing public discussion of gay civil unions. Public Opinion Quarterly, 69(2), 179–212. Price, V., & Tewksbury, D. (1997). News values and public opinion: A theoretical account of media priming and framing. In G. A. Barett & F. J. Boster (Eds.), Progress in communication sciences: Advances in persuasion (Volumen 13) (pp. 173–212). Price, V., Tewksbury, D., & Powers, E. (1997). Switching trains of thought: The impact of news frames on readers’ cognitive responses. Communication Research, 24(5), 481–506. Pyman, H., & Pammett, J. (2010). Environmental attitudes: A comparisson of Canada, Europe and the United States over time. Raworth, K. (2012). A safe and just space for humanity: Can we live within the doughnut? Rees, W. (1995). Acheiving sustainability: Reform or transformation? Journal of Planning Literature, 9(4), 343–361. Rees, W. (2003a). Carrying capacity and sustainability: Waking Malthus’ ghost. In D. Bell & A. Cheung (Eds.), Encyclopedia of life support systems. Oxford: EOLSS Publishers. Rees, W. (2003b). Economic development and environmental protection: An ecological economics perspective. Environmental Monitoring and Assessment, 86, 29–45. Rees, W. (2010). What’s blocking sustainability? Human nature, cognition, and denial. Sustainability: Science, Practice and Policy, 6(2), 13–25.   189  Rees, W. (2011). Toward a sustainable world economy. In Institute for New Economic Thinking Annual Conference - Crisis and Renewal: International Political Economy at the Crossroads. Bretton Woods, NH, USA. Rees, W., & Wackernagel, M. (2005). Ecological footprints and appropriated carrying capacity: Measuring the natural capital requirements of the human economy. In M. Redclift (Ed.), Sustainability: Critical Concepts in the Social Sciences (Volume III). Routledge. Rockstrom, J., Steffen, W., Noone, K., Persson, A., Chapin, S., Lambin, E., … Foley, J. (2009). Planetary boundaries: Exploring the safe operating space for humanity. Ecology and Society, 14(2), 1–32. Røpke, I. (2004). The early history of modern ecological economics. Ecological Economics, 50(3–4), 293–314. Rosner, W. (1995). Mental models for sustainability. Journal of Cleaner Production, 3(1–2), 107–121. Rothman, A., & Salovey, P. (1997). Shaping perceptions to motivate healthy behavior: The role of message framing. Psychological Bulletin, 121(1), 3–19. Salter, J. (2015). Energy in place: A case study and mental models analysis of engagement in community scale energy planning. PhD Thesis, University of British Columbia. Scheufele, D. (1999). Framing as a theory of media effects. Journal of Communication, 49(1999), 103–122. Scheufele, D., & Nisbet, M. (2007). Framing. In K. L.L & C. Holtz-Bacha (Eds.), Encyclopedia of political communication (Vol. 1, pp. 254–257). Thousand Oaks, California: Sage Publications, Inc. Schmidt, A., Ivanova, A., & Schafer, M. S. (2013). Media attention for climate change around the world: A comparative analysis of newspaper coverage in 27 countries. Global Environmental Change, 23(5), 1233–1248. Schneider, F., Kallis, G., & Martinez-Alier, J. (2010). Crisis or opportunity? Economic degrowth for social equity and ecological sustainability. Introduction to this special issue. Journal of Cleaner Production, 18(6), 511–518. Schreiber, J. B., & Pekarik, A. J. (2014). Technical note: Using Latent Class Analysis versus K-means or Hierarchical clustering to understand museum visitors. Curator: The Museum Journal, 57(1), 45–59. Schuldt, J. P., & Roh, S. (2014). Media frames and cognitive accessibility: What do “global warming” and “climate change” evoke in partisan minds? Environmental Communication, 8(4), 529–548. Schultz, P. W., Shriver, C., Tabanico, J. J., & Khazian, A. M. (2004). Implicit connections with nature. Journal of Environmental Psychology, 24(1), 31–42. Shafer, W. (2006). Social Paradigms and attitudes toward environmental accountability. Journal of Business Ethics, 65(2), 121–147. Shen, F. (2004). Effects of news frames and schemas on individuals’ issue interpretations and attitudes. Journalism & Mass Communication Quarterly, 81(2), 400–416. Sheppard, S. R. J. (2012). Visualizing Climate Change: A Guide to Visual Communication of Climate Change and Developing Local Solutions. Routledge. Shome, D., & Marx, S. (2009). The psychology of climate change communication: A guide for scientists, journalists, educators, political aides, and the interested public.   190  Center for Research on Environmental Decisions. New York, USA. Simon, J. (1981). The Ultimate Resource. Princeton, New Jersey: Princeton University Press. Smith, N., & Leiserowitz, A. (2014). The role of emotion in global warming policy support and opposition. Risk Analysis, 34(5), 937–948. Snyder, C. R., Rand, K. L., & Sigmon, D. R. (2005). Hope theory: A member of the positive psychology family. In C. R. Snyder & J. Lopez, Shane (Eds.), Handbook of positive psychology (pp. 257–276). Oxford University Press. Spash, C. (2011). Social Ecological Economics: Understanding the past to see the future. American Journal of Economics and Sociology, 70(2), 341–375. Spence, A., & Pidgeon, N. (2010). Framing and communicating climate change: The effects of distance and outcome frame manipulations. Global Environmental Change, 20(4), 656–667. Steffen, W., Persson, A., Deutsch, L., Zalasiewicz, J., Williams, M., Richardson, K., … Svedin, U. (2011). The anthropocene: From global change to planetary stewardship. Ambio, 40, 739–761. Steffen, W., Richardson, K., Rockström, J., Cornell, S., Fetzer, I., Bennett, E., … Carpenter, S. (2015). Planetary boundaries: Guiding human development on a changing planet. Science, 347(6223), 736–747. Stern, P. C. (2012). Psychology: Fear and hope in climate messages. Nature Climate Change, 2(8), 572–573. Stevenson, K., & Peterson, N. (2016). Motivating action through fostering climate change hope and concern and avoiding despair among adolescents. Sustainability, 8(1), 1–10. Stone-Jovicich, S., Lynam, T., Leitch, A., & Jones, N. (2011). Using consensus analysis to assess mental models about water use and management in the crocodile river catchment. Ecology and Society, 16(1), 45. Thompson, S. C. G., & Barton, M. A. (1994). Ecocentric and anthropocentric attitudes toward the environment. Journal of Environmental Psychology, 14, 149–157. Trainer, T. (2010). De-growth – is not enough. The International Journal of INCLUSIVE DEMOCRACY, 6(4), 1–14. Tversky, A., & Kahneman, D. (1981). The framing of decisions and psychology of choice. Science, 211, 453–458. UNDP. (2016). Human development report 2016: Human development for everyone. New York, USA. UNECE, & FAO. (2014). Rovaniemi Action Plan for the Forest Sector in a Green Economy. Geneva. UNEP. (2011). Towards a Green Economy: Pathways to sustainable development and poverty eradication. Union of Concerned Scientists. (1992). World scientists’ warning to humanity. Retrieved from http://www.ucsusa.org/about/1992-world-scientists.html#.V7XpeSgrKUk Van de Velde, L., Verbeke, W., Popp, M., & Van Huylenbroeck, G. (2010). The importance of message framing for providing information about sustainability and environmental aspects of energy. Energy Policy, 38(10), 5541–5549. van den Bergh, J. C. J. M. (2011). Environment versus growth — A criticism of “degrowth” and a plea for “a-growth.” Ecological Economics, 70(5), 881–890.   191  Van den Bergh, J. C. J. M. (2017). Don’t worry, be happy. Alternatives Journal, 43(1), 22–24. van der Linden, S. (2015). A conceptual critique of the cultural cognition thesis. Science Communication, 38(1), 128–138.  Vermunt, J. K., & Magidson, J. (2005). Latent Gold 4 . 0 User’s Guide. Belmont, Massachusetts: Statistical Innovations Inc. Victor, P. (2008). Managing without growth: Slower by design, not disaster. Cheltenham, Glos, UK: Edward Elgar Publishing Limited. Victor, P. (2017). We’ve outgrown growth. Alternatives Journal, 43(1), 16–20. Wackernagel, M., & Rees, W. (1996). Our ecological footprint: Reducing human impact on the Earth. Gabriola Island, BC, Canada: New Society Publishers. Ward, J. D., Sutton, P. C., Werner, A. D., Costanza, R., Mohr, S. H., & Simmons, C. T. (2016). Is decoupling GDP growth from environmental impact possible? PLOS ONE, 11(10), 1–14. Weaver, D. H. (2007). Thoughts on agenda setting, framing, and priming. Journal of Communication, 57(1), 142–147. Weiss, M., & Cattaneo, C. (2017). Degrowth – Taking stock and reviewing an emerging academic paradigm. Ecological Economics, 137, 220–230. Wiest, S., Raymond, L., & Clawson, R. A. (2012). All climate politics are local? Framing effects on attitudes toward policies on climate change. World Values Survey. (2015). Online Data Analysis. Retrieved May 19, 2016, from http://www.worldvaluessurvey.org/WVSOnline.jsp Zelezny, L. C., Chua, P.-P., & Aldrich, C. (2000). Elaborating on gender differences in environmentalism. Journal of Social Issues, 56(3), 443–457.      192  Appendices Appendix A: Interview protocol  Brief introduction to economic growth: Most politicians, media, businesses and even citizen’s attention are greatly focused on the economy and specifically on economic growth. In this context, I would like to know...  1. Topic: Economic growth a. When the media and government talk about economic growth, what do you think they are referring to? Anything else?  b. What is the first image that comes to your mind when you think about ‘economic growth’? Probe: How strong are your negative or positive feelings about___ (-5 very negative / +5 very positive) Interviewer: So you’re right in that economic growth is... but it also is the increase in the production and consumption of goods and services in an economy over time. It’s measured by the increase of Gross Domestic Product (GDP).   2. Topic: Benefits and costs of growth a. What good things have come from economic growth? How has your generation benefitted from this? Why do you think that? Is your life very different from the one your parents had?  b. Can you come up with any negative things that have arisen from economic growth?  c. In general terms, do you think that the positive benefits of economic growth outweigh any negative consequences, or is it the other way around?   3. Topic: Growth and development a. Is economic growth different from development? How? Is it different from progress or prosperity? How? b. What is/should be the end goal of the economy/economic process?   4. Topic: Future material abundance and limits to growth a. There is an assumption by many that economic growth can continue indefinitely. What do you think about that?  b. Are there any limits or constraints to how much an economy can grow? What would impose those limits? Explain why you think that.   5. Topic: Connections between the economy and the environment a. Please depict with these figures the relationship between economy, environment and society.  b. Which one is the most important? Why? c. Can the economy exist without the environment? How?     193  6. Topic: Sustainability a. What is sustainability for you? How important do you think it is? b. On a scale of 1 to 10 (1 the lowest, 10 the highest), how important should environmental sustainability be for governments?  c. Is indefinite economic growth compatible with environmental sustainability? How?   7. Topic: Optimal scale for the economy Introduction to steady state economy concepts: Some scientists, environmentalists and economists believe that the long-term, indefinite expansion of the economy is not possible or desirable on a finite planet. They believe that eventually the economy should reach a relatively stable size. They have called this model a ‘steady state economy’. [Introduce ecological economics worldview of embedded systems]. In a steady state economy the physical components of the economy (e.g. natural resources, human populations and stocks of human-built capital) should be kept relatively stable. Any non-physical components of an economy (e.g., knowledge, quality of life, human connection and development) can grow indefinitely. The main objective is to establish it at a sustainable scale that does not exceed ecological limits, so an economy could reach a steady state after a period of growth or after a period of downsizing or degrowth. a. What comes to your mind when you hear this description? List the 3 initial thoughts that come to you. On a scale of -5 to +5, how strong are your positive or negative feelings about these? Did it resonate with you? b. Hypothetically, if your preferred political candidate runs on a platform like this during the next elections… What would be your initial reaction? What would be your major fear? Can you see any benefits? (Explore doubts, misconceptions, fear to change, etc.) c. Is there anything that you would like to add, comment?       194  Appendix B: Visual aids used during the interviews B.1 Economic growth of the world economy since the 1960s          B.2 Figures of the economy, society and the environment used to aid participants in explaining the relationship between these systems     ENVIRONMENT     ENVIRONMENT    ENVIRONMENT    SOCIETY     ECONOMY    SOCIETY    ECONOMY   SOCIETY  ECONOMY  01020304050607080196019631966196919721975197819811984198719901993199619992002200520082011Trillion US$TimeWorld Economic Growth (Current US$)  195  B3. Figure used to explain ideas related to a steady state economy    Use of planetary resources   STEADY STATE ECONOMY  Natural resources Population  Material goods    196  Appendix C: Sociodemographics survey   What is your age? Under 20 years old 20 – 29 years old 30 – 39 years old  40 – 49 years old  50 – 59 years old 60 – 69 years old 70 years or older  What is the highest degree or level of school you have completed? If currently enrolled, highest degree received. Please check one.   Less than high school Some high school Finished high school  Technical/vocational training  Some college/university education Bachelor’s degree Post-graduate degree Other (please specify)  What is your occupation? If you are a homemaker, student, retired or unemployed, please state this and list your former occupation (if applicable)    Which of the following, best represents your racial or ethnic heritage? White/Caucasian South Asian East Asian Black Latino/Hispanic  Native American Arab Other  What is your place of residence (e.g. name of city, town) ____________________________  In your political ideology, you consider yourself:  Conservative Moderate Liberal Other  The following questions ask about you. Your answers will be kept confidential.    The following questions ask about you. Your answers will be kept confidential.    The following questions ask about you. Your answers will be kept confidential.    The following questions ask about you. Your answers will be kept confidential.     197  Appendix D: Cultural cognition of risk survey  People in our society often disagree about how far to let individuals go in making decisions for themselves. How strongly you agree or disagree with each of these statements? Strongly Disagree Moderately Disagree Slightly Disagree Slightly              Agree Moderately Agree Strongly Agree 1. The government interferes far too much in our everyday lives.             2. The government should do more to advance society's goals, even if that means limiting the freedom and choices of individuals.              3. It's not the government's business to try to protect people from themselves.             4. Sometimes government needs to make laws that keep people from hurting themselves.             5. The government should stop telling people how to live their lives.             6. Government should put limits on individual choices so they don't get in the way of what's good for society.              People in our society often disagree about issues of equality and discrimination.  How strongly you agree or disagree with each of these statements?  Strongly Disagree Moderately Disagree Slightly Disagree Slightly              Agree Moderately Agree Strongly Agree 1. Discrimination against minorities is still a very serious problem in our society.             2. We need to dramatically reduce inequalities between the rich and the poor, whites and people of color and men and women.             3. Society as a whole has become too soft and feminine.             4. We have gone too far in pushing equal rights in this country.             5. It seems like minorities don't want equal rights, they want special rights just for them.             6. Our society would be better off if the distribution of wealth was more equal.             Note: Survey items obtained from Kahan et al. (2011).   198  Appendix E: Output examples of NVivo 10 cluster analysis  E.1 Cluster analysis using Jaccard’s coefficient similarity metric. The most consistent clusters are indicated with rectangular boxes.   CLUSTER A CLUSTER B  CLUSTER B  CLUSTER B  CLUSTER B CLUSTER X  CLUSTER X  CLUSTER X  CLUSTER X CLUSTER E  CLUSTER E  CLUSTER E  CLUSTER E CLUSTER D  CLUSTER D  CLUSTER D  CLUSTER D CLUSTER C  CLUSTER C  CLUSTER C  CLUSTER C   199  E.2 Analysis removing less directly relevant codes (e.g. views on the supernatural, opinions about the past) and using Sorensen’s coefficient as a similarity metric. Rectangular boxes represent the most consistent clusters.      CLUSTER A CLUSTER E CLUSTER D CLUSTER X CLUSTER B   200  E.3 Analysis removing an increasing number of less directly relevant codes (e.g. views on the future, opinions about overpopulation) and using Pearson’s coefficient as a similarity metric. Rectangular boxes represent the most consistent clusters.     CLUSTER X CLUSTER E CLUSTER D CLUSTER B CLUSTER A   201  Appendix F: Survey example SURVEY ON THE ECONOMY, SOCIETY AND THE ENVIRONMENT SECTION 1. In this section we are interested in understanding your opinions about economic growth.  The following is a list of statements about economic growth and other related issues. In this survey, economic growth is understood to mean the annual increase in value of all goods and services produced in an economy (commonly measured as Gross Domestic Product or GDP). What is your level of agreement or disagreement with each of the following statements?   Strongly Disagree Moderately Disagree Neither Disagree nor Agree Moderately Agree Strongly Agree Economic growth is largely a good thing           The negative consequences of economic growth are greater than its benefits           Economic growth and environmental sustainability are compatible           There are no limits to the capacity of the economy to keep expanding            Humans are as much a part of nature as other animals           Technology will eventually solve our problems with scarce natural resources           The world is currently not environmentally sustainable            Developing countries have a lower impact on the environment than developed nations           Economic growth will eventually be limited by the availability of natural resources           The so called 'ecological crisis' facing humankind has been greatly exaggerated           Humans depend on nature to survive           Human ingenuity will ensure that we do not make the Earth unlivable            Note: Order of variables was randomized.      202  SECTION 2. In this section we want to learn about your interest and concern with economic and environmental issues.  Generally speaking, how concerned are you about the state of the natural environment?  a. Not at all concerned b. A little concerned c. Somewhat concerned d. Very concerned e. Extremely concerned  Generally speaking, how concerned are you about the state of the Canadian economy?  a. Not at all concerned b. A little concerned c. Somewhat concerned d. Very concerned e. Extremely concerned  How often do you think about how the economy and the environment affect each other? a. Never b. Not very much c. A fair amount d. A great deal     203  SECTION 3. Please read the following paragraph carefully. The information in the passage is based on current arguments being presented in various scientific articles. We will ask you about your opinions and perspectives after you have read the information.   Treatment surveys included versions of the framed messages – Frame 1, Frame 2, Frame 3 or Frame 4 (see Table 4.2 for the exact wording of each message frame).  The control survey did not include this section.    SECTION 4.   Treatment conditions wording:   In this section we are interested in hearing your opinions about the information you just read. Please write down up to 2 or 3 thoughts or ideas that you are currently thinking.   Control condition wording:  Please write down up to 3 thoughts or ideas that you currently have about the economy, society or the environment, and/or how they affect each other.       204  SECTION 5* Overall, how credible were the arguments provided in the previous paragraph? a. Very credible b. Somewhat credible c. Slightly credible d. Not credible at all  Overall, how convincing were the arguments provided in the previous paragraph? a. Very convincing b. Somewhat convincing c. Slightly convincing d. Not convincing at all  Overall, how weak or strong were the arguments provided in the previous paragraph? a. Very weak b. Somewhat weak c. Somewhat strong d. Very strong  * This section was not included in the control condition.    205  SECTION 6*. In this section we are interested in knowing what emotions you felt when you were reading the previous paragraph.   While reading the paragraph did you feel hopeful? a. Not at all hopeful  b. A little hopeful c. Somewhat hopeful d. Very hopeful e. Extremely hopeful  In a sentence or two, could you describe why did you feel this way?  While reading the paragraph did you feel fearful? a. Not at all fearful  b. A little fearful c. Somewhat fearful d. Very fearful e. Extremely fearful  In a sentence or two, could you describe why did you feel this way?  While reading the paragraph did you feel angry? a. Not at all angry b. A little angry c. Somewhat angry d. Very angry e. Extremely angry  In a sentence or two, could you describe why did you feel this way?  * This section was not included in the control condition.         206  SECTION 7. In this section we are interested in knowing your opinions and beliefs about economic growth in Canada.  The following statements refer to economic growth IN CANADA. What is your level of agreement or disagreement with each of the following statements?    Strongly Disagree Moderately Disagree Neither Disagree nor Agree Moderately Agree Strongly Agree Continued economic growth is essential for improving people’s quality of life           Economic growth is the best measure of social progress      Politicians should give less priority to economic growth as a major public policy goal          A 'good life' is possible without continuous economic growth           In view of limited natural resources, people should figure out ways to increase quality of life while reducing overall material consumption           Economic growth will not be limited by the availability of natural resources           The benefits of economic growth outweigh its negative consequences      We should continue growing our economy despite any large negative consequences           We should eventually transition into an economic model based on reduced levels of consumption      A sustainable economic model will only be possible if we stabilize the size of our population       Note: Order of variables was randomized.      207  SECTION 8.  How likely or unlikely are you to support a Canadian politician that does NOT pursue economic growth as a major policy goal? a. Very likely b. Somewhat likely c. Somewhat unlikely d. Very unlikely  In a sentence or two, could you describe why you are [very likely, somewhat likely, somewhat unlikely, very unlikely] to support a Canadian politician that does NOT pursue economic growth as a major policy goal?   Please assign points to the following issues. The total of all three should be 100. The more important you think an issue is, the more points you should assign to it.  A. Economic growth  B. Environmental issues C. Social well-being  _______  _______  _______   In the past decade, the economy has generally grown every year. What level of economic growth do you think the government should aim for in the next 10 years? a. More than in the previous decade b. About the same as in the previous decade c. Less than in the previous decade d. I don’t know    208  SECTION 9. In this section we are interested in knowing more about your awareness and initial reaction towards some economic terms.  The following terms are sometimes used in the media. How aware are you of their meanings?  Not at all Aware  Not very Aware Somewhat Aware Very Aware Economic growth     Steady state economy     Green economy     Sustainable degrowth     Postgrowth       Regardless of your awareness or knowledge about each term, how do you associate each with the idea of moving backwards or moving forward?          Note: Order of variables was randomized.   Regardless of your awareness of the following terms, which 3 terms appeal to you the most?  Click first on your best preferred option (which will be aligned with Choice 1), then click on your second preferred option (which will be aligned with Choice 2), and then click on your third preferred option (which will be aligned with Choice 3). To undo a choice, click on the term you would like to change. Green economy   _____ Economic growth _____ Sustainable degrowth _____ Postgrowth _____ Steady state economy _____  Which of the following terms appeal to you the least? Green economy Economic growth Sustainable degrowth Postgrowth Steady state economy  Note: Order of variables were randomized.     Moving Backwards Neutral Moving Forward Economic growth    Steady state economy    Green economy    Sustainable degrowth    Postgrowth      209  SECTION 10. In this section we are interested to know more about you. Please remember that your answers will remain anonymous.   What is your gender? a. Female b. Male  c. Other  What is your age? Please check one.  a. Under 18 b. 18-24 c. 25-34 d. 35-44 e. 45-54 f. 55-64 g. 65 or Above h. Prefer Not to Answer  What is the highest degree or level of education you have received? Please check only one.  a. Part of primary school b. Completed primary school c. Part of high school d. Completed high school e. Some college or university f. Received a college or technical school certificate g. Received a university bachelor’s degree h. Some graduate training i. Received a graduate university degree j. Other (please specify) ____________  In what province or territory do you live? a. Alberta b. British Columbia c. Manitoba d. New Brunswick e. Newfoundland and Labrador f. Nova Scotia g. Ontario h. Prince Edward Island i. Quebec j. Saskatchewan k. Northern Territories l. Nunavut m. Yukon  In what city, town or village do you live in? __________________ City, town or village   210  What is your occupation? (if you are a homemaker or a student please state this. If you are retired, or unemployed, please state this and list your former occupation)  What industry or sector do you work in? ___________________  Most people in Canada think of themselves as Canadians but also partly identify themselves based on the ethnic background of their ancestors. What would you say is the main ethnic background (or nationality) of your ancestors?  a. White b. Chinese c. South Asian (for example: East Indian, Sri Lankan) d. Black e. Arab f. West Asian (for example: Iranian, Afghan) g. Filipino h. Southeast Asian (for example: Vietnamese, Cambodian) i. Latin American  j. Japanese  k. Korean l. Aboriginal (that is, First Nations, Metis or Inuit) m. Other ______________________________  Below are listed several categories of income. Please choose the category that gives the best estimate of your total household income before taxes last year. a. No personal income      b. Under $5,000 c. $5,000 to 9,999 d. $10,000 to 14,999 e. $15,000 to 19,999 f. $20,000 to 24,999 g. $25,000 to 34,999 h. $35,000 to 44,999 i. $45,000 to 54,999 j. $55,000 to 64,999 k. $65,000 to 74,999 l. $75,000 to 84,999 m. $85,000 to 94,999 n. $95,000 to 114,999 o. $115,000 to 134,999 p. $135,000 to 149,999 q. $150,000 to 199,999 r. $200,000 to 249,999 s. $250,000 and over   211  In your political ideology, with which of the following political parties do you identify the most? Only choose one.  a. Conservative Party of Canada b. Liberal Party of Canada c. New Democratic Party (NDP) d. Green Party of Canada e. Bloc Québécois f. Other (please specify) ____________ g. None  Thank you very much for your participation! If you have any additional comments, please write them in the text box below.       212  Appendix G: Parameter estimates for ordinal regressions (main effects) G.1 Parameter estimates for the dependent variable hope  Parameter Estimate Std. Error 95% Wald Confidence Interval Hypothesis Test Exp(B) 95% Wald Confidence Interval for Exp(B) Lower Upper Wald Chi-Square df Sig. Lower Upper Threshold [Hope =1] -0.218 0.3449 -0.894 0.458 0.399 1 0.528 0.804 0.409 1.581 [Hope =2] 0.963 0.3472 0.282 1.643 7.686 1 0.006 2.618 1.326 5.171 [Hope =3] 2.839 0.3678 2.118 3.560 59.567 1 0.000 17.099 8.315 35.163 [Hope =4] 4.91 0.4936 3.942 5.877 98.948 1 0.000 135.600 51.538 356.772 [Frame 1] 0.883 0.2172 0.457 1.309 16.534 1 0.000 2.419 1.58 3.702 [Frame 3] 1.056 0.2153 0.634 1.478 24.060 1 0.000 2.875 1.885 4.384 [Frame 4] 0.666 0.2183 0.238 1.094 9.304 1 0.002 1.946 1.269 2.985 [Frame 2] 0a             1     [Audience segment] - Cluster 1 0.527 0.2082 0.119 0.935 6.415 1 0.011 1.694 1.127 2.548 [Audience segment] - Cluster 2 0.295 0.2135 -0.124 0.713 1.902 1 0.168 1.342 0.883 2.040 [Audience segment] - Cluster 3 0a             1     [Gender=1] - Female -0.014 0.1535 -0.315 0.286 0.009 1 0.925 0.986 0.730 1.332 [Gender=2] - Male 0a             1     [Political=1] - Green Party -0.645 0.3200 -1.272 -0.018 4.064 1 0.044 0.525 0.280 0.982 [Political=2] - Liberal Party -0.097 0.3006 -0.686 0.493 0.103 1 0.748 0.908 0.504 1.637 [Political=3] - NDP -0.188 0.3246 -0.824 0.449 0.334 1 0.563 0.829 0.439 1.566 [Political=4] - Conservative Party 0a             1     [Concern Environment =1] - Less concerned -0.237 0.1677 -0.566 0.091 2.001 1 0.157 0.789 0.568 1.096 [Concern Environment =2] - More concerned 0a             1     [Concern Economy =1] - Less concerned -0.149 0.1619 -0.467 0.168 0.850 1 0.356 0.861 0.627 1.183 [Concern Economy =2] - More concerned 0a             1     (Scale) 1b                     213  G.2 Parameter estimates for the dependent variable fear  Parameter Estimate Std. Error 95% Wald Confidence Interval Hypothesis Test Exp(B) 95% Wald Confidence Interval for Exp(B) Lower Upper Wald Chi-Square df Sig. Lower Upper Threshold [Fear =1] -1.978 0.3575 -2.679 -1.277 30.614 1 0.000 0.138 0.069 0.279 [Fear =2] -0.483 0.3478 -1.165 0.198 1.932 1 0.165 0.617 0.312 1.219 [Fear =3] 1.165 0.3601 0.459 1.871 10.468 1 0.001 3.206 1.583 6.492 [Fear =4] 2.867 0.4558 1.974 3.760 39.575 1 0.000 17.585 7.198 42.962 [Frame 1] -0.956 0.2197 -1.387 -0.526 18.949 1 0.000 0.384 0.250 0.591 [Frame 3] -0.909 0.2158 -1.332 -0.486 17.734 1 0.000 0.403 0.264 0.615 [Frame 4] -0.678 0.2176 -1.104 -0.251 9.704 1 0.002 0.508 0.331 0.778 [Frame 2] 0a             1     [Audience segment] - Cluster 1 -0.473 0.2093 -0.883 -0.062 5.100 1 0.024 0.623 0.414 0.939 [Audience segment] - Cluster 2 -0.269 0.2139 -0.688 0.150 1.584 1 0.208 0.764 0.502 1.162 [Audience segment] - Cluster 3 0a             1     [Gender=1] - Female 0.386 0.1571 0.078 0.694 6.044 1 0.014 1.472 1.081 2.002 [Gender=2] - Male 0a             1     [Political=1] - Green Party -0.210 0.3239 -0.845 0.425 0.421 1 0.516 0.810 0.430 1.529 [Political=2] - Liberal Party -0.245 0.3048 -0.842 0.352 0.647 1 0.421 0.783 0.431 1.422 [Political=3] - NDP -0.078 0.3286 -0.722 0.566 0.056 1 0.813 0.925 0.486 1.762 [Political=4] - Conservative Party 0a             1     [Concern Environment =1] - Less concerned -0.536 0.1712 -0.871 -0.200 9.804 1 0.002 0.585 0.418 0.818 [Concern Environment =2] - More concerned 0a             1     [Concern Economy =1] - Less concerned -0.709 0.1671 -1.036 -0.381 17.988 1 0.000 0.492 0.355 0.683 [Concern Economy =2] - More concerned 0a             1     (Scale) 1b                       214  G.3 Parameter estimates for the dependent variable anger  Parameter Estimate Std. Error 95% Wald Confidence Interval Hypothesis Test Exp(B) 95% Wald Confidence Interval for Exp(B) Lower Upper Wald Chi-Square df Sig. Lower Upper Threshold [Anger =1] -1.571 0.3635 -2.283 -0.858 18.675 1 0.000 0.208 0.102 0.424 [Anger =2] -0.648 0.3584 -1.351 0.054 3.272 1 0.070 0.523 0.259 1.056 [Anger =3] 0.852 0.3730 0.121 1.583 5.224 1 0.022 2.345 1.129 4.872 [Anger =4] 1.940 0.4275 1.102 2.778 20.598 1 0.000 6.959 3.011 16.084 [Frame 1] -0.753 0.2308 -1.206 -0.301 10.651 1 0.001 0.471 0.300 0.740 [Frame 3] -1.020 0.2342 -1.479 -0.561 18.984 1 0.000 0.361 0.228 0.570 [Frame 4] -0.525 0.2277 -0.972 -0.079 5.317 1 0.021 0.591 0.378 0.924 [Frame 2] 0a             1     [Audience segment] - Cluster 1 -0.358 0.2215 -0.793 0.076 2.616 1 0.106 0.699 0.453 1.079 [Audience segment] - Cluster 2 -0.229 0.2266 -0.673 0.215 1.024 1 0.312 0.795 0.510 1.240 [Audience segment] - Cluster 3 0a             1     [Gender=1] - Female -0.089 0.1682 -0.418 0.241 0.278 1 0.598 0.915 0.658 1.273 [Gender=2] - Male 0a             1     [Political=1] - Green Party -0.448 0.3304 -1.096 0.200 1.839 1 0.175 0.639 0.334 1.221 [Political=2] - Liberal Party -0.954 0.3144 -1.571 -0.338 9.218 1 0.002 0.385 0.208 0.713 [Political=3] - NDP -0.700 0.3392 -1.365 -0.035 4.259 1 0.039 0.497 0.255 0.965 [Political=4] - Conservative Party 0a             1     [Concern Environment =1] - Less concerned -0.514 0.1862 -0.879 -0.149 7.609 1 0.006 0.598 0.415 0.862 [Concern Environment =2] - More concerned 0a             1     [Concern Economy =1] - Less concerned -0.509 0.1809 -0.863 -0.154 7.919 1 0.005 0.601 0.422 0.857 [Concern Economy =2] - More concerned 0a             1     (Scale) 1b                        215  Appendix H: Parameter estimates for ordinal regressions (significant interaction effects) H.1 Parameter estimates for interaction effects for frame * audience segment for the dependent variable hope Parameter Estimate Std. Error 95% Wald Confidence Interval Hypothesis Test Exp(B) 95% Wald Confidence Interval for Exp(B) Lower Upper Wald Chi-Square df Sig. Lower Upper Cluster 1 Threshold [Hope =1] -0.487 0.2171 -0.912 -0.061 5.028 1 0.025 0.615 0.402 0.941   [Hope =2] 0.489 0.2171 0.063 0.914 5.071 1 0.024 1.630 1.065 2.495   [Hope =3] 2.246 0.2572 1.742 2.750 76.249 1 0.000 9.450 5.708 15.644   [Hope =4] 4.237 0.4557 3.343 5.130 86.436 1 0.000 69.172 28.317 168.971 [Frame 1] 0.498 0.2886 -0.068 1.064 2.975 1 0.085 1.645 0.934 2.896 [Frame 3] 0.456 0.2955 -0.123 1.035 2.384 1 0.123 1.578 0.884 2.816 [Frame 4] 0.184 0.2962 -0.396 0.765 0.387 1 0.534 1.202 0.673 2.149 [Frame 2] 0a             1     (Scale) 1b                   Cluster 2 Threshold [Hope =1] -0.146 0.2205 -0.578 0.286 0.439 1 0.507 0.864 0.561 1.331   [Hope =2] 1.039 0.2295 0.589 1.488 20.491 1 0.000 2.826 1.802 4.430   [Hope =3] 3.109 0.3162 2.490 3.729 96.723 1 0.000 22.407 12.058 41.638   [Hope =4] 6.070 1.0225 4.066 8.074 35.246 1 0.000 432.735 58.332 3210.250 [Frame 1] 0.417 0.3105 -0.191 1.026 1.805 1 0.179 1.518 0.826 2.790 [Frame 3] 0.844 0.3083 0.240 1.449 7.499 1 0.006 2.326 1.271 4.257 [Frame 4] 0.372 0.3119 -0.240 0.983 1.420 1 0.233 1.450 0.787 2.672 [Frame 2] 0a             1     (Scale) 1b                   Cluster 3 Threshold [Hope =1] 0.694 0.3182 0.070 1.318 4.757 1 0.029 2.002 1.073 3.735   [Hope =2] 1.895 0.3460 1.217 2.573 30.001 1 0.000 6.653 3.377 13.107   [Hope =3] 3.644 0.4170 2.827 4.462 76.365 1 0.000 38.253 16.893 86.624   [Hope =4] 5.433 0.6688 4.122 6.744 65.985 1 0.000 228.764 61.675 848.528 [Frame 1] 1.525 0.4354 0.672 2.379 12.271 1 0.000 4.596 1.958 10.790 [Frame 3] 1.938 0.4339 1.088 2.788 19.950 1 0.000 6.944 2.967 16.253 [Frame 4] 1.478 0.4260 0.643 2.313 12.037 1 0.001 4.384 1.902 10.104 [Frame 2] 0a             1     (Scale) 1b                     216  H.2 Parameter estimates for interaction effects for frame * gender for the dependent variable fear Gender B Std. Error 95% Wald Confidence Interval Hypothesis Test Exp(B) 95% Wald Confidence Interval for Exp(B) Lower Upper Wald Chi-Square df Sig. Lower Upper Female Threshold [Fear =1] -1.364 0.1863 -1.729 -0.998 53.561 1 0.000 0.256 0.177 0.368 [Fear =2] 0.064 0.1732 -0.275 0.404 0.139 1 0.710 1.067 0.760 1.498 [Fear =3] 1.568 0.2028 1.171 1.965 59.784 1 0.000 4.797 3.224 7.138 [Fear =4] 3.137 0.3410 2.468 3.805 84.622 1 0.000 23.027 11.803 44.924 [Frame 1] -0.984 0.2473 -1.469 -0.500 15.842 1 0.000 0.374 0.230 0.607 [Frame 3] -1.297 0.2571 -1.800 -0.793 25.443 1 0.000 0.273 0.165 0.453 [Frame 4] -0.656 0.2445 -1.135 -0.177 7.204 1 0.007 0.519 0.321 0.838 (Scale) 1c                   [Frame 2] 0b             1     Male Threshold [Fear =1] -0.512 0.2059 -0.915 -0.108 6.180 1 0.013 0.599 0.400 0.897 [Fear =2] 0.800 0.2087 0.391 1.209 14.708 1 0.000 2.226 1.479 3.352 [Fear =3] 2.720 0.2936 2.145 3.296 85.808 1 0.000 15.181 8.538 26.994 [Fear =4] 4.966 0.7284 3.539 6.394 46.482 1 0.000 143.488 34.417 598.215 [Frame 1] -0.259 0.2803 -0.809 0.290 0.855 1 0.355 0.772 0.446 1.337 [Frame 3] -0.345 0.2733 -0.881 0.191 1.592 1 0.207 0.708 0.415 1.210 [Frame 4] -0.326 0.2779 -0.870 0.219 1.374 1 0.241 0.722 0.419 1.245 (Scale) 1c                   [Frame 2] 0b             1         

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0357253/manifest

Comment

Related Items